Navo he 5-HTi antagonist devejop'ed with the patie --- Navohan.t,opisetron) Always once a day. Always 5 mg. A SANDO.Z Sandoz Pharma Ltd, Pharma Basle Operations, Marketing & Sales, CH 4002 Basle/Switzerland RADIOLOGY AND ONCOLOGY Established in 1964 as Radiologia Iugoslavica in Ljubljana, Slovenia. Radiology and Oncology is a journal devoted to publication of original contributions in diagnostic and interventional radiology, computerized tomography, ultrasound, magnetic resonance, nuclear medicine, radiotherapy, clinical and experimental oncology, radiophysics and radiation protection. Editor in chief Tomaž Benulic Ljubljana, Slovenia Associate editors Gregor Serša Ljubljana, Slovenia Viljem Kovac Ljubljana, Slovenia Editorial board Tullio Giraldi Branko Palcic Udine, Italy Vancouver, Canada Marija Auersperg Andrija Hebrang ]urica Papa Ljubljana, Slovenia Zagreb, Croatia Zagreb, Croatia Matija Bistrovic Durtla Horvat Dušan Pavcnik Zagreb, Croatia Zagreb, Croatia Ljubljana, Slovenia Haris Boko Laszlo Horvath Stojan Plesnicar Zagreb, Croatia Pecs, Hungwy Ljubljana, Slovenia Malte Clausen Berta Jereb Ervin B. Podgoršak Kiel, Germany Ljubljana, Slovenia Montreal, Canada Christoph Clemm Vladimir Jevtic Miran Porenta Miinchen, Germany Ljubljana, Slovenia Ljubljana, Slovenia Mario Corsi H. Dieter Kogelnik Jan C. Roos Udine, Italy Salzburg, Austria Amsterdam, The Netherlands Christian Dittrich Ivan Lovasic Horst Sack Vienna, Austria Rijeka, Croatia Essen, Germany Ivan Drinkovic Marijan Lovrencic Slavko Šimunic Zagreb, Croatia Zagreb, Croatia Zagreb, Croatia Gillian Duchesne Luka Mi/as Lojze Šmid London, Great Britain Houston, USA Ljubljana, Slovenia Bela Pomet Maja Osmak Andrea V eronesi Budapest, Hungary Zagreb, Croatia Gorizia, Italy Publishers Slovenian Medica[ Society -Section oj Radiology, Croatian Medica[ Association -Croatian Society oj Radiology Affiliated with Societas Radiologorum Hungarorum Friuli-Venezia Giulia regional groups of S.I. R. M. (ltalian Society of Medica! Radiology) Correspondence address Radiology and Oncology Institute of Oncology Vrazov trg 4 61000 Ljubljana Slovenia Phone: + 386 61 1320 068 Fax: +386 61 1314180 Reader for English Olga Shrestha Design Monika Fink-Serša Key words und UDC Eva Klemencic Secretaries Milica Harisch Betka Savski Printed by Tiskarna Tone Tomšic, Ljubljana, Slovenia Published quarterly Bank account number 501 OJ 678 48454 Foreign currency account number 50100-620-133-27620-5130/6 Nova Ljubljanska banka d.d. -Ljubljana Subscription fee for institutions 100 USD, individuals 50 USD. Single issue for institutions 30 USD, individuals 20 USD. According to the opinion of the Government of the Republic of Slovenia, Puhlic Relation and Media Office, the journal RADIOLOGY AND ONCOLOGY is a publication of informative value, and as such subject to taxation by 5 % sales tax. Indexed and abstracted by BIOMEDICINA SLOVENICA CHEMJCAL ABSTRACTS EXCERPTA MEDICA/ELECTRONIC PUBLISHING DIV/SION LUNG CANCER BIOLOGY AND CLINICAL ASPECTS 13-16 April 1994 LJUBLJANA -SLOVENIA LOCAL ORGANIZING COMMITTEE Chairman J. Orel Secretary M. Bitenc Treasurer M. Sok Members B. Hrabar V. Kovac T. Rott F. Šifrer S. Vidmar INTERNATIONAL SCIENTIFIC COMMITTEE AND INVITED SPEAKERS B. Corrin (U. K.) M. Mermolja (Slovenia) A. Debeljak (Slovenia) K. Moghissi (U. K.) M. Debevec (Slovenia) U. Pastorino (ltaly) D. Ferluga (Slovenia) M.I. Pelerman (Russia) P. Goldstraw (U. K) V. Pompe-Kirn (Slovenia) H.H. Hansen (Denmark) P. Rocmans (Belgium) K. Havemann (Germany) J.B. S5C. Patients' survival was strongly related to the number of stem lines, S-phase related fraction, percentage of tumor cells with an JOD> 5C and expression of binding sites to BLOOD A and BLOOD H. The expression of binding sites to blood-group antigens H was associated with decreased cellular heterogeneity as demonstrated by the percentage of tumor cel! nuclei with abnormal JOD. The analysed prognostic factors were found to be independent from cel! type and tumor stage (TNM). Key words: bronchial neoplasms-pathology; blood groups; DNA, neoplasm Introduction The following study was performed to assess the prognostic importance of blood-group anti­gens, macrophage inhibitory factor, DNA con­tents of tumor cell nuclei as defined by the integrated optical density (IOD), and syntactic Correspondence to: Prof. Klaus Kayser, M.D., Ph.D., Head, Department of Pathology, Thoraxklinik, Ama­IienstraBe 5, D-69126 Heidelberg, Germany. UDC: 616.24-076:612.118.221.2 structure analysis in operated bronchus carci­noma patients. The most important prognostic factors in surgically treated lung cancer patients are in descending order lymph node involve­ment (pN-stage), tumor size and its localization (pT-stage), inflammatory response of host tis­ 2• 3 sue, and invasion of bronchial cartilage. 1• Additional factors related to cellular hetero­geneity and proliferation (number of stem lines, percentage of proliferating tumor cells), and to the variability of tumor cell textures analyzed by syntactic structure analysis ( distance between Binding capacities to blood-group antigen.1· proliferating tumor cells, distance between tu­mor cells ancl immunocompetent cells, etc) have been suggestec1;4 however, cletailecl etata are stili missing. Liganclohistochemistry of bron­chial carcinoma revealecl a large variety of bincling capacities of carbohyclrates ancl signifi­cant clifferences in tumor cells comparecl to cells of healthy Jung parenchyma. s, 6 Expression of epiclermoicl growth factor (EGF) ancl that of corresponcling bincling sites coulcl be clemons­tratecl in nearly ali epiclermoicl ancl aclenocarci­noma cases.7 The cleterminants of ABH bloocl group antigens are carbohyclrate sicle chains of glycoproteins. Their expression is probably as­sociatecl with the clifferentiation of cancer cells. s, 9 In aclclition, Lee et al. 10 reportecl a favourable survival of Jung carcinoma patients whose tumors hacl not !ost their bloocl-group A antigens. Similar etata have been publishecl for patients sufering from epiclermoicl carcinoma of the heacl ancl neck.11 Liganclohistochemistry is a newly clevelopecl technique which pennits histochemically the clemonstration of biologi­cally important bincling capacities. This has alreacly been clemonstratecl by the macrophage inhibitory factor (MIF) , which is a lymphokine able to prevent the migration of macrophages from the capillary tubes in guinea pigs, ancl probably, a mecliator regulating the activity of 12 macrophages in host clefense, or sarcolectin (SAR) which is an interferon antagonist, ancl hacl been isolatecl ancl purifiecl from human placenta.13 For example, the presence ancl ex­pression of bincling sites to MIF ancl SAR coulcl be clemonstratecl in nearly ali analysecl cases with healthy human lung parenchyma;14 howe­ver in only a minority of human lungs of patients wtio clevelopecl autoimmune-relatecl cli­seases, such as biliary cirrhosis.15 -We have employecl the bincling capacities of BLOOD A, BLOOD B, BLOOD H in human lung carcinomas, ancl report their rela­tion to tumor celi type, ancl survival of the patients. In aclclition, their relationship to nuc­lear ancl structural features is analysecl. Material and methods This stucly inclucles 149 patients who unclerwent surgery for lung cancer with potentially curative resection at the Thoraxklinik Heidelberg. The tumor size, pTNM stage, ancl celi types of the resectecl ]obes ancl lungs were obtainecl as usual, i.e. the surgical specimens were fixecl with buferecl formalin, cut into serial sections ancl analysecl as clescribecl in cletail elsewhere. 5 The antiboclies ancl biotinylatecl probes were appliecl to 4-5 um thick histological slicles. After clepa­raffination ancl rehyclration through garcled al­cohol, and blocking of the enclogenous peroxi­dase by incubation in 0.1 % methanolic hyclro­gen peroxide for 30 min an equilibration with 0.1 M Tris buffer and treatment with 1 % bovine serum albumin was followed by incuba­tion with biotinylated probes ( oligosaccharides boncl to biotinylated polyacrylamide) at a con­centration 10 µ.g/ml at room temperature for 60 min. Visualization of the bincling capacities was performed by the application of ABC complex (Camon, Wiesbaden) and development of the chromogenic product from the substrates diami­nobenzidinehydrogen peroxicle. Finally, the sections were counterstainecl and mountecl. Po­sitive and negative controls were performed as usual, i.e. by parallel performance of positive cases, any by replacement of biotinylated pro­bes with BSA. Cases were consiclered to have stained positively only if a dark brown staining could be seen in ali or in clusters of tumor cells. The survival of patients was assessecl by repeat­ed questionnaires to the house physicians or physicians responsible for postsurgical tumor care of patients. The analysis of survival rates was perfonned by the Kaplan-Meier and log­rank tests. 16 IOD measurements were perfor­med on Feulgen stained histological slides using an automated image analysing system (DIAS, Towersoft, Berlin). The extractecl features in­cluded IOD, nuclear size, S-phase relatecl frac­tion as definecl 2. 75 < IOD < 3.25, percentage of tumor celi nuclei with an IOD < 5C ( < 5C), number of stem lines, 2CV standard deviation, entropy as defined by Stenkvist, 17 and IOD entropy-flouss ( current of entropy) as describ­ed. 18 The features of syntactic structure analysis were obtained by construction of the minimum Kayser K et al. Table 1. Synopsis of material (number of patients) blood-group A antigen was seen in 68/149 cases, Feature Celi Type that of blood-group B antigen in 82/149 cases, Epidermoid Adeno Large Celi Small Celi and that of blood-group H antigen in 72/149 Men 27 23 cases respectively, without any preference to a Women 5 13 8 3 certain celi type. No association of binding of T-Stage blood group trisaccharides to tumor stage, celi type, or sex could be established. The median T-l 10 18 7 8 T-2 21 14 22 15 T-3 T-4 N-Stage N-0 N-1 N-2 N-3 14 8 3 l survival rates as defined by Kaplan-Meier esti­ 1 2 2 2 mations for the patients, including the number 17 12 15 8 of deaths, grouped according to the expression 12 8 5 7 of blood-group A antigen, blood-group H anti­ 11 9 10 8 6 13 5 3 gen, and IOD features (number of stem lines, Tota! 46 42 26 1OD-entropy, 1OD-entropiefluss ), structural spanning tree, and comprised mmnnum dis­tance between of centers of ali tumor celi nuclei (DMST), proliferating tumor celi nuclei only, tumor celi nuclei with an IOD < 5C only, bet­ween tumor celi nuclei and immunocompetent cells, MST entropy and MST entropy-fluss as described originally.18 The use of external and interna! standards provided a reproducibility of measurements as described previously. 19 Results A synopsis of patients and tumor celi types included in the study is given in Table l. Ali in ali, primary bronchial carcinoma of 149 pa­tients (121 men and 28 women) could be analys­ed. The relation of positive stained cases in respect to the celi type and IOD and MST features is given in Table 2. Expression of features (MST-entropy, MST-entropiefluss), are given in Table 3. The differences between the different median values of survival rates marked with an asterisk are at a statistically significant leve! (p < 0.05). Discussion Lung function depends on complex interactioHs between many celi types, their matrix, and organized preserved morphological structures. Development of Jung cancer, i.e. the occurence of a new uncontrolled proliferating biological system within the preexisting biological system (lung) may retain, or loose certain cellular properties of the original system, or express new characteristics. Some of these properties may be of advantage for cellular propagation while others may not. It has been demonstrated by various authors that the expression of carbo- Table 2. Expression of histo blood-group antigenes in relation to !OD and MST features (mean values). Feature Histo Blood Group A-A+ B-B+ H-H+ (82) Number of cases (81) (68) (72) IOD -features Stem Lines 2.9 3.0 2.7 S-Phases (%) 12 12 12 12 11* Entropy 2.2 2.2 2.2 2.2 2.2 2.2 2CV-Std (%) 11 9 10 10 II 9 5C Exc (%) 7.1 7.6 7.6 8.1'' Structural features (MST) Dist Tu-Tu (um) 12 12 12 12 12 12 Dist Pr-Pr (um) 40 38 39 Dist 5C-5C (um) 25* 28* 26 27 26 27 Entropy l1 9 10 10 11 9 Entropy-flow 29 28 30 24* * Statistically significant (p < 0.05) Binding capacities to blood-group antigem· Table 3. Expression of histo blood-group antigens, A antigens, and found no influcence of blood !OD-, and MST features, and median survival of group B and blood group H antigens on survi­patients (in months). val. Our data are in agreement with the pre­Feature Patients Deaths Survival vious reports with respect to blood-group A (N) (N) (months) Blood A-81 antigen expression: the survival of patients with Blood A + 68 30 12* 37 .-expression of blood-group A antigens of tumor Blood B-67 12 Blood B + 82 16 cells is favourable compared to that of patients without this tumor characteristics.21 28 Blood 1-1-77 1. In our ma­ Blood 1-1 + 72 n• terial, this observation holds true for both pa­ l Stem line 28 4 .· tients with blood type A and those with other 2 Stem lines 32 6 30* blood types. In our material, patients with > 2 S tem lines 89 59 11 * expression of blood-group H antigens had < 5 % S-phases 10 2 a 5-IO 0/c, S-phases 45 3 > 1 O% S-phases 94 prolonged survival compared to those with loss 64 of blood-group H 10 MST Entropy-flow of blood-group B had no influence. These data 55 33 11* < 10 MST Entropy-flow > 10 < 30 Entropy-flow 54 are in accordance with the findings reported by 20 An influence of blood types on 14 n• 26 II' Miyake et al. 29 JI* > 30 MST .ntropy-flow survival of patients could not be seen in our material. Miyake et al. 22 developed two monoc­ * Statistically significant ( < ().05) hydrate-binding capac1t1es of Jung carcinoma cells differs from that of cells of a normal lung in a broad range.4 Patients with epidermoid carcinoma and adenocarcinoma who have a favourable prognosis compared to small celi or large celi carcinoma patients, commonly express a broad panel of carbohydrate binding capaci­ 4 ties in contrast to other celi types.3• The expression of carbohydrate binding capacities in human lung cells reflects to both injury and differentiation of the corresponding cells. 3 The expression of blood group antigens is closely associated with that of carhobydrate-binding capacities. The loss of blood-group A antige­neity has been reported to be an unfavourable prognostic indicator in various cancer types. In patients suffering from head/neck cancer Wolf and Carey11 observed a median survival of 8.1 months in tumors with corresponding loss of blood-group A antigens, and a median survival of 38 months in those with retained antigens. 11 Similar observations have been reported by Mijake et al. 20 in Jung carcinoma patients ( ex­pression of blood-group H antigens) and by Lee et al. 10 ( expression of blood-group A anti­gens ). Lee et al. 10 could, however, only demons­trate the prognostic significance of blood group lonal antibodies that recognize migration-inhibi­tory-related carbohydrate sequences. The aut­hors reported a significantly improved survival of patients whose Jung tumors exhibited no detectable binding sites to the antibodies. In accordance with our findings, these data again support the idea, that carbohydrate-related cel­lular binding capacities may play an important role in tumor propagation and tumor celi inte­ractions with their environment. This hypothe­sis is supported by the findings that nuclear abnormalities (abnormal chromosome content) is associated with loss of trisaccharide blood A and H binding. The MST entropy-fluss as esti­mated from the Onsager equation can be assu­med to be a stable indicator for the "distance of a thermodynamically open system from its equilibrium", i.e. reflects the amount of heat produced inside the tumor which has to be removed through its surface. Similar to the number of stem lines, S-phase related tumor celi fraction, percentage of tumor celi nuclei > SC, this factor, as well as blood-group A antigeneity, blood-group H antigeneity, and expression of binding sites to sarcolectin are associated with an improved survival of poten­tially curable bronchial carcinoma patients tre­ated by surgery, and may thus be added to the list of prognostic factors in Jung cancer. Kayser K et al. References l. Kayser K. Blilzebruek H, Probst G, Vogt-Moy­kopf I. Retrospective and prospective tumor stag­ing evaluating prognostic factors in operatcd bron­chus carcinoma patients. Cancer 1987; 59: 335-61. 2. Kayser K. Analytical lung pathology. Heidelberg, New York, Springer, 1992. 3. Kayser K, Liewalcl F, Kremer K, Tacke M. Inte­grated optical density (!OD), syntactic structure analysis, and survival in operatecl lung carcinoma patients. Pathol Res Pract 1994, in press. 4. Kayser K, Stute H, Tacke M. Minimum spanning tree, integrated optical clensity and lymph nodc metastasis in bronchial carcinoma. ANACAL 1993; 5: 225-34. 5. Kayser K, Heil M, Gabius HJ. Is the profile of bincling of a panel of neoglycoproteins useful as a cliagnostic marker in human lung cancer? Pathol Res Pract 1989; 184: 621-9. 6. Kayser K, Gabius HJ. Neoglycoproteins and lect­ins in human lung cancer. In: Gabius HJ, Gabius S (ecls) Lectins and Cancer Heidelberg, New York, Springer, 71-84. 7. Kayser K, WeiBe G, Gabius HJ, Bubenzer J, Eberstein M, Hintze T. Differentiation-relatecl expression of epidermal growth factor receptors in human lung carcinoma dernonstratecl histoche­mically by biotinylated EGF. Modem Pathol 1990; 3: 327-31. 8. Lloyd KO. Blood group antigens as markers for normal clifferentiation and malignant change in human tissues. Am J Clin Pathol 1987; 87: 129-39. 9. Coon JS, Weinstein RS. Blood-group related an­tigens as markers of malignant potential ancl hete­rogeneit y in human carcinomas. Human Pathol 1986; 17: 1089-106. 10. Lee JS, Ro JY, Sahise AA, Hong WK, Brown BW, Mountain CF, Hittelmann WN. Expression of bloocl group antigen A: a favourable prognostic factor in non-srnall celi Jung cancer. N Engl .! Med 1991; 324: 1084-90. 11. Wolf GT, Carey TE. Tumor antigen phenotype, biologic staging, ancl prognosis in head ancl neck squamous carcinoma . ./ Natl Cancer Inst Monogr 1992; 13: 67-74. 12. Pozzi LM, Weiser WY. Human recombinant mi­gration inhibitory factor activates human macro­phages to kill tumor cells. Celi fmmwwl 1992; 145: 372-9. 13. Zeng FY, Weiser WY, Kratzin H, Stahl B, Karas M, Gabius HJ. The major bincling protein of the interferon antagonist sarcolectin in the human placenta is a macrophage migration inhibitory factor. Arch Biochem Biophys 1993; 303: 74-80. 14. Kayscr K, Zeilinger C, Zeng FY, Gabius S, Gabius HJ, Weiser WY. Detection of the lympho­kine migration inhibitory factor in normal and clisease-affected Jung by antibody and by its major bincling protein, the interferon antagonist sarcolec­tin. Pathol Res Pract 1993; 189: 992-5. 15. Kayser K, Bohrer M, Kayser C, Weiser WY, Zeng FY, Gabius HJ, Tuengerthal S, Schulz V. Alteration of human lung parenchyrna associatecl with primary biliary cirrhosis, Zentralbl Pathol 1993; 139: 377-80. 16. Kaplan EL, Meier P. Nonparametric estimations from incomplete observations. Am Stat Ass .! 1958; 53: 457-81. 17. Stenkvist B, Strancle G. Entropy as an algorithm for the statistical description of DNA cytometric clata obtained from image analysis microscopy. Anacel 1990; 2: 159-66. 18. Kayser K, Kremer C, Tacke M. Integrated optical density and Entropiefluss ( current of entropy) in bronchial carcinoma. In Vivo 1993; 7: 387-92. 19. Kayser K, Stute H, Bubenzer J, Paul J. Combined morphological and syntactic structure analysis as tools for histomorphological insight into human lung carcinoma growth. Anacel 1990; 2: 167-78. 20. Miyake M, Hakomori S. A specific celi surface glycoconjugate controlling celi motility: evidence by functional monoclonal antibodies that exhibit celi motility and tumor celi metastases. Biochemi­stry 1991; 30: 3328-34. 21. Kayser K, Bovin NV, Korchagina EY, Zeilinger C, Zeng FY, Gabius HJ. Correlation of expression of binding sites for synthetic blood group A-, B­and H-trisaccharides and for sarcolectin with survival of patients with bronchial carcinoma. Eur J Cancer 1994; 30A: 653-657. 22. Miyake M, Taki T, Hitomi S, Hakomori S. Cor­relation of expression of H/LeY/Leb antigens with survival of patients with carcinoma of the lung. N Engl .! Med 1992; 327: 14-8. Radio/ Oncol 1994; 28: 287-9. Methodology and results of bronchopulmonary cancer detection in Slovenia 1970-1992 Jurij Šorli Institute far Respiratory Diseases, Golnik, Slovenia The results of bronchopulmonary cancer detection in pneumonology departments represent over 90 % of alf microscopically verified cases of bronchopulmonary cancer in Slovenia. The incidence rose from 370 in 1970 to 782 in 1992, with male prevalence (85 %). Passive detection was preferentially used as mass fluoroscopy has been gradually abandoned after 1972. Bronchoscopy was always the main diagnostic method (92.l % in 1970 and 90.7 in 1992), followed by perthoracic fine neddle aspiration biopsy (3.8 % and 4, 5 %). Other methods (sputum cytology, mediastinoscopy, thoracoscopy, thoracotomy) accounted for 4.1 and 4.8 % respectively. There is a slight predominance of central type ( 53. 7 vs. 46.3 % in 1992). Planocellular type was predominating (68.9 % in 1970, 39.8 % in 1992), followed by small-cell (10.1 and 19.2 %), adeno-(8 and 16.5 %), and large-cell (7 and 13.6 %) carcinomas. In 1992 surgery was the therapy of choice in 26,3 %, chemotherapy in 15 % and radiotherapy in 22,5 % of newly detected cases. Key words: lung neoplasms-diagnosis; Slovenia lntroduction Today, bronchopulmonary carcinoma is the most prevalent type of cancer, and is responsi­ble for more cancer-related deaths than any other tumour in man. In some countries in the world it has surpassed breast carcinoma in women as the most lethal tumour. 1 These facts stress the importance of research into the modalities and results of Jung cancer detection in every country facing rise in the incidence of bronchopulmonary carcinoma, in Corrcspondcncc to: Prof. Jurij Šorli, M.D., Ph.D., Institute for Rcspiratory Discascs, 64204 Golnik, Slo­vcnia. UDC: 616.24-006.6-07 order to find the way to detect and properly diagnose lung carcinoma as early as possible, and to institute adequate therapy. In Slovenia, in contrast to some other coun­tries, diagnosis and partly also therapy of bron­chopulmonary cancer is almost exclusively car­ried out in the pneumonology departments and out-patient clinics. Our aim was, therefore, to investigate methodology and evaluate the re­sults of lung cancer detection in these units in the period from 1970 to 1992. Material and methods The data from annual Golnik Epidemiology Reports were used as a source, together with 288 Šorli J the data from individual pneumonology depart­ments of regional hospitals in Slovenia. Statistical evaluation was performed by means of SAS software. Results Figures la and lb present newly detected pa­tients with lung cancer by age and sex. There LEGEND; Agogroup [ill O-« D ..... ... ;/1 1910-n 197a.85 1986-92 PERIOD Figure la. Distribution of newly detected male paticnts with lung cancer according to agc (1970-92). LEGEND: Agogroup [m O·« D ..... ... * 1910-n 1978-85 1986-92 PERIOD Figure lb. Distribution of ncwly dctectcd fcmale pa­tients with Jung cancer according to age (1970-92). is an obvious increase in the number of patients, especially in the 45-64 year age group in both males and females. Passive detection was the most frequently used mode of detection throughout the obser­ved period: 82.3 % in 1970-77 and 93.4 % in 1986-92 period. Mass X-ray method made an important contribution (14.7 % ) in the period 1970-77 only; after 1972 it has been gradually abandoned because of a high cost-benefit ratio in tuberculosis detection. Diagnostic methods were not grossly changed in the observed period, with bronchoscopy lea­ding and others (perthoracal fine needle biopsy, sputum cytology, mediastinoscopy, thoracosco­PY, thoracotomy) contributing less than 10 % (Figure 2). cf.. 10 1970-77 1978-85 1986-92 PERIOD PTFNB Pcrthoracic fine ncedlc biopsy. Figure 2. Diagnostic methods for vcrification of lung cancer. Cytology alone or in combination with histo­logy was the most frequent method of cancer verification. In 1970-77, cytology alone was used in 60 % of cases, with gradual decrease to 51 % in 1986-92. In the same periods both cytology and histology were used in 35 % and 46 % of cases respectively. Histologic type of the detected carcinomas (Figure 3) significantly changed over tirne, with a decrease in squamous and an increase in adeno-and partly also small-and macrocellular types. The detection of carcinoma required ade­quate therapy. The progress in therapeutic de­cision making in 1970-92 period is shown in Figure 4. Methodology and results of lmmchopulmonary cancer detec/ion in Slovenia 1970--/992 * PERIOD Figure 4. Moclcs of thcrapy in ncwly clctcctccl paticnts. Discussion Growing trends in lung cancer incidence are characteristic for industrial developed coun­tries. 2 It is quite sad that by this standard Slovenia is a well developed, industrial country. The incidence of Jung cancer in the age group 45-64 years increased by almost 50 % , mostly in men, but with an important contribution of female population in the last third of the obser­ved period. Smoking could be blamed for 90 % of causes in males but only for 73 % in female population as observed also by others.3 It is clear from our data that the increase in lung cancer incidence could not be ascribed to a better or modified mode of detection. Also in the period of mass X-ray, more than 50 % of detected patients already had symptoms of the disease. Changes in bronchoscopy, with prevalent use of flexible instruments have dramatically chan­gecl the distribution of lung cancer histology. As more peripheral tumours were being diagno­secl, the percentage of squamous type steadily clecreased and of adeno increased. This pattern was observed also by others.4·5 The increase in large and small celi types was less significant and mostly due to better classification.6 In contrast to other authors, we used cytology more frequently as the only method of cancer 7 verification.This was mostly due to a better access to this type of service. In the last obser­vation period both methods have been used with more balanccd use of different thcrapies, with only a small proportion of patients being treated only symptomatically. It is most promi­sing that thc increase is grcatest in surgery and chemotherapy as thesc two methods contribute 8 most to thc survival of lung cancer patient. References l. Beckett WS. Epiclcmiology and ctiology of lung canccr. In: Matthay RA cel. Lung cancer. Phila­delphia: WB Saundcrs, 1993; 1-15. 2. Whclan SL, Parkin DM, Masuycr E ccls. Pattems of cancer in .five continents. Lyon: IARC, 1990 (IARC Sci Pubi; 102). 3. Garfinkcll, Stcllman SD. Smoking ancl Jung canccr in womcn: Finclings in a prospcctivc stucly. Cancer Res 1988; 48: 695 l. 4. Zavala DC. Diagnostic fibcroptic bronchoscopy. Clzest 1975; 68: [2-19. 5. Popp W, Rauschcr H, Ritschka L, ct al. Diagno­stic scnsitivity of diffcrcnt tcchniqucs in thc cliag­nosis of lung tumours with flcxiblc fibcroptic bron­choscopy. Cancer J 991; 67: 72-75. 6. Truong LD, Undcrwoocl RD, Grccnbcrg SD, ct al. Diagnosis and typing of Jung carcinomas by cytopathologic mcthocls. Acta Cytol [985; 29: 379­81. 7. Mcrmolja M. Possibilitics and limitations of cyto­logy in thc cliagnosis of lung tumors. Radio/ Oncol 1994; 28: (in prcss). 8. Narukc T, Goya T, Tsuchiya R, ct al. Prognosis ancl survival in rcscctcd Jung carcinoma basccl on thc ncw intcrnational staging systcm. J Thorac Cardivasc Surg 1988; 96: 440-7. Radio/ Oncol 1994; 28: 290-7. Epidemiological features of Jung cancer in Slovenia Vera Pompe-Kirn, Maja Primic Žakelj, Neva Volk Institute of Oncology, Zaloška 2, Ljubljana, Slovenia Lung cancer is the most frequent cancer in the world with wide geographical variations in risk. In Europe its incidence trends in men are decreasing in the most affected countries such as Scotland and Finland, increasing moderately in Eastern Europe, and increasing steeply in Southern Europe. The incidence trends in women are increasing everywhere. Many risk factors have been identified, and the overwhelming role of tobacco smoking has been repeatedly demonstrated. According to the data of the Cancer Registry of Slovenia in the tirne period 1961-1990, the incidence of lung cancer in Slovenia was increasing. In the 80's the increase was moderate in men and steep in women. The cumulative rates in men were in the middle of those established for selected European states and regions while the rates in women were at the bottom. The results of the birth cohort analysis indicated a stabilisation of the rates in men and further increase of the rates in women. About 25 % of cases in both sexes were diagnosed in a localised stage. In men the percentage of the localised stage was increasing by age which was explained by a decreasing percentage of the more aggressive small cel! carcinoma. More squamous cel! carcinomas were registered in men, and more adenocarcinomas in women. The observed survival of lung cancer patients was around 7 % for men and 6 % far women, and has not changed since 1970. In 1989, 42 % of adult men and 24 % of adult women in Slovenia were smokers. In the period 1975-1994, the percentage of smokers was decreasing in men, and increasing in women. These results are a challenge for more efficient antismoking campaigns, especially among women. Key words: Jung neopJasm-epidemioJogy; SJovenia Introduction Lung cancer is now the most frequent cancer in the worJd. Pisani, Parkin and Ferlay estima­ted that in the year 1985 the worJd burden had been: 896 000 new Jung cancer cases (677 000 in men and 219 000 in women) and 785 000 Correspondence to: Professor Vera Pompe -Kirn, MD, PhD, Institute of Oncology, Zaloška 2, 61105 Ljublja­na, Slovenia. Phone: + 38661 13 24113. UDC: 611.24-006.6-0 36.2 deaths due to lung cancer (600000 in men, and 185 000 in women). 1 There are wide geographicaJ variations in the risk. In Europe alone the crude incidence rates are varying from 180-45/100 000 in men, and from 55-5/100 000 in women.2 In men the time­trends in incidence are different. They are decreasing in most affected countries such as Scotland, EngJand and FinJand, increasing mo­derateJy in Eastern Europe, and increasing steepJy in Southern Europe. In women the trends are increasing everywhere.3 Epidemiological f eatures of lung cancer in Slovenia Standard methods in descriptive epidemio­ logy were used.7 Crude incidence rate has been defined as the rate of total annual number of new cases per 100 000 popuJation in the relevant year. Comulative incidence rate is a speciaJ age-standardised rate. It is the sum over each year of age of the age-specific incidence rates taken from birth to age 74. It can be interpreted either as a directly age standardised rate with the same popuJation size in each age-group, or as an approximation to the cumulative risk. Staging was based on the international cancer registries regulation. According to this reguJa­tion, ali investigation methods including surgery A ro' ro":J roCo roOJ :'.'\'1, :'.'\' ro' R>o .... Oj .... Oj .... Oj .... Oj .._OJ .._OJ .._OJ .._ .._05 .._05 .._05 Year Figure 1. Crudc annual Jung canccr incidcncc ratcs by scx, Slovcnia 1961-90. Study of the epidemiology of lung cancer has been one of the most rewarding aspects of medica! research in the past 40 years. It has been shown how a disease that has become the most common type of cancer throughout the world can be made to become relatively rare.4 Many risk factors have been identified: tobacco smoking, atmospheric pollution, occupational hazards, ionising radiation, some familial, gene­tic and other host factors as well as the protect­ive role of diet rich in fruit and vegetable. The overwhelming role of tobacco smoking in the causation of lung cancer has been repeatedly 5 demonstrated. 4• The aim of our study was to analyse lung cancer incidence in Slovenia in depth and to compare the obtained results with the available 0 "qj "qj "qj "qj , ,<1> "qj Year l ioio:'.\'l, :'.\0 :'.\'o fb " fb. . ?>0 "qj , "qj Year Figure 3. Cumulativc annual lung canccr inciclcncc ratcs by scx, Slovcnia 1961-90. ancl with Norway and Sweden where the rates were the lowest 3-4/100. With the cumulative rates in women (0.9/100), Slovenia was placecl at the bottom, very close to different regions of Germany (Figure 5). In Slovenia itseJf, the cumuJative rates varied. In men high rates were found in inclustrial and mining settJements, whiJe in women they were founcl in the capitaJ region of Slovenia. 9 In the commune of Idrija stable high rates were found in both sexes. To preclict properly further trends of Jung cancer in SJovenia a detailecl analysis on trencls in the inciclence of specific age-groups by tiine periods (Figure 6a, b) and by birth cohorts was neecled (Figure 7 a, b). In men the peak of the rates plotted by tirne periocls shiftecl to the left ancl a stabilisation in the eighties was observed. Epidemiological features of lung cancer in Slovenia NL, Eindhoven PL, Lower Silesia UK, Scotland CZ, Boh.& Moravia ltaly, Trieste Slovakia D, Saarland Finland H, Szabolcs UK, South Thames D, former GOR CH, Geneva S/ovenia Denmark Bel aru s S, Tarragona Norway Sweden Figure 4. Cumulativc avcragc annual lung canccr incidcncc ratcs in men, Europc 1983-87. In women the peak remained in the oldest age groups and the curve was higher in each subse­quent time period. We tried to explain such a different behaviour by sex with a birth cohort analysis. In the birth cohort analysis (Figure 7) we presented the same < 0.1 O''\I>< 1>< I>< < [!,\() r.,"-r.,'0 ." .\() ." ,05 ,05 , , , , , , ,05 ,05 , . r.,"-r.,'0 ." .\() ." ,05 ,05 , , , , , , ,05 ,05 ,05 Year of birth Figure 7b. Agc-spccific lung canccr inciclcnce ratcs in womcn by birth cohots, Slovenia 1961-90. N=59 N=54 N=104 N=144 N=175 N=163 N=155 N=151 N=110 0% 25% 50% 75% 100% il localised ¦ regional remote O unknown 80+ Figure 9. Stage clistribution of Jung canccr in womcn by agc, Slovcnia 1983-90. Epidemiological features of lung cancer in Slovenia N 168 N 256 N 601 N 1036 N 985 N 710 N-604 N -533 N -263 II squnII10us li.111;1llcell [] adeno u1 nonsp i Jn1ln 11011sp Figure 10a. Histologic types of lung cancer in men by age Slovenia 1983-90. N" 53 N 52 N 88 N 117 N 1,19 -'.=,-----·-·····i N 137 N 171 N 13<1 N 10\J Figure 10b. Histologic types of lung cancer in women by agc, Slovenia 1983-90. 60· llf rnen [Jwomcn 3 . .'b ,0 ,'b'l, CJ:>fb g,'b ...,_ OJ g,0 ,0 l #,cs , Year Figure 11. Pcrccntagcs of smokcrs by scx, Slovcnia 1975-94. Reference l. Pisani P, Parkin DM, Ferlay J. Estimates of the worldwide mortality from eighteen major cancers in 1985: implications for prevention and project­ions of future burden. lnt I Cancer 1993; 55: 891-903. 2. Cancer incidence in five continents. Vol 6. IARC Sci Pubi 1992; No 120. 3. Trenc/s in cancer incidence and mortality. IARC Sci Publ 1993; No 121. 4. Doli R. Introduction and overview. In: Samct JM cd. Epidemiology of lung cancer. New York: Marcel Dckker Inc, 1994: l-14. 6. Žakelj M. Zdravstvena prosvetljenost polnoletnih Slovencev o raku. Specialisticna naloga. Ljubljana: Onkološki inštitut, 1991. 7. Cancer registra/ion: principles and methods. IARC Sci Publ 1991; No 95. 8. Statisticni letopis R Slovenije. Ljubljana: Zavod R Slovenije za statistiko, 1993. 9. Pompe-Kirn V, Primic-Žakelj M, Fcrligoj A, Škrk J. Zemljevidi incidence raka v Sloveniji 1978-1987. Ljubljana: Onkološki inštitut, 1992. 10. Tcctcr SM, Holmes FF, McFarlanc MJ. Lung carcinoma in thc cldcrly population: influence of histology on thc invcrsc rclationship of stagc to age. Cancer 1987; 60: 1331-6. 11. Mason TJ. Thc descriptivc epidemiology of lung canccr. In: Samct JM cel. Epidemiology of lung cancer. New York: Marcel Dckkcr Inc, 1994: 51-69. 12. Canccr: causes, occurrcncc and control. IARC Sci Pub] 1990; No 100. Radio/ Oncol 1994; 28: 298-300. Epidemiologic data of lung cancer during 1984-1993 in Albania Elez Selimi, Shaqir Karaulli, Perlat Kapisyzi, Hektor Cocoli, Sokol Mulosmani University Hospital of Pulmonary Disease, Tirana, Albania Since we cannot attain greatness, let us have our revenge by railing at it. (Montaigne) During the last ten years in Albania, the incidence of lung cancer was 10 per 100.000 inhabitant. The diagnosis was based on conventional examinations, biopsy taken through fiberbronchoscopy, or by using other invasive methods. In 75 % of cases the histopathologic examination has determined the cytologic type; 45 % of cases belong to the 55-64 year age group, and 27.1 % of those were over 65 years old. According to the place of living, there were 50.9 % of cases from urban areas, and f!).l % from rural areas. Considering that the greatest part of Albanian population lives in rural areas, the frequency of lung cancer is greater in urban areas. The smoking is the main risk factor, accounting for 88 % of cases. Radical surgical treatment was done only in 9 % of cases and palliative surgical treatment was performed in 3.5 % of cases as the patients seek medica! help in too advanced stages of diseases. Key words: lung neoplasms-epidemiology; Albania Introduction Many epidemiologic studies have concluded that lung cancer is a disease the prevalence of which has increased after the turn of the centu­ry, especially after 1950. 1-4 Many attempts have been made to stop this disease of modem times. Though medicine have failed to cure this disease, it is railing by hundred and thousand attempts to treat it as easily as is treating tuberculosis. Correspondence to: Prof. Elez Selimi, M.D. Ph.D., Hospital of Pulmonary Disease (Sanatorium), Tirana, Albania. UDC: 616.24-006.6-036.22 Among these efforts, the epidemiologic ones take the first place, being the starting point of ali others. Material and methods In our country, the diagnostic methods of lung cancer have been as follow: X-ray examination; cytologic examination of sputum; fiberoptic bronchoscopy; bronchial biopsy; diagnostic tho­racotomy. This study is the first one in the epidemiology of lung cancer in Albania. We have studied some of the main epidemiologic characteristics of lung cancer in the last ten years (1984-1993), such as incidence, incidence by sex; by age; incidence rates in urban and Epidemiologic data of lung cancer during 1984-1993 in Albania rural areas, smoking-specific, male-female ratio and the percentage of cases treated by radical and palliative resection. During the studied period, 3000 lung cancer cases were diagnosed. A geographic variation study in occurrence of lung cancer was carried out. Results and discussion The incidence of lung cancer during this period was 10, and has not been found to vary not any evident by years (Figure 1). Lung cancer 14 11 ll 1985 198.ll Year O 1984 -1986 1991 --1992 1993 Figure 1. Lung canccr incidcncc ratcs (pcr 100 000). in our cases is more frequent in males, (8.7 per 100 000) than in females (1.23 per 100 000) (Figure 2). 12 Year .... Male ... Female Figure 2. Lung canccr incidcncc by scx (pcr !00 000). Different epidemiologic studies have pointed out the predominance of Jung cancer in urban areas compared with rural ones.5 Our = 65 yr Figure 4. Lung cancer incidencc rates by age, (per 100 000). Within Albania, the geographic variation in the frequency of lung cancer is as follows: the disease is more frequent in Tirana, Elbasan, Durres and in north districts than in South areas. This is explained by the fact that in Selimi E et al. Tirana, Elbasan and North districts there is mineral and metal Industry and mines. Risk factors for lung cancer A lot of former facts can be explained by risk factors. Since Adler6 in 1912 first suggested a relationship betwcen smoking and Jung cancer, there are many studies which have provided evidence that is now regarded as conclusive in establishing the causal link between cigarette smoking and lung cancer. The cigarette smoking as a risk factor encom­pas multiple aspects of smoking behaviour, including age at start the nmnber of cigarettes smoked, the product smoked, and inhaling pat­tern. In our study, 57 % of cases were male smokers, 0.6 % female smokers, O% male non­smokers and 12.4 % female non-smokers (Fi­gure 5). -Female nonsmokers, 4 -Male nonsmokers Figure 5. Smoking and lung cancer. Specific malc-fc­male. Almost ali women in our cases with Jung cancer were non-smokers but there was a family history of smoking. Besides smoking as an important risk factor in Jung cancer, there are other non-smoking risk factors, such as asbestos, radon, and other exposures, diet, hormone, acquired host charac­ 68 teristics and genetic factors.3· · From 3000 cases with lung cancer, only 9 % have been treated by surgical resection and 3.5 % by palliative surgical procedures. These results can be explained by a single reason: These late diagnosis of the disease. The histological types of Jung cancer in our patient have been as follow: Epidermoid 47 % ; adenocarcinoma 13 % ; small-cell Jung cancer 7 % ; Iarge-cell lung cancer 5 % ; carcinoid 1 % ; secondary malignant versament 3 % . In 24 % of cases there is no data on the histologic type. Conclusions During the last ten years, the incidence of lung cancer has remained unchanged being 10. per. 100 000 population. This incidence shows a tendency to decline in the middle-aged and those above 65 years old. Thc cigarette smoking is a main risk factor of Jung cancer, accounting for the majority of lung cancers in our country. The male-to-female ratio of incidence was 7:1. The predominance of lung cancer in men can be explained by the fact that ali of them were smokers. As a result of late diagnosis only 9 % of lung cancer cases were treated by surgical resection ancl 3.5 % by palliative interventions. References l. U.S. Department of health and Human Services. Smoking and f-lealth. Government Printing Office, 1964: 1103. 2. White C. Research on smoking ancl lung cancer: a lanclmark in the history of chronic disease epide­miology. Yale I Biol Med 1990; 63: 29-46. 3. Samet JM. The epiclemiology of lung cancer. Chest 1993; 103 ( 1 ): 205. 4. Morgan W, Hales M. Bronchogenic Carcinoma. In: Baum G ed. Textbook of Pulmonary Diseases. Boston: Little Brown and company, 1983 : 1046. 5. Shy C. Lung Cancer and the urban environment: a review. in: Finke! Ay, Duel WC, eds. Clinical implications of air pollution research. Acton, Mass: Publishing Science Group: 1976; 3- 6. Haddad R, Massaro D. ldiopathic diffuse intersti­tial pulmonary fibrosis. Am .1 Med 1968; 45: 211-19. 7. Hinds MW, Cohen HI, Kolorel LN. Tuberculosis and lung cancer risk in smoking women. Am Rev Respir Dis 1992; 125: 776-78. 8. Saracci R. Thc interaction of tobacco smoking ancl other agents in canccr ctiology. Epidemiol Rev 1987; 9: 175-93. Radio! Oncol 1994; 28: 301-8. Prognostic factors in non-small cell lung cancer Jens Benn Sfi)rensen Department oj Oncology, Finsen Center/National University Hospital, Copenhagen, Denmark The literature on prognostic factors evaluated in multivariate analyses in patients with non-small cel! lung cancer (NSCLC) was reviewed. Twenty studies inclucling a total of 3500 resectecl patients revealecl that definite prognostic variables in the resectable patient group were performance status, stage, category of tumor size and location (T), and category of lymph node involvement (N). Fourteen studies including a total of 5875 patients with inoperable NSCLC, inclucled in chemotherapy tria/s with or without radiotherapy, revealed that solely performance status and stage were definite prognostic variables in this non-resectable patient group. Possible predictors of /ong survival were low LDH, jemale gender, and high plasma albumin leve/, while weight loss, histology, and age were of minor importance. A large pari of the variation in survival outlook remains unexplained, calling for further stuc/ies on the prognostic influence of biological features of the tumors. Key words: carcinoma, non-small celi Jung; lung neoplasms; prognosis Introduction The current treatment results for non-small celi lung cancer (NSCLC) clearly call for improved therapy and also for careful selection of patients for the treatment options from which they are most likely to benefit. A detailed knowledge of prognostic factors, meaning variables with a well-established relation to the prognosis, is important for achieving these goals. Any clinical tria] must therefore include as assessment of the possible influence of these prognostic fac­tors on the therapeutic results. Correspondence to: Jens Benn SQ>rensen, M.O., Deparlmenl of Oncology, Finsen Ccntcr/National Uni­vcrsity Hospital, 9 Blcgclamsvcj, DK-2100 Copcnha­gcn. Dcnmark. Fax: + 45 35 45 69 66. The endpoints against which the prognostic variables are tested are usually disease-free survival and overall survival for surgically treat­ed patient populations, and response rate and survival for 11011-resectable patients. Analyses with response duration or quality of life as endpoints have been infrequently cornpared with prognostic variables. This is especially the case for quality of life, which for this reason will not be clealt with in the following. The ideal prognostic test divides the patients into two or more groups with very different outlook ancl, accorclingly, without any overlaps. However, this situation is relatively uncornmon in clinical practice and a more likely situation is using a prognostic test which is able to divide the patients into "high-risk" and "low-risk" subgroups with significant differences in out­ UDC: 616.24-006.6-08-036 look. However, there is stili a high degree of 302 S6rensen J B uncertainty, since not ali the "high-risk" pa­tients will recur and not ali the "low-risk" patients escape the bad outcome. If such test results were used to decide treatment, many patients would receive an appropriate therapy, but not without the occurrence of some over­treatment and some under-treatment. One way to overcome this obstacle and make a more accurate prediction of outcome is to consider a number of predictors simultaneou­sly. 1 This theme was de alt with in a consensus report on prognostic factors in NSCLC in Bru­ges in 1993.2 Besides stressing the importance of different, specific variables on prognosis, the report also emphasized the need for multiva­riate analyses to determine independent signifi­cant predictors of prognosis. Having defined which prognostic factors are the most useful and reproducible, it then becomes possible to generate a prognostic index for individual pa­tients. Such an index would combine all accept­ed prognostic factors into a single statement of outcome. It is recognized that, in spite of multivariate evaluations, the current variables describing anatomic stage, clinical, histological, and clinical chemistry features do not competely predict the prognosis and a large fraction of the variability remains unexplained. 1 One possible explanation for this phenome­non is that the most frequently used prognostic variables are in reality epiphonomena of the trne cellular and molecular characteristics of the disease, and little is known about the biolo­gical model of the disease itself. Thus, more knowledge on these cellular and molecular cha­racteristics is needed. The objective of the following is to provide an update on the current knowledge of progno­stic variables in NSCLC which have been esta­blished through multivariate analyses, and which have a documented impact on endpoints. Material and methods An update on the current knowledge of progno­stic variables in NSCLC which have been esta­blished through multivariate analyses is given in the following. Studies are reviewed provided that they are describing prognostic factors for survival solely in NSCLC patients and provide clear descriptions of the variables included in the multivariate analyses. The studies are divid­ed according to whether the study populations consisted of resectable patients or non-resecta­ble patients. Results presented in abstracts are not included. Results Multivariate analyses of prognostic factors far survival in resectable NSCLC 22 Twenty studies3 -including a total of 3500 patients with resected NSCLC are summarized in Table l. It is apparent that none of the variables have been evaluated in ali 20 studies. Definite prognostic factors for Iong survival include good performance status, low stage, and low lymph node category (N). Also low tumor category (T) was a major predictor of survival in many trials, but the variables descri­bing T and N were not independent, significant predictors of survival in ali studies evaluating these factors. This observation might be explai­ned by the phenomenon that T and N categories may be less important in Cox multivariate re­gression analysis when stage of disease is inclu­ded in the analyses as well. In most studies including these variables, there was a significant prediction by either T and N or by stage. Possible prognostic factors in Table 1 include variables which have not been evaluated exten­sively, but have been atributed an independent impact on survival in half or more of the studies. This is the case for postoperative infec­tions, perioperative blood transfusions, and DNA ploidy. The impact of infections, including empyema, pneumonia, or wound infection in the postope­rative period was evaluated by Gail et al. 3 while Deslauriers et al.7 included patients with major postoperative complications, such as pleura­pneumonia, bronchopleural fistulas and respira­tory failure into one group. Both trials reported Prognostic factors in non-small celi lung cancer a significant deterioration of survival outlook, a finding which, however, most probably is Iess associated with the biological characteristics of the malignant disease than with the treatment. The effect of perioperative blood transfusion has been a significant prognostic variable in two 18 studies. 8· Both studies included exclusively stage I patients and observed 5-year survival rates of 53 % and 62 % for patients with blood transfusion and 81 % and 76 % for patients without, respectively. Thus, any perioperative transfusions significantly worsens the patients prognosis. This association may be due to an adverse effect of the transfusion itself, or may serve as a marker for another, yet undetermi­ned, risk factor. The ploidy status of the deoxyribonucleic acid (DNA) has been investigated in five stu­dies13 · 17· 19• 21• 22 and was an independent predic­tor of survival in three studies. 17• 19• 22 Survival was poorest in patients with DNA aneuploid tumors, and the same pattern, though not sig­nificant, was observed in one of two negative studies21 but not in the other. 13 A number of variables have been relatively intensively evaluated and shown to be of impor­tance in only few of the studies. These variables may thus be of minor importance as predictors of survival in resected NSCLC patients and are outlined in Table l. This was the case for variables describing age, gender, histology, hi­stologic degree of differentiation, and extent of Table l. Prognostic factors in 20 multivariate analyses including a total of 3500 resected NSCLC patients. Prognostic factor No. of positive studies total no. of studies evaluated Definite Performance status 3/3 Stage T 5/13 N Possible Postoperative infection 2/2 Perioperative blood transfusion 2/2 DNA ploidy 3/5 Minor importance Age 1/8 Gender 1/9 Histology 3/12 Differentiation 1/4 Extent of resection Preliminary Tumor giant cells 1/1 Plasma celi infiltration 1/1 Solid ACL/other ACL subtypes 1/1 Satellite pulmonary nodules 1/1 FEV1 1/1 WBC Hemoglobin leve! preoperatively 1/1 Intratumoral blood vessel invasion 1/2 Ras mutation 1/1 p53 mutation 1/1 p53 protein expression 1/1 Tumor-associated antigen 43-9F 1/1 Proliferative activity 1/2 Celi line stablishment Abbreviations: ACL, adenocarcinoma of the lung; FEV 1, forced expiratory volume; WBC, white blood celi count. 304 Sd,rensen J B surgery (usually evaluated as lobectomy against larger resection). With respect to histology, both Gail et al. 3 and Deslaurier et al. 7 observed an independent and significant prediction for long survival among patients with squamous celi carcinoma as opposed to patients with non-squamous hi­stology, while these observations were not con­firmed in a study by Lipford et al. 4 However, the latter study reveaied a significantly worse prognosis for patients having large celi carcino­ 13-15-17-20, 22 ma. In contrast, 9 studies9 • did not find any independent prognostic impact of hi­stology. Overall, these 12 studies suggest that the influence of different histologic types of resected NSCLC is only minor when also other variables are taken into consideration. Other less frequently evaluated prognostic factors are outlined in Table l. Being evaluated in one tria! only, or being positive in one tria! only, none of these can be claimed to be established prognostic factors, but should be further evaluated for confirmation or rejection as predictors for survival in this group of pa­tients. Multivariate analyses of prognostic factors for survival in non-resectable NSCLC Fourteen studies23-63 including a total of 5875 patients with inoperable NSCLC included in chemotherapy trials with or without radiothe­rapy have reported data on multivariate anaiys­es of variables predicting survival. The progno­stic variables carrying significant and indepen­dent information are outlined in Table 2. Included in this table is also a large recent study evaluating independent prognostic factors among 1565 patients treated with radiotherapy in four clinical trials.37 Solely performance sta­tus has been evaluated in ali 15 studies in Table 2 and was significant in 14 studies, indicating that performance status is stili the best docu­mented prognostic variable in this patient group. The new international staging system, as des­cribed by Mountain38 has been evaluated in six multivariate analyses, of which it carried inde­pendent and additional significant information 34-37 on survival outlook in five.30• Thus, the importance of the new staging system has now been firmly documented as an independent prognostic factor for survival both among resec­ted patients and among non-recetable patients receiving other treatment modalities. Other variables, which may be possible pre­ dictors of good prognosis, though evaluated . less intensively, are low LDH, female gender, and normal plasma albumin leve! (Table 2). Variables having minor importance, as indicat­ed by lack of significant influence in the majo­rity of studies in which they are evaluated, include weight loss, histology, and age (Table 2). The relative importance of other potential variables should be further explored in non-re­ sectable patients. This holds trne for a number of biochemical variables, such as white blood celi count (WBC), aspartic aminotransferase (AST), hemoglobin, and calcium. However, these variables may have a less logical theore­ tical background for having a prognostic impact than the anatomical stage of the disease has. The impact of chemotherapy as opposed to best supportive care and of cisplatin-based chemo­ therapy against cytostatic treatment without cis­ platin has been evaluated by Rapp et al.28 and by Albain et a!.31 respectively. The Canadian multicenter study by Rapp et al. showed both an enhanced survival for patients receiving che­ motherapy and, in addition, that chemotherapy as opposed to best supportive care only -was an indepent predictor of survival among the 137 patients evaluated in the study.28 Albain et al. analysed data of 2531 patients with inoperable NSCLC from the Southwest Oncology Group treated from 1974 to 1988.31 The use of cisplatin was an additional independent predictor of im­ proved outcome_ in multivariate analyses after adjustments for year of accrual and ali other prognostic variables. Although the effect of cisplatin was modest, the data allow the conclu­ sion that it is real and not due to concurrent occurrence of other favourable prognostic varia­ bles. These data allow a cautious optimism regarding the development of more effective Prognostic factors in non-small celi lung cancer Table 2. Prognostic factors in multivariate analyses of non-resectable NSCLC including 14 studies with a total of 5875 paticnt trcated with chcmothcrapy or chcmothcrapy plus radiothcrapy and one study including 1565 patients trealcd with radiothcrapy alone. Prognostic factor No. of positivc studics/total no. of studics cvaluated Chemotherapy ± radiotherapy Radiotherapy alone (n = 5875) (n = 1565) Definite Pcrformancc status [3/14 lil New international staging Possible LDH 1/1 Gender 7/l 1 1/1 Albumin Minor importance Weight loss I-listology Age Preliminary Tstage N stage Mctastases to Liver Bone Subcutaneous 0/1 cxtrathoraeic metastasis WBC AST Hemoglobin Calcium Chemotherapy/BSC Cisplatin/no cisplatin 3/11 1/ l 2/7 1/9 1/1 1/1 ]/l 1/1 3/7 2/6 lil 1/1 1/2 J/2 1/2 1/1 1/1 1/l Abbreviations: LDH -lactatc dehydrogenase; WBC •·· white blood celi count; BSC -best supportive care. chemotherapy against this disease, but the issue should also be further elucidated. Biological characteristics In addition to stage of disease and perfonnance status of the patient, many other factors are important for predicting the prognosis in NSCLC. These factors include the biological properties inherent in the tumor cells themsel­ves (Table 3). The majority of these factors are at present preliminary as they have usually not been evaluated against other variables in multi­variate testing. However, the literature is ra­pidly expanding in this field, as recently outlin­ed by Szabo and Mulshine.39 These biological characteristics should be subjected to further evaluation in multivariate settings for asses­sme11t of prognostic impact. Discussion Based 011 a review of 20 multivariate studies including a total of 3500 patients with complete resected NSCLC it appears that definite prog­nostic factors in the resectable patient group are perfonnance status and stagc. Also T and N status are important, though Iess so when stage is included in the analysis, serving as a summarizing variable. The two former variables are also the most important amo11g patients with 11011 resectable disease based on a review of 14 studies including a total of 5875 patients. Variables describing age, histology, gender, and 306 Sibrensen J B Table 3. Biological charactcristics attributed progno­stic impact among NSCLC patients in univariate ana­lysis. Variablcs Neuroendocrine differentiation Oncogenes K-ras mutation p53 mutation L-myc mutation Oncogene products p53 protein c-crb B-2 protein p 185"0" protein DNA ploidy Carbohydrate antigens ABH blood group LeY and Lcb antigenes Lex antigen (4C9 antigen) Dolichos biflorus agglutinin (DBA) binding sitc Growth factors Epidermal growth factor receptor (EGFr) Prolifcrative activity Ki67 related antigen weight loss have been confined with a progno­stic impact in fewer studies than performance status and stage. A large number of variables describing biolo­gical features of the tumors have been confined with a prognostic impact among NSCLC pa­tients in univariate analyses. This is the case for variables such as neuroendocrine differen­tiation, oncogenes, and oncogene products, DNA-ploidy status, presence of carbohydrate antigens, growth factors, and proliferative acti­vity of the tumors. Further studies are warrant­ed in order to further document the value of these variables in order to more accurately predict the prognosis of the patients with NSCLC. In conclusion, severa! prognostic variables have been indentified to have an impact on the endpoints for clinical studies of NSCLC. Howe­ver, a large part of the variability in the prog­nosis among patients stili remains to be explain­ed. The results from studies on the cellular and molecular characteristics of the tumors open prospects for scientifically well-founded forms of therapy in the future, which may in tiine lead to a better therapy for patients with an otherwise poor prognosis. References l. Cox, DR. Rcgression modcls and lifc-tables. J Roy Stat Soc 1972; 34: 187-200. 2. Feld R, Bongcr M, Giner V, Ginsberg R, Harper P, Klastersky J, Lacquet L, Paesman M, Payne D, Rosell R, Sause W, Sculier J-P, Shaw E, S,»rensen JB, Splinter T, Stahel R, Bunn P. Prog­nostic factors in non-small celi lung cancer a concensus report. Lung Cancer (in press). 3. Gail MH, Eagan RT, Feld R et al. Prognostic factors in patients with resected stage I non-small ce]l Jung cancer. A report from the Lung Cancer Study Group. Cancer 1984; 54: 1802-13. 4. Lipford HH, Sears DL, Eggleston JC, More GW, Littlemore KD, Baker RR. Prognostic factors in surgically reseeted limited-stage, non-small celi carcinoma of the lung. Am J Surg Pathol 1984; 8: 357-65. 5. Chastang C, Lebeau B, Charpak Y, Decroix G. Prognostie faetors from a ranclomized clinical tria! in resected lung cancer. Stat Med 1985; 4: 279-85. 6. S,»rensen JB, Badsberg JH. Prognostie factors in resected stage I and II aclenocarcinomas of the lung. A multivariate regression analysis of 137 conseeutive patients. J Thorac Cardiovasc Surg 1990; 99: 218-26. 7. Deslaurier J, Brisson J, Cartier R et al. Carcinoma of the lung. Evaluation of satellite nodules as a faetor influeneing prognosis after resection. J Tho­rac Cardiovasc Surg 1989; 97: 504-12. 8. Little AG, Wu H-S, Ferguson MK et al. Periope­rative Bloocl Transfusion Adversely Affects Prog­nosis of Patients with Stage I Non-Small-Cell Lung Cancer. Am J Surg 1990; 160: 630-2. 9. Haracla M, Dosaka-Akita H, Miyamoto H, Kuzu­ maki N, Kawakami Y. Prognostic Significance of the Expression of ras Oncogene Product in Non­Small Celi Lung Cancer. Cancer 1992; 69: 72-7. 10. Fontanini G, Macchiarini P, Pepe S et al. The Expression of Proliferating Celi Nuclear Antigen in Paraffin Sections of Peripheral, Nade-Negative Non-Small Celi Lung Cancer. Cancer 1992; 70: 1520-7. 11. Macchiarini P, Fontanini G, Hardin JM, Pingitore R, Angeletti CA. Most peripheral, node-negative, non-small-cell Jung eancers have low prolifcrative rates ancl no intrutumoral and peritumoral bloocl and lymphatic vessel invation. J Thorac Cardio­vasc Surg 1992; 104: 892-9. Prognostic factors in non-small celi lung cancer 12. Stipa S, Danesi DT, Modini C ct al. Thc impor­tancc of hcterogcncity and of multiplc sitc sam­pling in the prospectivc detcrmination of dcoxyri­bonucleic acid flow cytometry. Surg Gynecol Obst 1993; 176: 427-34. 13. Mwrkvc O, Halvorsen OJ, Skjaervcn R, Stange­land L, Gulsvik A, Laerum OD. Prognostic Signi­ficancc of p53 Protein Expression and DNA Ploidy in Surgically Trcated Non-small Celi Lung Carci­nomas. Anticancer Res 1993; 13: 571-8. 14. Horio Y, Takahashi T, Kuroishi T et al. Progni­stoc significance of p53 Mutations and 3p Dcle­tions in Primary Resectcd Non-Small Celi Lung Canccr. Cancer Res 1993; 53: 1-4. 15. Pena CM, Ricc TW, Ahmad M, Mcdendorp S. Significance of Pcriopcrativc Blood Transfusions in Paticnts Undergoing Resection of Stage I and II Non-small-ccll Lung Canccrs. Chest 1992; 102: 84-8. 16. Battifora H, Sorenscn HR, Mehta Pet al. Tumor­Associated Antigen 43-9F Is of Prognostic Valuc in Squamous Celi Carcinoma of thc Lung. Cancer; 70: 1867-72. 17. Liewalcl F, Hatz R, Stork M et al. Prognostic value of deoxyribonuclcic acid aneuploidy in pri­mary non-small-cell lung carcinomas and thcir metastascs. I Thorac Cardiovasc Surg 1992; 104: 1476-82. 18. Tarttcr PI, Burrows L, Kirschncr P. Pcriopcrative blood transfusion advcrscly affects prognosis after rcsection of Stagc I (subset NO) non-oat celi Jung canccr. .l Thorac Cardiovasc Surg 1984; 88: 659­62. 19. Zimmerman PV, Hawson GAT, Bint MH, Par­sons PG. Ploidy as a prognostic determinant in surgically treated lung cancer. Lancet 1987; 230: 530-3. 20. Alama A, Costantini M, Rcpetto L ct al. Thymi­dine Labcllcing Index as prognostic Factor in Resccted Non-Small Celi Lung Canccr. Eur J Cancer 1990; 26: 622-5. 21. van Bodegom PC, Baak JPA, Stroet-van Galen C et al. The Pcrccntagc of Ancuploid Cclls is Significantly Correlatcd With Survival in Accura­tely Stagcd Patients With Stage 1 Rcsccted Squa­mous Celi Lung Canccr and Long-Term Follow Up. Cancer 1989; 63: 143-7. 22. Ichinosc Y, Hara N, Ohta M ct al. 1s T factor of thc TNM staging system a predominant prognostic factor in pathologic stagc I non-small-cell Jung cancer? .l Thorac Cardiovasc Surg 1993; 106: 90-4. 23. Miller TP, Chcn TI, Coltman CA et al. Effcct of altcrnating combination chcmothcrapy on survi­val of ambulatory patients with metastatic largc­ccll and adcnocarcinoma of thc lung. A Southwest Oncology Group Study. J Ciin Oncol 1986; 4: 502-8. 24. Finkclstein DM, Ettinger DS, Ruckdcschel JC. Long-tcrm survivors in mctastatic non-small celi lung canccr: an Eastcrn Cooperativc Oncology Group study . .T Ciin Oncol 1986; 4: 702-9. 25. Einhorn LE, Loehrer PJ, Williams SD ct al. Random prospectivc study of vindesinc vcrsus vindcsinc plus high-dosc cisplatin vcrsus vindcsinc plus cisplatin plus mitomycin C in advanccd non­small-ccll Jung canccr. I Clin Oncol 1986; 4: 1037-43. 26. Evans WK, Nixon DW, Daly JM ct al. A rando­mized study of oral nutritional support vcrsus ad lib nutritional intakc during chcmotherapy for advanccd colorectal and non-small celi lung can­ccr. .T Ciin Oncol 1987; 5: 113-24. 27. O'Conncll JP, Kris MG, Gralla RJ ct al. Fre­qucncy and prognostic importancc of pretreatmcnt clinical charactcristics in paticnts with advanccd non-small celi lung canccr trcatcd with combina­tion chcmothcrapy. I Ciin Oncol 1986; 4: 1604-14. 28. Rapp E, Pater JL, Willan A ct al. Chcmotherapy can prolong survival in paticnts with advanccd non-small-ccll Jung canccr: report of a Canadian multiccntcr randomizcd tria!. J Clin Oncol 1988; 6: 633-41. 29. Sukurai M, Shinkai T, Eguchi E ct al. Prognostic factors in non-small celi Jung cancer: multircgrcs­sion analysis in thc National Cancer Center Hospi­tal (Japan) . .T Cancer Res Ciin Oncol 1987; 115: 563-6. 30. Swrcnscn JB, Badsbcrg JH, Olsen J. Prognostic factors in inopcrable adenocarcinoma of the lung: a multivariatc rcgrcssion analysis of 259 paticnts. Cancer Res 1989; 49: 5748-54. 31. Albain KS, Crowley JJ, LeBlanc M, Livingston RB. Survival dcterminants in cxtensivcstage non­smal]-ccll Jung canccr: thc Southwcst Oncology Group cxpcricncc. Ciin Oncol 1991; 9: 1618-26. 32. Shinkai T, Eguchi K, Sasaki Y ct al. A prognostic­factor risk index in advanccd non-small-ccll lung cancer trcatcd with cisplatin-containing combina­tion chcmothcrapy. Cancer Clzemother Plzarmacol 1992; 30: l-6. 33. Kojima A, Shinkai T, Eguchi E ct al. Analysis of thrcc-ycar survivors among paticnts with advanccd inoperablc non-small celi lung cancer. .Tpn I Ciin Oncol 1991; 21: 276-81. 34. Kawahara M, Furusc K, Kodama N ct al. A Randomized Study of Cisplatin Vcrsus Cisplatin Plus Vindcsinc for Non-Small Celi Lung Carcino­ma. Cancer 1991; 68: 714-9. 35. Bonomi P, Gale M, Rowland K ct al. Prc-trcat­ment prognostic factors in stagc III non-small celi lung canccr paticnts rccciving combined modality trcatment. Int .T Radia/ Oncol Biol Phys 1991; 20: 247-52. 36. Pujol JL, Cooper EH, Lehmann M ct al. Clinical cvaluation of serum tumour markcr CA 242 in non-small celi lung canccr. Br .T Cancer 1993; 67: 1423-9. 308 S6rensen J B 37. Graham MV, Gcitz LM, Byhardt R ct al. Compa­38. Mountain CF. A ncw intcrnationaJ staging systcm rison of Prognostic Factors and SurvivaJ Among for Jung canccr. Chest 1986; 89: 225-33. Black Paticnts and White Paticnts Trcatccl With 39. Szabo E, Mulshinc J. EpidcmioJogy, prognostic Irracliation for Non-Small-Ccll Lung Canccr. factors ancl prcvcntion of Jung canccr. Curr Opi­ J Nat Cancer lnst 1992; 84: 1731-5. nion Oncol 1993; 5: 302-9. Radio/ Oncol 1994; 28: 309-15. Transbronchial needle aspiration with fiberoptic and rigid bronchoscope in the diagnosis and staging of lung cancer Andrej Debeljak, 1 Milivoj Mermolja, 1 Janez Orel, 2 Tomaž Rott3 1 Institute for Respiratory Diseases Golnik, 2 Department of Thoracic Surgery, University 3 Medica! Centre, Ljubljana, Institute for Pathology, Medica! Faculty Ljubljana, Slovenia In the prospective study from May 1992 til! March 1994, the sensitivity of transbronchial neeclle aspiration (TENA) pe,formed means of Olympus flexible NA-2C and Storz rigid /0436 or 10438 needles was compared. Fourty patients were included into the study, 14 with already confirmed central lung cancer for the purpose of staging, and 26 patients because central lung cancer had been suspected. Thirty patients had lung cancer and five had metastases of extrapulmonary malignant tumours to the lung and mediastinal lymph nodes. in 5 patients no cancer was found. TENA of the carina was performed 30 times, of the trachea 7 times and of the right upper lobe bifurcation 8 times. in five patients, minor bleeding occurred after TENA with a rigid needle. The sensitivity of TENA for cancer with Storz needle was 64 % (21 positive out of 33) and 45 % with the Olympus needle (15 positive out of 33). The difference between the two needles was statistically not significant. in 12 patients the TENA was false negative. A satisfactory histological sample was obtained in 18 patients with Storz needle. in 6 patients the histologic diagnosis was carcinomai in 4 nonspecific inflammation and in one patient tuberculosis. TENA with a rigid needle gives some additional information in comparison with a flexible one. TENA helps establishing the diagnosis and staging of central lung cancer and improves the bronchoscopic diagnosis. Jts sensitivity is satL1factory and complications are few. Key words: Jung neoplasms-diagnosis; biopsy, needle-methods, bronchoscopy; neoplasms staging Introduction The transbronchial needle aspiration biopsy (TBNA) of mediastinal Iymph nodes was first Correspondence to: Doc. Andrej Debeljak, M.O., Ph.D., Institute for Respiratory Diseases Golnik, Uni­versity Medica! Centre Ljubljana, 64204 Golnik, Slo­ vcnia. UDC: 616.24-006.6-076 described by Schieppati. 1 The method was fur­ther developed and introduced into practice in the USA2 and Europe.3 It became commonly used with the introduction of flexible needles suitable for TBNA through fiberoptic broncho­scope.4 The fiberoptic bronchoscope has largely replaced the rigid one and only a few authors stili report cytologica!5 • 6 or histologicai7 exami­nations of TBNA by rigid bronchoscope. The efficiency of TBNA with fiberoptic broncho­ Debeljak A et al. scope is comparable with the results that had been previously obtained by the rigid instru­ment. 8 TBNA can often replace the surgical methods of staging, i.e. cervical mediastinoscopy and anterior mediastinoscopy in patients with Jung cancer. In general, it is agreed that those pa­tients, in whom malignant cells are found on 6910 TBNA are considered to be inoperable.•• The patient is certainly inoperable if the meta­stasis has overgrown the capsule, if the metasta­ses are multiple or are present on the contrala­teral side of the mediastinum.11-13 If only one micrometastasis into a mediastinal lymph node is present, the prognosis of patients who under­went surgery is better. 9 TBNA is used also in the diagnostics of necrotic endobronchial tumours in whom fre­quently only necrosis is obtained by forceps biopsy.14 In submucously growing bronchial tu­mours TBNA is successful as we!L 15-1 6 Good results were obtained also in the diagnostics of peripheral tumours.17 Even better results in the diagnostics of peripheral tumours were obtained by the use of needle brush.18 TBNA is success­ful also in the diagnostics of pulmonary metasta­ 19 ses from extrapulmonary malignant tumours. False positive results of TBNA are rare; they may be due to the aspiration of malignant cells from bronchial secretion. 20• 2 1 Transtracheal needle aspiration biopsy of a tumour of the right upper lobe has been described.2 • 23 Com­plications such as pneumothorax after TBNA of peripheral Iesions, 17 haemomediastinum,24 bacteriemia25 or smaller bleedings into the bronchi26 are rare. 27 The frequent false negative findings represent the greatest problem of TBNA. 28 In our study we tried to establish the sensiti­vity of TBNA with both a flexible and a rigid needle. We wanted to find out which of them provides better results, and whether TBNA improves the bronchoscopic diagnostics and sta­ging of patients with central Jung cancer and metastases of extrapulmonary malignant tu­mours into the lungs and mediastinal lymph nodes. TBNA related complications were stu­died as well. Material and methods The prospective study of TBNA was approved by the ethic commission of the Institute for Respiratory Diseases Golnik. From May 1992 till March 1994, 40 patients were included into the study. In 26 patients clinically and roentgenologically a central lung cancer was suspected (tumour visible by fiber­optic bronchoscope), while in 14 patients with comfirmed central Jung cancer staging was ne­cessary prior to the plonned surgical interven­tion. Thirty-two males and 8 females with the mean age of 53 years were included into the study. Prior to bronchoscopy, posteroanterior and laterni X-rays of the thorax were performed. In patients, candidates for surgical treatment, computed tomography (CT) of the thorax was performed as well, mostly after bronchoscopy. Bronchoscopy was performed with a fiberop­tic bronchoscope Olympus 20 D and a Storz bronchoscope No. 8. Premedication was Atro­pini sulphas 1 mg subcutaneously. We used local anaesthesia with Xylocain 4 % 5 ml in drops or transtracheally, together with the application of narcotic analgetic fentanyl citrate 0,1 mg intra­venously. In 16 patients, diazepam was given intravenously as an additional premedication. During the procedure, up to 20 ml of 2 % Xylocain was added through the working chan­nel of the bronchoscope, and the patient recei­ved 3 L of oxygen min by nasal catheter. Fiberoptic bronchoscope was introduced through the nose, through the mouth with an orotracheal tubus or through the channel of a rigid bronchoscope. TBNA was performed first with an Olympus NA-2C flexible needle. This 21 gauge, needle is 13 mm in length, with a side hole. In every site of aspiration biopsy the needle was introduced three times to the area of 1 to 2 cm. Afterwards aspiration biopsy of the same sites was performed also by a rigid straight needle 10436 of by a curved needle Storz 10438. This needle is of 22 mm Iength with 17 gauge. The carina was aspirated with the straight needle, and the trachea and the right upper lobe bifurcation with the curved Transbronchial needle aspiration biopsy needle. The flexible needle was introduced by a shaft through the working channel. When we had seen the top of the shaft, the needle :was protruded from the shaft and advanced by a quick thrust between two cartilages. Sometimes the patient helped the advancing of the needle by coughing. Even if this manoeuvre had not been successful, the needle was fixed on the upper part of the working channel of the bron­choscope, and both the needle and the bron­choscope were advanced. The rigid needle was thrusted into the wall between two cartilages and to the shaft. We aspirated by a 20 ccm syringe. TENA was te­chnically easier with the rigid needle. Then we examined the entire tracheobronchial system and obtained the forceps biopsy and brushing. The patients were monitored for 24 hours after bronchoscopy for possible complications. X-ray was not made routinely. The first and the second aspirations were expelled from the needle on to a glass slide. The third aspiration was rinsed in Haemacell (Hoechst) and the smears from Haemacell were prepared by cytocentrifuge (Shandon Cytospin II). The smears were stained by May Grtinwald Giemsa and Papanicolau. The aspirations in which the elements of lymph nodes were found together with a few epithelial cells were consi­dered as satisfactory. This criterion was sugge­sted by Baker. 29 Ali three aspirations obtained by one needle were considered as a single result. The particles of the tissue obtained by the rigid needle, were fixed in the 10 % formalin and examined by standard histological methods. The sensitivity was calculated as the quotient of the number of TENA with malignant cells in ali patients with metastases in the mediastinal lymph nodes. The sensitivity of TENA with both flexible and rigid needle was compared by Chi square test. Results From May 1992 till March 1994 TENA was performed in 40 patients by fiberoptic and rigid bronchoscopy. Thirty patients had primary Jung cancer, while 5 had metastases of extrapulmo-nary malignant tumours into the lungs and mediastinal Iymph nodes (Table 1). In 5 pa- Table l. Diagnoscs of paticnts with malignant tumours includcd into thc TBNA study. Lung canccr Mctastatic Jung Icsions Small celi 7 Adcnocarcinoma-brcast Largc celi 4 Adcnocarcinoma-pancreas Adcnocarcinoma 7 Adenocarcinoma-unknown Squamous celi 12 origin 1 Hodgkin's Disease 2 Tota! 30 Tota! 5 tients nonmalignant Jung diseases were found as follows: tuberculosis, sarcoidosis and connec­tive tissue disease, and pneumonia in two pa­tients, but no cancer. The site of the aspiration biopsy was the main carina in 30 patients, the carina of the right upper lobe in 8 and the trachea in 7 patients. In 3 patients the TENA was done both through the main carina and through the trachea, and in two patients through the main carina and through the carina of the right upper lobe. TENA with a Storz needle was positive in 60 % (21 of 35) and with an Olympus needle, at the same sites, in 43 % of patients (15 of 35). One of the aspirations obtained by the Olympus needle was cytologically suspicious. In the assessment of sensitivity it was included among the negative findings. There were no statistically significant differences between the sensitivity of TENA performed with either Storz or Olympus needles. In 6 patients, malig­nant cells were found in TENA by the Storz needle only. In no case the malignant infiltra­tion of mediastinal lymph nodes was confirmed by the Olympus needle alone (Table 2). Table 2. Comparison of Storz and Olympus needlc TBNA results in 33 patients with metastatic mcdiasti­nal lymph nodes. Storz Ncedle + Olympus needle + 15 O 6 12 312 Debeljak A The sensitivity of TBNA in patients with lung cancer and extrapulmonary malignant tumours with metastatic mediastinal lymph nodes is pre­sented in Table 3. Table 3. Scnsitivity of TBNA in Jung canccr and cxtrapulmonary malignant tumours with mctastatic mcdiastinal lymph nodcs. Tumour Positivc TBNA Scnsitivity (%) Srna!] celi 6 of7 86 Large celi 3 of 4 75 Adenocarcinoma 5 of7 71 Squamous celi 5 of 10 50 Lung canccr -total Wci. . Extrapulmonary 2cl5 ® malignant tumours Tota! 21 of 23 64 % In the bronchoscopic diagnostics, the forceps biopsy and brushing were used. The sensitivity was 77 % (27 of 35). In two patients with small cell lung cancer and in one patient with adeno­carcinoma, TBNA was the only bronchoscopic method confirming the malignoma and increa­sing the sensitivity of bronchoscopic methods to 86%. In 7 patients with no mediastinal involvement TBNA was negative. There were no false posi­tive findings, the specificity being 100 % . Among the patients with positive TBNA, 17 had bronchoscopic signs of extrinsic pressure on the trachea, carina or on the right upper lobe bifurcation at the site of TBNA. Only 4 patients had a normal bronchoscopic finding in the site of TBNA (Table 4). Table 4. Frequcncy of positive TBNA in relation to bronchial cxtrinsic comprcssion. TBNA positivc TBNA negative Extrinsic compression 71 % (17 of 24) 29 % (7 of 24) Without compression 36 % ( 4 of 11) 64 % (7 of 11) Tota! 60 % (21 of 25) 40 % (14 of 35) Tissue for histological examination was obtai­ned by Storz needle only in 40 % of patients (14 of 35) with malignant lung tumours. In 6 et al. patients histology confirmed metastases while in 8 patients it was negative. The histological and cytological type of malignoma were consi­stent in 5 cases; in one patient the histological diagnosis was large cell carcinoma while the cytological one was adenocarcinoma. In three TBNA with histological signs of inflammation, malignant cells were found only cytologicaly. Histologically, five patients had TBNA with no malignant tissue and aspiration with no malignant cells. In these patients, cartilage and other parts of bronchial wall were found histo­logically. These samples were deemed unsatis­factory. In 14 patients TBNA was without malignant cells. In 6 of these, the mediastinal node meta­stases were not found on surgery in 2. Other 12 (34 % ) cases were considered to have false negative TBNA. Mediastinal lymph node metastases were found in 4 patients; in 2 by pneumonectomy, in one by explorative thoracotomy and in one by left lower lobectomy. In two of them, the affected lymph nodes had not been accessible by TBNA: one in paraesophageal -N8 and the other in the lymph nodes of pulmonary ligament -N9. In 8 patients, metastases in the mediastinal Iymph nodes were confirmed: once by transtho­racic needle aspiration and twice by mediasti­noscopy. In 5 patients we indirectly estimated metastatic involvement of the mediastinal lymph nodes, which were roentgenologically enlarged. Two patients had Hodgkin's disease, one patient had small cell Jung cancer and two patients had distant metastases in the brain and parietal pleura. In one patient, right pneumonectomy was performed even though TBNA of the right upper lobe bifurcation revealed malignant cells. TBNA of the main carina was not performed. Metastases in the carina! Iymph nodes were found on pneumonectomy. In patients with benign Jung diseases in 4 cases (80 % ) tissue was obtained by a Storz needle. In one patient tuberculous Iymphadeni­tis and in the other nonspecific inflammation was found. In the third patient the cytological Transbronchial needle aspiration biopsy examination showed epitheloid cells and lym­phocytes. In this patient transbronchial lung biopsy showed nonnecrotising granulomas, which was consistent with the diagnosis of sar­coidosis. After TBNA with a Storz needle in 5 pa­tients, minor bleeding was observed which cea­sed spontaneously. Hemorrhage was observed 4 times after TBNA of the main carina and once after TBNA of the right upper lobe bifur­cation. Twenty-four hours after the intervention no other complications were observed. Discussion In our study the efficiency of TBNA ( 64 % ) is similar to that obtained by other authors: 34 % ,8 46 % ,26 71 % . 16 We consider that, in addition to the technical pe1formance of TBNA, the selection of patients is of great importance. In our patients the sensitivity was established only in those patients who had central Jung cancer visible by fiberoptic bronchoscope, and in whom roentgenograms showed enlarged lymph nodes in the mediastinum where metastases were confirmed or strongly suspected. The sen­sitivity in our study is therefore approaching that achieved by Shure and Fedullo, 16 whose criterion for inclusion into the study was submu­cous growth or peribronchial tumour. Even greater sensitivity may be reached if the pre­sence and site of enlarged lymph nodes in the mediastinum were proved by computed tomo­graphy. 22 According to the success in our pa­tients, CT is not considered to be essential prior to TBNA. Other authors are of the same opi­nion.Malignant cells were most frequently 31 found in TBNA in patients with signs of exter­nal pressure on the bronchi. This finding was in agreement with those of other authors confir­ming malignant cells in TBNA of such patients in 33 % , 8 53 % 31 or even 62 % . 26 The lowest sensitivity of TBNA, only 7 % , was described by British authors,32 probably because TBNA was applied during routine bronchoscopies and performed by a needle for sclerosation; only samples with abundant malignant cells were considered positive. We found that TBNA increased the bron­choscopic diagnosis of Jung cancer by almost 9 % (3 of 35). TBNA showed a similar impro­vement of bronchoscopic diagnosis also in other authors: by 7%,31 14%33 or even 18%.34 TBNA significantly improves the bronchoscopic diagnostics by forceps biopsy, brushing and washing.35 The sensitivity of TBNA with Storz needle was insignificantly higher from that with Olympus needle. Despite that, in 6 patients malignant cells were confirmed by aspiration with Storz needle only. We think that TBNA with a rigid broncho­scope has been abandoned because it requires general anaesthesia, while Iocal anaesthesia is considered more inconvenient for the patient,36 and not because it would be Iess efficient. The results of TBNA with rigid and fiberoptic bron­choscope were comparable.3• 8 We had similar results. By application of sedatives and modem narcotic analgetics, the patient can tolerate bronchoscopy either with a rigid or flexible instrument. We find it important that in 40 % of patients also samples of tissue for histological examina­tion were obtained by Storz needle which ena­bled the diagnosis of malignoma in 6, and of nonmalignant disease in 2 patients. The sensiti­vity of 20 % (8 out of 40) cannot be compared with that to new flexible needles for histological examination: 72 % ,37 80 %38 or even 85.5 % .39 The histological examination of the tissue in TBNA enables the differentiation of bronchial wall or peribronchial tissues involvement, espe­cially lymph nodes. Possibly, rare false positive results could be excluded as well. We are not quite sure that there were no false positive results among our TBNA findings be­cause the involvement of mediastinal lymph nodes in TBNA positive patients was not veri­fied by the most reliable methods of surgical staging,9 the fina! test being the follow up. To prevent possible aspiration of malignant cells from the bronchial secretion, we have followed the principles of other authors33 and performed TBNA first, and only afterwards on examina­tion of the bronchial system and forceps biopsy were performed together with brushing. Since Debeljak A et al. false positive TENA are rare, this method often allows omission of surgical staging. 39 However, the greatest problem is encounte­red with patients who have had negative finding of TENA; there were 14 ( 40 % ) such cases among our patients. In these patients, surgical staging is necessary if they are potential candi­dates for resection. Ey obtaining six aspirations from the same site, in each patient the efficacy of the method was improved. The expelling of the sample on a glass slide had been used as recommended by Ndukwu.40 According to our experience, washing of the needle with Haemacell is equally or even more efficient; namely in the third aspiration we frequently found malignant cells and they were more abundant. Due to technical reasons, the cytologist was unable to examine the material immediately so that in the case of negative aspiration the bronchoscopist might repeat TENA immediately as it had been re­commended by Davenport.41 Mediastinal metastases were found by TENA most frequently in patients with small-cell Jung cancer, followed in descending order, by those with large celi, adenocarcinoma and squamous celi Iung cancer. This is in agreement with most other studies. 31 As was found by other authors,6• 31• 33 also in our patients only minor bleedings during TENA occured. They were more abundant after aspi­ration biopsy by Storz needle, as it had been expected since this needle is longer and broa­der. We avoided aspiration biopsies of the posterior tracheal wall and the right main bron­chus since the biopsy in these sites is more hazardous. 42 The sensitivity of TENA by Olympus needle is satisfactory and comparable with TENA per­formed by Storz needle. Histological diagnosis of malignant and benign diseases is sometimes possible by Storz needle. TENA improves the bronchoscopic diagnosis of pulmonary malig­nant tumours and in many patients enables the staging of Jung cancer in an easy, cost-effective and safe way. References l. Schieppati E. La punction mediastinal traves del espolon traquel. Rev As Med Argent 1949; 663: 497. 2. Wang KP, Tcrry P, Marsh B. Bronchoscopic needle aspiration biopsy of paratracheal tumours. Am Rev Respir Dis 1978; 118: 17-21. 3. Loddenkemper R, Groser H, Brandt HJ. Thirty ycars cxpcricnce with pcrbronchial fine necdle aspiration. In: Nakhostecn JA, Maascn W eds. Bronchology. Haguc, Boston, London: Martinus Nijhoff, 1981: 299-301. 4. Wang KP, Brower R, Haponik F, Sigelman S. Flcxiblc transbronchial ncedle aspiration for sta­ging of bronchogenic carcinoma. Chest 1983; 84: 571-6. 5. Pongrac I, Roglic M, Mrakovcic M, Jakaša G. The value of cytomorphologic finding in transtrac­heal and/or transbronchial biopsy for the prognosis of bronchial carcinoma. Pluc Bol 1984; 36: 28-33. 6. Svcnsson G, Hambraeus GH, Pettersson KI. Die transbronchiale Feinnadelbiopsie bei der Diagnose parabronchialer Tumoren. HNO 1985; 33: 370-3. 7. Robinson CLN. An instrument for transbronchial biopsy through a rigid bronchoscope. Lancet 1986; 2: 195. 8. Schaberg T, Mai J, Thalmenn U, Loddenkemper R. Die transbronchialc Lymphknotcnpunktion liber das Fiberbronchoskop bei pulmonalen Nco­plasmcn. Prax Klin Pneumol 1986; 40: 306-11. 9. Ratto GB, Mereu C, Motta G. The prognostic significance of preoperative assessment of media­stinal lymph nodes in patients with Jung cancer. Chest 1988; 93: 807-12. 10. Wang KP. Flexiblc bronchoscopy with transbron­chial necdle aspiration: biopsy for cytology speci­mens. In: Wang KP ed. Biopsy techniques in pulmonary disorders. New York: Raven Prcss, 1989: 63-71. 11. Naruke T, Sucmasu K, Ischikawa S. Lymph node mapping and curability at various levcls of meta­stasis in resected Jung cancer. J Thorac Cardiovasc Surg 1978; 76: 832. 12. Lung Cancer Study Group. Should subcarinal lymph nodes be routinely examined in patients with non-small lung canccr? J Thorac Cardiovasc Surg 1988; 95: 883-7. 13. Martini N, Flehingcr BJ, Zama MB, Beattic EJ. Rcsults of resection in nonoat celi carcinoma of thc lung with mediastinal lymph node metastases. Ann Surg 1983; 198: 386-97. 14. Buirski G, Calverlcy P, Douglas NJ et al. Bron­chial needlc aspiration in the diagnosis of bron­chial carcinoma. Thorax 1981; 36: 508-11. Transbronchial needle aspiration biopsy 15. Rosenthal DL, Wallacc JM. Fine needle aspiration of pulmonary lesions via fiberoptic bronchoscopy. Acta Cytol 1984; 28: 203-10. 16. Shure D, Fedullo PF. Transbronchial necdlc aspi­ration in thc diagnosis of submucosal and peri­bronchial bronchogcnic carcinoma. Chest 1985; 88: 49-51. 17. Wang KP, Haponik EF, Britt EJ, Khouri N, Erozan Y. Transbronchial nccdlc aspiration of pcriphcral pulmonary nodulcs. Chest 1984; 86: 819-23. 18. Wang KP, Britt EJ. Nccdlc brush in thc diagnosis of lung mass or nodule through flexible broncho­scopy. Chest 1991; 100: 1148-50. 19. Gay PC, Brutincl WM. Transbronchial nccdle aspiration in thc practicc of bronchoscopy. Mayo Ciin Proc 1989; 64: 158-62. 20. Cropp AJ, DiMarco AF, Lankcrani M. Falsc-po­sitive transbronchial necdlc aspiration in broncho­genic carcinoma. Chest 1984; 85: 696-7. 21. Harrow EM, Oldcnburg FA, Lingcnfcltcr MS, Smith AM. Transbronchial ncedlc aspiration in clinical practicc. A fivc-year cxpcricnce. Chest 1989; 96: 1268-72. 22. Schcnk DA, Chasen MH, McCarthy MJ, Duncan CA, Christian CA. Potcntial falsc positive media­stinal transbronchial ncedlc aspiration in broncho­genic carcinoma. Chest 1984; 86: 649-50. 23. Schenk DA, Bowcr JH, Bryan CL ct al. Trans­bronchial necdlc aspiration staging of bronchogc­nic carcinoma. Am Rev Respir Dis 1986; 134: 146-8. 24. Kuccra RF, Wolfe GK, Pcrry ME. Hcmomediasti­num aftcr transbronchial ncedlc aspiration. Chest 1986; 90: 466. 25. Witte MC, Opal SM, Gilbcrg JG ct al. Incidcncc of fcvcr and bacteriemia following transbronchial needlc aspiration. Chest 1986; 89: 85-7. 26. Harrow EM, Oldenburg FA, Smith AM. Trans­bronchial needle aspiration in clinical practice. Thorax 1985; 40: 756-9. 27. Sherling BE. Complication with a transbronchial histology needle. Chest 1990; 98: 783-4. 28. Wagner ED, Ramzy I, Greenberg SD, Gonzales JM. Transbronchial fine necdle aspiration. Am J Ciin Pathol 1989; 92: 36-41. 29. Baker JJ, Solanki PH, Schenk DA, Pelt CV, Ramzy I. Transbronchial fine nccdle aspiration of mediastinum. Importancc of lymphocytes as an indicator of specimen adcquacy. Acta Cytol 1990; 34: 517-23. 30. Anon. Clinical staging of primary Jung canccr. Am Rev Respir Dis 1983; 127: 659-64. 31. Harrow E, Halber M, Hardy S, Haltcman W. Bronchoscopic and rocntgenographic corrclatcs of a positivc transbronchial necdlc aspiration in the staging of Jung cancer. Chest 1991; 100: 1592-2. 32. Blainey AD, Curling M, Gree M. Transbronchial aspiration of subcarinal lymph nodes. Br J Dis Clzest 1988; 82: 149-54. 33. Salathe M, Soler M, Bollinger CT, Dalqucn P, Perruchoud AP. Transbronchial needle aspiration in routine fiberoptic bronchoscopy. Respiration 1992; 59: 5-8. 34. Bhat N, Bhagat P, Pearlman ES et al. Transbron­chial ncedle aspiration in the diagnosis of pulmo­nary ncoplasms. Diagn Cytopatho!l990; 6: 14-7. 35. Shure D. Transbronchial biopsy and necdlc aspi­ration. Chest 1989; 95: 1130-8. 36. Mehta AC, Kavuru MS, Meeker DP, Gephardt GN, Nunez C. Transbronchial ncedle aspiration for histology spccimens. Chest 1989; 96: 1228-32. 37. Wang KP. Flexible transbronchial needle aspira­tion biopsy for histologic specimcns. Chest 1985; 88: 860-3. 38. Schcnk DA, Strollo PJ, Pickard JS et al. Utility of the Wang 18-gauge transbronchial histology necdlc in the staging of bronchogcnic carcinoma. Chest 1989; 96: 272-4. 39. Schenk DA, Chambcrs SL, Dcrdak S ct al. Com­parison of the Wang 19-gauge and 22-gauge necd­lcs in the mcdiastinal staging of Jung canccr. Am Rev Respir Dis 1993; 147: 1251-8. 40. Ndukwu I, Wang KP, Davis D, Welk P, Sutula M. Direct smear for cytological examination of transbronchial nccdle aspiration spccimens. Chest 1991; 100: 88S. 41. Davenport RD. Rapid on site evaluation of trans­bronchial aspirates. Chest 1990; 98: 59-61. 42. Hammersley JR, Grecn RA. Transbronchial needlc aspiration -evaluation of penetration depth and areas of high risk usc. Am Rev Respir Dis 1984; 129: A43. Radio! Oncol 1994; 28: 316-9. Carcinoembryonic antigen (CEA) levels in pleural effusion and serum in differentiation of malignant and benign pleural effusion Peter Berzinec,1 Jan Plutinsky,2 Katarina Zrubcova2 , Vladimir Vondrak1 , Maria Letkovicova1 1 Institute of TB and Respiratory Diseases, Nitm, 2 Institute of TB and Respiratory Diseases, Kvetnica Poprad, Slovakia We determined CEA leve/s in pleural effusion (CEAP) and in serum (CEAS) o.f 67 patients. Thirty-eight patients had histologically and/or cytologically praven malignant effusion (Group I) ancl 29 patients had benign ejfusion (Group II). The obtained data were analysed using multifactorial Kruskal-Wallis test, Bayesian theorema and Spearman rank correlations. Results: 1) There was a signi.ficant difference between the leve/s of CEAP in Groups I and l1 (p < 0.001), 2) there was no signi.ficant clifference between the leve/s o. f CEAS in Groups l nad II, 3) there was satisfactory sensitivity, specificity and preclictive value of CEAP in the cliagnosis o.f malignant pleural effusion, 4) combined measurement o.f CEAS-CEAP in the diagnosis o.f malignant pleural effusion by cut-off leve/s JO ng/ml-7 ng/ml had a positive predictive value of 0.95. Key worcls: pleural effusion carcinoembryonic antigen, pleural effusion, malignant Introduction Carcinoembryonic antigen (CEA), first describ­ed by Gold and Fredman in 1965, has been an extensively used serologic marker. Increased levels are detected in carcinoma of the large intestine, anus, pancreas, !iver, stomach, eso­phagus, pharynx, ovaries, breast, Jung, thyroid gland, some hemoblastoses and sarcomas as well as in some benign diseases: bronchitis, pulmonary emphysema, tuberculosis, hepatic cirrhosis, hepatitis, ureamia, or in heavy smo­kers. 1 Correspondence to: Peter Berzinec, M.D., Puskinova 134, 94901 Nitra, Slovakia. UDC: 616.25-002.3-096 CEA determined in serum, may be used in monitoring the course of disease or as a prog­nostic factor in patients with some types of cancer, whereas CEA serum determination for 3 diagnostic purposes1 -is less effective. CEA levels in pleural effusion have been investigated by severa! authors to distinguish malignant effusion from benign.4-7 However, the diagnostic usefulness of CEA measurement in pleural effusion is stili unexplained. The aim of the present study was to deter­mine and compare CEA levels in pleural effu­sions (CEAP) and in sera (CEAS) of patients with malignant and benign pleural effusion and to determine the value of CEAP and CEAS in differentiation between malignant and benign pleural effusion. Carcinoembryonic antigen (CEA) Patients and methods Patients 67 patients with pleural effusion were enroled into this study. 38 patients (Group I) had histologically and/or cytologically proven mal­ignant effusion: in 23 patients by Jung cancer, in 10 patients by breast cancer, in 3 patients by other types of cancer, in 2 patients by dissemi­nated cancers of unknown origin. 29 patients (Group II) had benign effusion: in 18 patients by tuberculosis, in 8 patients by TB-unrelated pneumonia, in 3 patients by ischemic heart disease. In Group l were 20 men, l8 women, age-me­ dian: 62 years (range: 42-82), in Group II were 16 men, 13 women, age-median: 60 years (range [6-91). CEA cletermination The samples of pleural fluid and sera were obtained on the same day in each patient. CEA was measurecl with commercially available ra­dioimmunoassay kits. Statistical analysis For the statistical analysis the multifactorial Kruskal-Wallis test, Bayesian theorema, ancl Spearman rank corrclations were used. 8· 9 Results Table 1 shows the median CEA Ievels in pleural fluid ancl sera of patients with malignant (Group I) ancl bening (Group II) pleural effusion. In Group I CEAP levels exceeclecl the cut-off Ievels of 5 ng/ml in 28 patients, 7 ng/ml in 27 patients, ancl 10 ng/ml in 23 patients. In Group Table l. CEA levels in pleural effusions (CEAP) and serum (CEAS). Median CEAP CEAS (x l -xn) (ng/ml) (ng/ml) I Malignant effusion 20.0 15.0 (n 38) (0.1-360.0) (0.9>80.0) II Benign effusion 3.75 5.55 (n = 29) (0.1 > 7.0) (2.0 > [4.0) II the CEAP levels exceecling the above-men­tionecl cut-off Ievels were founcl in 6, 5, and 4 patients respectively. CEAS levels exceecling the cut-off leve\ of 5 ng/ml were found in 31 patients, 7 ng/ml in 26 patients, ancl 10 ng/ml in 23 patients of Group I. In Group II the cut-off levels were exceeclecl only in 18, 10, and 5 patients respectively. There was a correlation between the levels of CEAS and CEAP in Group I: r = 0.88, p 0.0000 (+ + + ). Significant correlation (but still less strong) bctween the levels of CEAS ancl CEAP was found also in Group II: r 0.41, p = 0.25 (+). The comparison of CEAP and CEAS levels in Group I ancl Group II is shown in Figure l. Figure l. Comparison of CEAP and CEAS in paticnts with malignant (Group 1) and bcnign (Group 11) plcural effusion. There was a significant difference between the levels of CEAP in Groups I ancl II (p < 0.001), but the difference between the levels of CEAS in Groups I and II was not significant. The sensitivity, specificity, and predictive va­lue (positive ancl negative) of different CEA cut-off levels in the diagnosis of malignant pleural effusion are shown in Table 2. Positive preclictive value of different com­bined CEAS -CEAP cut-off levels in the cliagnosis of malignant pleural effusion is shown in Table 3. Discussion Elevated levels of CEAP were found to corre­late with malignancy at sensitivities ranging Berzinec P et al. Table 2. Sensitivity, specificity, and predictive value (positive and negative) of different CEA cut-off levcls in the diagnostic of malignant pleural effusion. CEA Cut-off Sensi- Specifi- Predictive value tivity city Posi- Nega- (ng/ml) (%) (%) tive tive Pleural 5 72.05 78.57 0.83 0.65 effusion 7 71.05 82.71 0.86 0.67 10 60.53 85.71 0.87 0.59 Serum 5 76.05 40.11 0.65 0.52 7 63.00 66.00 0.73 0.54 10 56.01 83.00 0.83 0.65 Table 3. Positive predictive value of different combined CEAS -CEAP cut-off levels in the diagnostic of malignant pleural effusion. CEAS - CEAP Test results and (ng/ml) (ng/ml) predictive value Cut-off leve] 1-1 1-0 0-1 0-0 5 5 0.86 0.40 0.74 0.24 5 - 7 0.88 0.39 0.78 0.23 5 - 10 0.89 0.46 0.79 0.28 7 5 0.90 0.49 0.73 0.23 7 7 0.92 0.49 0.77 0.22 7 10 0.93 0.62 0.61 0.33 10 5 0.94 0.63 0.72 0.22 10 7 0.95 0.63 0.76 0.21 10 - 10 0.95 0.69 0.77 0.26 Tests results: l -positive, O negative from 25 % to approximately 70 % using diffe­ 10• 11 rent cut-off levels from 5ng/ml to 15ng/ml.4• By using lower cut-off levels, it has been impos­sible to detect a significant difference between benign and malignant effusion. 12 In our study the cut-off leve! of CEAP 10 ng/ ml has yielded a sensitivity of 60.53 % , specifi­city of 85.71 % , and positive predictive value of 0.87. The combined measurement of CEAP and CEAS has further enhanced the positive predictive value. In other words, the probability of malignant etiology by elevated levels of CEA both in serum and pleural effusion over 5 ng/ml is 86 % , over 7 ng/ml -92 % . In the presence of CEA levels in serum exceeding 10 ng/ml, and CEA levels in pleural effusion exceeding 7 ng/ml the probability of malignancy is 95 % . Nevertheless, these probabilities are influenc­ed by the studied group of patients referred to our institute. The prevalence of malignant effu­sion in this group of patients was 0.6. By changing prevalence, according to the situation in other institutions, different probabilities may be expected. In conclusion, the measurement of CEA le­vels in pleural effusion, together with the mea­surement of CEA levels in serum, in addition to cytology, pathology, and other studies, may contribute to the diagnostic evaluation of pa­tients with pleural effusion. For the institutions dealing with differential diagnosis of pleural effusion in larger groups of patients we would recommend the statistical analysis which would take into account the local situation -prevalence of the malignant pleural effusion -before choosing the optimal values of CEAP and CEAS for diagnostic pur­poses. References l. Kausitz J. Radioimmunoassay of tumour markers. Bratislava: Veda, 1991. 2. Stieber P, Fateh-Moghadam A. Tumormarker und ihr sinnvoller Einsatz. Klin Lab 1993; 39: 291-306. 3. Fletcher R. Carcinoembryonic antigen. Ann Intern Med 1986; 104: 66-73. 4. Thomas NC, Rana B, Ratcliffc JG. Carcinoem­bryonic antigen assay in pleural effusions. Ann Intern Med 1979; 90: 720-l. 5. Straehe RR, Briese V, Willroth PO, Voss P, Seyfarth M, Roepeke G, Broek J. Differential­diagnostisehe Wertigkeit von Tumormarkern in Pleuraergussen benigner und maligner Aetiologie unter besonderer Berueksiehtigung des eareinoem­bryonalen Antigens (CEA). Z Klin Med 1988; 43: 1963-8. 6. Chung-Hung Chen, Jen-Fon Lin, Jen-Ho Wen. Comparison of plcural effusion and serum ear­cioembryonie antigen (CEA) leve] in differentiat­ing benign and malignant pleural effusion. ARRD 1992; 145: A425. 7. Plutinsky J, Biza J. Berezova M, Vondrak, V. Letkovieova H. Examination of adenosine deami­nase and eareino-embryonal antigen in differential diagnosis of pleural effusion. Stud Pneumol Phty­seol 1993; 53: 18-22. 8. Keller H, Trendelenburg Ch. Data presentation/ Interpretation. Berlin, New York: Walter de Gruyter, 1989. Carcinoembryonic antigen (CEA) 9. Letkovicova M, Plutinsky J, Neuschl S. Diagnostic probability according to one or more tests. L&T 1992, 23: 126-30. 10. Ferroni P, Szpak C, Greiner JW, Simpson JF, Guadagni F, Johnston WW, Colcher C. CA 72-4 radioimmunoassay in the diagnosis of malignant effusions. Comparison of various tumour markers. Int J Cancer 1990; 46: 445-51. 11. Tamura S, Nishigaki T, Moriwaki Y, Fujioka H, Nakano T, Fuji J, Yamamoto T, Nabeshima K, Hada T, Higashino K. Tumor markers in pleural effusion diagnosis. Cancer 1988; 61: 298-302. 12. Whiteside TK, Dekker A. Diagnostic significance of carcinembryonic antigen levels in serous effu­sions. Correlation with cytology. Acta Cytol 1979; 23: 444--8. Radio! Onco/ 1994; 28: 320-2. Tumor markers and hormones as a diagnostic and prognostic test in lung cancer Gennady S. Moroz Ternopol Meclical Institute, Ternopol, Ukraine A study of the blood-levels of carcinoembrionic antigen (CEA), alfa-fetoprotein, ferritin, ACTH, cortisol, triiodothyronin (T3) and thyroxin (t4) carried out by means of radioimmunoassay in 61 patients with lung cancer, 57 patients with lung tuberculosis and in 12 healthy controls has shown that the leve/s of CEA, ferritin and cortisol in patients with lung cancer exceed the leve/s in patients with lung tuberculosis (p < 0.05) and the controls (p < 0.05). Low levels of T_, and T4 were found in more than 70 % of lung cancer patients (p > 0.05). The highest leve! of ACTH was nateci in patients with adenocarcinoma (20.68nmol/L) and distant metastases (21.12nmol/L). There was no correlation between the clinical stage of the disease, the histological pattern of tumor and the levels of tumor markers. These data can be used as a diagnostic and prognostic test. Key words: lung neoplasms-diagnosis; tumor markers, biological; hormones; prognosis Introduction A study of the blood-Jevels of tumor markers and hormones has been recently wideJy used as a diagnostic and prognostic test in Jung cancer simultaneousJy with other routine methods. l-4 The aim of our work was to determine the blood-levels of some tumor markers and some hormones for initial diagnosis and differential diagnosis of cancer and tuberculosis of the lung. Material and methods The bJood-JeveJs of carcinoembryonic antigen (CEA), cortisol, triiodothyronim (T3), thyroxin Correspondence to: Prof. Gennady S._Moroz, M.D., l Maidan Voli, Tcrnopol, Ukrainc. UDC: 616.24-006.6-096 (T4) (made by means of a standard kit produced in the former Soviet Union) and alfa-fetopro­tein, ferritin, ACTH (of the French firm CJS) were measured by means of radioimmunoassay in 61 patients with lung cancer (group 1), 57 . patients with tuberculosis of the Jung (group 2) · and in 12 healthy coontrols (the control group 3). The bJood was taken on an empty stomach in the morning. The serum was kept in a refrigerator at -20° C not more than 1-2 months. The obtained data underwent statistical analysis. Forty-seven man and 14 women for­med the first group of patients, aged 46-72. 22 patients had the second stage of cancer. The third stage was diagnosed in 23 cases while 16 patients had the fourth stage of the disease. Ali the cases were histologically or cytologically verified. Fourteen had adenocarcinoma and 27 squamous cell carcinoma. Small-cell lung cancer was diagnosed in 20 patients. In 38 cases it was Tumor mtirkers and lwrmones as a diagnostic and prognostic test in lung cancer 321 necessary to conduct differential diagnosis bet­tive form of tuberculosis was present in 28 ween the infiltrative form of tuberculosis and patients and the cavity phase of tuberculosis in the central cancer of the upper !obar bronchus 29 patients. Since in ali the examinecl patients with atelectasis or hypoventilation. Twenty­there has not been established a distinct depen­three patients hacl a peripheral tumor of the clence of tumor markers and hormones upon upper lobe which was difficult to differ from age, sex, stages of cancer, clifferent fonns of tuberculoma or the rounclecl shape of infiltrate. tuberculosis, etc., the clata in Table 1 are given Group 2 comprisecl 46 men ancl 11 women, for ali groups together. ranging in age form 25 to 67 year. The infiltra- Table 1. The blood levels of tumor markers and some hormones in the group observed. Tumor markers, hormones (nmol/) Control group (n = l2),M±m Group 1 (n = 61),M±m Group 2 ll = 57), M± 111 CEA 0.04± 0.01 (0.01-0.06) 0.17 ± 0.015*,** (0.I0-0.19) 0.081 ± 0.015* (0.04-0.10) Ferritin 0.10±0.026 0.668 ± 0.52* 0.486 ± 0.44** (0.08-0.13) (0.52-0.86) (0.10-0.61) ACTH 11.52 ± 0.73 16.97 ± 1.6* 11. 96 ± 1.32 (9.22-13.41) ( 11.0-18.33) (10.28-13.0) Cortisol 182.87 ± 33.68 652.10 ± 60. 3*.** 330.15 ± 46.9* (146.4-20.27) 594.0-730.0) (290.0-400.0) T, 2.26 ± 0.12 (2.13-3.0) 1.33±0.03 (0.96-2.30) l. 23 ± 0.35 (0.88-2.20) T4 67.9±9.9 59.15 ±3.5 83. 5 ± 7.5** (57.9-78.0) (50.1-68.0) (72.4-93.0) * Significantly highcr (p < (l.05) than thc control. ** Statistically significant diference (p < 0.05) between the first ancl the seconcl groups. The limits of the levels variation of the substances stucliecl, are given in parentheses. Results and discussion As evident from Table 1 in patients with Jung cancer much higher blood levels of CEA, ferri­tin ancl cortisol were founcl than in the control group (the lowest bloocl-levels of these substan­ces in group l were higher than the highest ones in group 3.) The inter group comparison revealecl significan clifferences (p < 0.05). The excess of ACTH was found in 70 % of Jung cancer patients (more than 13.5 nmol/ L). The excess was significant (p < 0.05). Lower Ievels of T3 and T4 were founcl in more than 70 % of lung cancer patients. The clifference was not statistically significant. The blood Ievels of CEA ancl cortisol in patients with Jung cancer were significantly higher (p < 0.05) than those in patients with tuberculosis. The same substances and ferritin in patients with tuberculosis were significantly higher (p < 0.05) than those of the controls. This phenomenon is difficult to explain. It is stili harder to explain the significantly higher (p < 0.05) blood levels of thyroxin in patients with tuberculosis than in the by can group. The correlational analysis demonstrated a high posi­tive Pearson r (r 1 in r2 more than 0.8) between the levels of ACTH, cortisol ancl the levels of ACEA. There was no correlation between the clinical stage of the disease, the hystological pattern of tumour and the leve! of CEA found, although there was a two-threefolcl excess in individual cases. The bloocl-levels of CEA, for 322 Moroz G S example, in the individual patients with a cen­tral cancer of the lung stages III-IV reached 1.075-1.525 nmol/L. The average levels of CEA in patients with small-cell lung cancer was 0.335 ± 0.02 nmol/L and two-threefold higher in patients with squamous cancer and adenocar­cinoma (0.007 ± 0.01 and 0.126 ± 0.01 nmol/L) respectively. The highest leve! of ACTH is found in patients with adenocarcionoma (20.68nmol/L) and with distant metastases (21.12 nmol/L). As a rule, an increase in the leve! of ACTH associated with an increase in the level of cortisol. These 50 % bron­chial luminal obstruction (range 50-95 % , mean 75%). Table 2 presents the post-treatment data. Table 2. Endoscopic Y AG laser treatment for exten­sive inoperable broncho-pulmonary cancer (n 350). Results No procedure related mortality Improvement in bronchial opening (43 % mean)* Improvement in FVC: 0.1 litre* Improvement in FEVl: 0.05 litre* Symptom relief: Immediate -lasting 4-12 weeks Survival 2-25 months * 1 month post trcatmcnt Group lb There were 99 patients in this group (62 male and 37 female) aged 33-80 years (mean 64.8). Eleven patients (11 % ) in this group had lymph node involvement on exploration. One patient died in hospital due to cardio-respiratory fail­ure. Three patients had prolonged air leak necessitating prolonged tube drainage. Hospital stay was 5-22 days (mean 9 days). Pulmonary function carried out 4-6 months post-operatively showed a mean FVC and FEVl deterioration compared with pre-opera­tive figure of 11 % and 4 % respectively. Unex­pectedly there was improvement post-operati­vely of FVC and FEVl in 15 % and 17 % of patients respectively. Long-term results have been related to the histology and staging. Group le In this group there were 179 patients (120 male and 59 female) age 45-71 (mean 64.5). Ali had conservative lung surgery of lobectomy in which laser was used as a complementary measure to excise small subsidiary lesions separately or as a continuation of the main tumour. Twelve patients (7 % ) in this group had bilateral thora­cotomy for a second tumour or metastases. The total mortality was 2 (1.1 % ). Long-term survi­val depended on the stage of the disease. Group 2 There were 59 patients ( 47 male and 12 female) in this group, age 52-80 years (mean 75.5). Ali patients had inoperable Stage III tumour. Table 3 shows the post-treatment data of these patients. Table 3. Endoscopic PTD for broncho-pulmonary cancer (n = 59). Results No procedure related mortality 1 patient with mild skin photosensitivity reaction lmprovement in bronchial % opening 71 % (mean)* Improvement in FVC: 0.47 litres (mean)* Improvement in FEVl: 0.35 litres (mean)* Symptom relief: Lasting for 8-18 weeks Survival: 2-36 months * 1 month aftcr trcatmcnt Discussion It is generally accepted that surgical resection of the tumour and its satellite lymph nodes, is the treatment of choice for ali resectable non­small cell lung cancer. It is also an undisputable fact that only the minority of lung cancer suffe­rers qualify for receive such surgical treatment. Illustration of this fact may be furnished by referring to statistics provided by the Yorkshire Regional Cancer Organisation, an organisation which registers ali cancer throughout the York­shire Region of the UK, with a population of 3.6 million. Some 2760 cases of Jung cancer are diagnosed yearly, of which 9. 6-11. 8 % receive surgical treatment. 5 When one considers that Lasers in broncho-pulmonary cancer many of the operated patients will have recur­rence of their turno ur, it becomes clear that at presentation a great many patients with Jung cancer are receiving treatment which can be classified as palliative. It is against this back­steet was perceived. In fact, from the beginning of its introduction to the medica! field, much of laser research has centred around evaluation of its benefit in cancer therapy generally and is directed towards palliation of patients with ex­tensive turno ur more specifically. In broncho-pulmonary cancer lasers have been used predominantly for endoscopic treat­ment. Since the early 1980's thousands of pa­tients have received bronchoscopic Y AG laser treatment for palliation of their broncho-pulmo­nary cancer. This is well reflected in the number 8 of surveys and review articles.6· Our series of 350 patients (Group la) confirms the previous findings. In addition our experience suggests that in a selected group of patients laser treat­ment coupled with external beam radiotherapy, thus combining intra-Iuminal with external ra­diotherapy achieves palliation over a Ionger period. Endoscopic PDT in our series of 59 patients had given stili better symptom relief results and Jonger survival confirming our previous fin­dings9 that not only is PDT beneficial in exten­sive Jung cancer with bronchial obstruction, but its benefits are more substantial and last for a longer duration thas those of YAG laser. Some authors9 · 10 believe that endoscopic PDT is indicated only in early endobronchial tumours. Our results and those of some other authors11 -15 demonstrate conclusively the value of PDT in early as well as extensive Jesions when there is substantial obstruction of the bronchial lumen. Intra-operative use of YAG laser in pulmo­nary surgery generally and Jung cancer more specifically has received Jittle attention. Never­theless our previous studies2• 3• 16 and the results of the present series clearly point out the advantages of the technique in which a nodular tumour is locally excised through a Iimited thoracotomy. In effect the use of laser allows patients with borderline pulmonary function to undergo Jung resection for cancer because the technique is attended by preservation of pulmo­nary parenchyma which is essential to such patients. The Jow operative mortality16 and insignificant post-operative complications are a reflection of this and alsoof the capability of YAG laser to seal off blood vessels and minute airways. 17 It is obvious that pre-operative asses­sment of pulmonary function and nodal status are particularly important in patients submitted to this type of surgery. The association of laser with conventional surgery in Group le is another way of achieving conservative parenchyma saving surgery. It is to be emphasised that the use of laser either for local excision of "coin" lesions or in associ­ation with conventional surgery will not and should not compromise the oncological integrity of the Jung cancer operation. To the contrary, it will add to the safety of the resectional surgery because of the destructive action of Jasers on cancer tissue. Conclusions The use of Iasers makes an important contribu­ tion to the treatment of Jung cancer. Employed endoscopically they can disperse obstructive intraluminal Jesions, thus improving ventilation. They can also reduce cough, frequency of hae­ moptysis and chest infection. YAG laser and/or PDT used individually (and in some cases alternately) can achieve desobliteration of the airway. In early tumours confined to the bronchial wall alone cure may be attained using endosco­ pic PDT. At the present tirne intra-operative indica­ tions of laser in Jung cancer is confined to YAG which is particularly useful in local excision of nodular tumours. The potential intra-operative usage of PDT has yet to be defined. References 1. Moghissi K, Jessop T, Deneh M. A new broneho­seopy set for laser therapy. Thorax 1986; 41: 485-6. 364 Moghissi K 2. Moghissi K, Dench M, Goebells P. Experience in non-contact Nd Y AG laser in pulmonary surgery: A pilot study. Eur J Cardio-thorac Surg 1988; 2: 87-94. 3. Moghissi K. Local excision of pulmonary nodular (coin) lcsion with non-contact yttrium aluminium garnet laser. J Thorac Cardiovasc Surg 1989; 97: (1): 147-51. 4. Moghissi K, Parsons RJ, Dixon Kate. Photodyna­mic therapy (PDT) for bronchial carcinoma with the use of rigid bronchoscope. Lasers in Medica! Science 1992; 7: 381-5. 5. Lung Cancer. In: Joslin C, Rider L eds. Cancer in Yorkshire Cancer Registry -Special Report Series l. 1992. 6. Toty L, Personne C, Golchen A, Vourc H. Bron­choscopic management of tracheal lcsions using the neodymium yttrium aluminium garnet laser. Thorax 1981; 36: 175-8. 7. Dumon JF, Reboud E, Garbe L, Aucomte F, Merk B. Treatment of tracheo-bronchial lesions by laser photoresection. Chest 1982; 81: 278-84. 8. Hetzel MR, Smith SGT. Endoscopic palliation of tracheo-bronchial malignancies. Thorax 1991; 46: 325-33. 9. Moghissi K, Dixon K, Parsons RT. A controlled tria! of Nd Y AG laser versus Photodynamic the­rapy for advanced malignant bronchial obstruct­ion. Lasers in Medica/ Science 1993; 8: 269-73. 10. Hayata Y, Kato H, Konaka C, et al. Haematopor­ phyrin derivative and photoradiation therapy in early stage Jung cancer. In: Berns MW ed. Haema­toporphyrin derivative photoradiation therapy of cancer. New York: Alan R Lisse Inc, 1984: 39-47. 11. Cortese DA, Bronchoscopic photodynamic the­rapy of early Jung cancer. Chest 1986; 90: 629-31. 12. Vincent R, Dougherty T. Photoradiation therapy in the treatment of advanced carcinona of the trachea and bronchus. Pro Ciin Biol Res 1984; 170: 759-66. 13. Doiron DR, Balchum OJ. Haematoporphyrin de­rivative photoradiation of endobronchial Jung can­cer. In: Andreoni A, Cubeddu R eds. Po171hyrin in tumour plzototherapy. New York and London: Plenum Press 1984; 355-403. 14. McCaughan JS Jr, Williams TE Jr, Bethel BH. Photodynamic therapy of endobronchial tumours. Lasers Surg Med 1986; 6: 336-45. 15. Locicero J III, Metzdorff M, Almgren C. Photo­dynamic therapy in palliation of late stage obstruc­tion non-small celi Jung cancer. Chest 1990; 98: 97-100. 16. Moghissi K. Experience in limited Jung resection with the use of laser. Lung 1990; Suppl: 1103­1109. 17. Moghissi K, Dench M, Neville E. Effect of the non-contact mode of YAG laser on pulmonary tissues and its comparison with electrodiathermy; An anatomo-pathological study. Lasers in Meclical Science 1989; 5: 17-23. Radio/ Oncol 1994; 28: 365-8. Long-term results of primary surgery for stage I small celi lung cancer Gaetano Rocco, Maurizio Tondini, Fabio Massera, Gerolamo Rossi, Claudio Della Pona, Mario Robustellini, Andriano Rizzi Division of Thoracic Surgery, E. Morelli Regional Hospital, Sandalo, Italy Between January 1979, and December 1993, 17 patients (15 females and 2 males; mean age 57.9 ± 2.8 (SEM) yrs., median 55 yrs, (range 43-78) with stage 1 small-cell lung cancer underwent primary surgery followed by adjuvant polychemotherapy and prophylactic cranial irradiation. In no instance was preoperative histological diagnosis available. The extent of resection included a lobectomy in 9 patients, a pneumonectomy in 1, and a segmenta! resection in 7. Postoperative chemotherapy included three cycles of cyclophosphamide -Z 000 mg!m2, vincristine -1.3 mg!ni2 , doxorubicin or epidoxoru­bicin 60 mg!ni2 , administered every 21 days, followed by three additional cycles of cisplatin ­25 mg!m2 , and VP16-120 mg!m2 , administered in 3 consecutive days every 21-28 days. Prophylactic cranial irradiaton was given concurrently with the first cycle of chemotherapy and immediately after surgery. Actuarial 5-year survival was 47% (Kaplan-Meier method). Median survival was 61 months. No statistically significant difference in terms of survival was detected between lobectomy and other procedures (p 0.2, log-rank test). Key words: Jung neoplasms-surgery; carcinoma, oat celi, survivaJ analysis Introduction Combined treatment modalities for stage I small-cell Jung cancer are advocated in order to obtain Jong-term survival. 1 Surgery of the peripheraJ stage I small-cell carcinoma is war­ranted both for diagnostic and therapeutic pur­poses. Recently, the thoracoscopic approach has gained the favor of surgeons deaJing with suspected lesions in the peripheraJ lung,2 thus reducing postoperative morbidity. Correspondence to; Gaetano Rocco, M.O., Division of Thoracic Surgery, E. Morelli Regional Hospital, via Zubiani 33, 23039 Sondalo (Sodrio), ltaly. We report our experience concerning the management of stage I (T1NO/T2NO subsets) small-cell carcinoma over a 15-year period. Material med methods Between January 1979, and December 1993, 17 patients (15 males and 2 femaJes; mean age 57.9 ± 2.8 (SEM) yrs., median 55 yrs., (range 43 -78) with stage I small-cell Jung cancer (SCLC) underwent primary surgery followed by adjuvant poJychemotherapy and prophyJactic craniaJ irradiation. In no instance was preopera­tive histoJogicaJ diagnosis avaiJabJe. In ali pa­ UDC: 616.24-006.6-089 tients, a peripheral noduJe was visible on stan­ Rocco G et al. dard chest x-ray. The staging procedure for Actuarial 5-year survival was 47 % (Kaplan­clinical stage I disease did not include a preope­Meier method) ( Figure 1). Mean and median rative mediastinal exploration because of the lack of mediastinal node enlargement at CT . evaluation. A muscle-sparing limited thoracotomy was used in our early series. Since 1992, video-assisted thoracoscopy was routinely used for resection of the nodule which was removed from the pleural cavity via an accessory thoracotomy ( comparable in lenght to the previously used muscle-sparing limited thoracotomy). An effort to sample accessible nodal stations was routinely made. The extent of resection included a lobectomy in 9 patients, a pneumonectomy in 1, and a segmenta! resec­tion in 7. Intraoperative frozen section of the surgical specimen was obtained in ali cases. Nine pa­tients with stage I disease had a Tl and 8 had T2 subset. Postoperative chemotherapy included three cycles of cyclophosphamide -1000 mg/m2 , vin­cristine -1.3 mg/m2, doxorubicin or epidoxoru­bicin 60 mg/m2 , administered every 21 days, followed, more recently, by three additional cycles of cisplatin -25 mg/m2 , and VP16 ­120 mg/m2 , administered in 3 consecutive days every 21-28 days. Prophylactic cranial irradia­tion (PCI) (30 Gy) was given concurrently with the first cycle of chemotherapy and immediately after surgery. The survival rates were calculated according to the Kaplan-Meier method.3 Results Histologic examination showed a pure small celi carcinoma in ali cases. There were no surgery-related deaths nor postoperative complications. As of December 1993, 11 patients are alive and without evidence of disease. The longest survival has been 150 months. Six patients died from disseminated disease with a longest survival of 48 months. In no instances was relapse of the disease in the primary site detected. ·;;100 ... 80 ;:l VJ 4-< 60 o ----'.,j Q 40 .s 20 o o.. o ... o... Years Figure 1. Overall survival of patients with stage SCLC (Kaplan-Meier method). survivals were 58 ± 10.4 (SEM) and 61 months, respectively. No statistically significant diffe­rence in terms of survival was detected between lobectomy and other procedures (p = 0.2, log­rank test) (Figure 2). Similarly, no significant o... _.._ TI _,__ T2 Figure 3. Survival of patients with stage I SCLC by stage (Tl vs TI). Lang-term results of primary surgery for stage I small celi lung cancer No signs of PCI-related toxicity to central nervous system (CNS) were noted. Discussion Only 10 % of patients with SCLC present with a peripheral nodule. 1 The 5-year survival of these patients is comparable to that of patients with other histotypes.1 In our limited series, the 5-year survival of patients with stage 1 SCLC reached an actuarial survival of 47 % which compares favorably with the 59.5 % for TINO and the 27.9 % for TINO reported in one study.1 Similar results have been obtained in a multi-institutional study focused on initial sur­gery followed by chemotherapy (55 % 4-year survival for stage I SCLC).4 The addition of chemotherapy, chest radio­therapy and prophylactic cranial irradiation in patients with Iimited disease has been advocated in order to obtain better Iocal control and to avoid systemic recurrences. 1 Based on the ob­servation of a very low incidence of brain metastases from stage I SCLC and the reported CNS toxicity, the routine resort to PCI is under debate.4 Furthermore, TXNO subsets are reported to have a 50 % chance of systemic reccurences after surgery, and an identical 50 % chance of local recurrence after chemotherapy. 1 In an effort to improve overall survival rates, we used this aggressive approach in our patients with stage I disease. Local control at the primary site was obtained and adjuvant treatments were well tolerated as no significant morbidity was reported. Preoperative diagnosis of a peripheral nodule can be obtained through either needle biopsy or surgical resection. The latter is achieved via a thoracotomy or, more recently, via a thora­coscopic approach.2 In our case, the significant false-negative rate of transthoracic needle biopsy has impaired the overall sensitivity of this procedure.5 Of greater importance is the possibility to draw distinction between SCLC and well-diffe­rentiated neuroendocrine tumors which are amenable to surgery alone.1 To do so, an adequate specimen is needed in order to obtain a fina! pathologic analysis.1 Accordingly, we believe that surgical resection only can provide an adequate tissue sample for this precise discri­mination. Before the thoracoscopic era, we favored a muscle-sparing limited thoracotomy for the diagnosis and resection of undefined peripheral nodules. Recently, the role of surgical staging procedures in the management of SCLC has been emphasized.1 According to an observation made by the Toronto Lung Cancer Study Group,6 nearly 48 % of the patients with clinical stage I disease will be upstaged after histological examination of the surgical specimen. As a result, prethoracotomy mediastinal exploration is indicated,4 for the evidence of N2 disease which is a contraindication to resection.7 In this setting, video-assisted thoracoscopic techniques may be of value,8 although, in our series, we have not observed a remarkable advantage of thoracoscopy over minithoracotomy as far as postoperative morbidity and hospital stay are concerned. In conclusion, we emphasize the role of sur­gery as a part of a multimodality approach to stage 1 SCLC, yielding better local control of the primary and attendant satisfactory Iong­term survival rates. 6· 9 References 1. Mcntzcr SJ, Rcilly JJ, Sugarbakcr DJ Surgical rcscction in thc managcmcnt of small-ccll carci­noma of thc lung. Chest 1993; 103: 349S-51S. 2. Landrcncau RJ, Hazclrigg SR, Fcrson FP, ct al. Thoracoscopic rcscction of 85 pulmonary lcsions. Ann Thorac Surg 1992; 54: 415-20. 3. Glantz SA. Statistica per discipline biomediche. 2nd Ed. McGraw-Hill Italia srl, 1988. 4. Shiclds TW. Surgical thcrapy for carcinoma of thc lung. In: Matthay RA cd. Lung Cancer -Clinics in Chest Medicine. Philadelphia: WB Saudcrs Phi­ladelphia, 1993: 139-40. 5. Rizzi A, Radaelli F, Robustcllini M, Rossi G, Rocco G, Dclla Pona C. The coin Icsions: our Rocco G et al. experience. In: Monduzzi ed. Proceedings of the Second World Week of Professional Updating in Surgery and Oneology Disciplines of thc University of Milan. Milan, 1990: II327-II329. 6. Shepherd FA, Ginsberg RJ, Feld R, Evans WK, Johansen E. Surgical treatment for limited small­cell Jung cancer. J Thorac Cardiovasc Surg 1991; 101: 385-93. 7. Karrer K, Shields TW, Denck H, Hrabar B, Vogt- Moykopf I, Salzer GM. The importance of surgical and multimodality treatment for small celi bronchial carcinoma . .l Thorac Cardiovasc Surg 1989; 97: 168-76. 8. Naruke T, Asamura H, Kondo, H, Tsuchiya R, Suemasu K. Thoraeoscopy for staging of Jung can­cer. Ann Thorac Surg 1993; 56: 661-3. 9. Ginsberg RJ. Operation for small cell lung cancer. Where are we? Ann Thorac Surg 1990; 39: 692-3. Radio/ Oncol 1994; 28: 369-72. Surgery for Jung cancer in the elderly Gaetano Rocco, 1 Fabio Massera, 1 Claudio Della Pona, 1 Gerolamo Rossi, 1 Mario Robustellini, 1 Dario Ballabio, 2 Adriano Rizzi2 1 Division of Thoracic Surgery, E. Morelli Regional Hospital Sonda/o and, 2 Division of Thoracic Surgery, S. Gerardo University Hospital, Monza, Italy Between January 1979 and December 1993, 162 patients aged more than 70 years underwent surgery far lung cancer. Thirty-eight patients (23 %) were last to follow-up. Overall median survival was 32 months. The overal five-year survival was 24 % (Kaplan-Meier method). Age, histological type, stage and adjuvant treatments far stage 11/A were not significantly realted to survival. Female sex (p = 0.03), and extent of resection (p 0.02) significantly correlated to survival. Lobectomies were associated with a better survival than pneumonectomies (p = 0.03), the latter showing a better outcome than lesser resections (p = O. 03). The extent of resection was the only variable indipendently influencing the overall survival (p = 0.007). Median disease-free survival was 24 months. Female sex (p = 0.03) was the only indipendent predictor of disease-free survival by multivariate analysis (p = 0.02). Surgery for lung cancer in the elderly can be performed with acceptable long-term results in terms of overall and disease-fi·ee survival. The indications for pneumonectomy should be carefully weighed against potential postoperative risks. Even in the elderly patients, resections lesser than lobectomy, although functionally more acceptable, entail a greater risk far local recurrences. Key words: lung neoplasms-surgery; aged; survival analysis Introdnction In the early 70's, it was perceived that Jung cancer patients older than 70 years could not be suitable surgical candidates. 1 The advances in anesthesia and the improve­ment in surgical technique enabled this popula­tion to benefit from the only effective treatment Correspondence to: Gaetano Rocco, M.D., Division of Thoracic Surgcry, E. Morelli Regional Hospital Sondalo, via Zubiani 33, 23039 Sondalo (Sondrio), Italy. for lung cancer.1 Conversely, surgical considera­tion for octogenarians is now under debate.2 We report our 15-year experience on the surgical treatment of Jung cancer in patients older than 70 years. Material and methods Between January 1979 and December 1993, 162 patients (146 males and 16 females) aged more than 70 years underwent surgery for Jung cancer. UDC: 616.2.4-006.6-05-053.9-089 A clinically significant chronic obstructive Rocco G et al. pulmonary disease was found in association with lung cancer in over 90 % of our patients. Preoperative workup included a careful eva­luation of the performance status, an accurate assessment of pulmonary reserve through spiro­metric tests and the determination of maximum oxygen consumption. Cardiologic profile was documented by routine examination, ECG, and ultrasound, as indicated. Borderline function required ventilation/perfusion seans or 24-hour ECG monitoring. In no instance was preopera­tive coronary arteriography indicated. Special attention was paid to preoperative physical therapy focused on postural drainage and diaphragmatic training. Until recently mediastinoscopy was not inclu­ded, among staging procedures for lung cancer. When an N2 disease was suspected, a muscle­sparing limited thoracotomy was used with me­diastinal node sampling and frozen section. CT evaluation of the brain was obtained routinely, whereas bone seans were done in the presence of symptoms suspicious for skleletal involvement. For the sale of simplicity, the patients were grouped according to their age (group 1, includ­ing 110 pts. aged 70 to 73 years; group 2 including 36 pts. aged 74 to 76 years; group 3 including 16 pts. aged 77 yrs and older), gender (males = 146; females = 16), stage (I = 101 pts.; II = 11 pts.; IIIA = 50 pts.), histologic type (epidermoid = 97 pts.; adenocarcinoma = 47; other = 18 pts.), the extent of resection (lobec­tomy = 110 pts; pneumonectomy = 27 pts.; oth­ers = 24 pts.; unknown = 1), and the admini­stration of adjuvant treatment for stage IIIA (yes = 20 pts.; no = 14 pts.). The influence on survival and disease-free interval of these variables was evaluated by univariate (log-rank test or Mann-Whitney U test for two groups, and Kruskal-Wallis test for multiple groups) and multivariate analysis (Cox proportional hazard regression model), as ap­propriate. 3· 4 Twenty-eight patients (17 % ) were lost to follow-up. These patients were evenly distribut­ed among the three groups (stage I = 10; stage II = 1; stage IIIA = 17). Accordingly, we be­lieve that this relatively high percentage of lost-to-follow-up patients has not affected the statistical analysis. Results Overall mean and median survivals were 22. 7 ± 2.1 (SEM) and 32 months (range, 1-132), respectively. The overall five-year survival was 24 % (Kaplan-Meier method) (Figure 1). The r :--1 p 100..,=------------:----­ - o p 80 o r t 60 1 o n 40 s u 20 v 1 0L------'-----'------'----. -'---­o 1 2 a 4 s 1 n Years g -tE-Onrall _.,... 7D :o 73 71. -+.-74 to 76 rs -a-77 and older Figure l. Survival by agc groups. 5-year survival rates for patients aged 70-73, 74-76, and 77 years and older, were 30 % , 19 % and 13 % , respectively. Age (p = 0.16), histological type (p = 0.5) (Figure 2), and stage (p = 0.09) were not signi- r o i ! '.:=:::t-,, o n. "' '">'.--­ s u 20 r -. v 1 1 v o 1 2 3 o 1 n Years g -OTerall -+-Sqaamou1 -+-Adenocarctnoma -s-Other Figure 2. Survival by histology. ficantly related to survival, although there is a trend towards better survival in patients with earlier stages of the disease. Surgery for lung Although no statistically significant difference was detected between stages I and II of the disease, stage I patients fared somewhat better than those with stage IIIA (p 0.03). The estimated 3-year survival for stage I and IIIA patients was 48 % and 36 % , respectively. For T3 and N2 subsets, 3-year survival was 39 % and 33 % , respectively (NS). While T3 patients showed a 19 % 5-year survival, no 5-year survi­val was established for N2 patients (Figure 3). p . I00 .. .=-----------------7 p o 80 60 o n 40 s u 20 v .---.--­ o.---.----.---. 1 g o v 1 Years n Overall . Slago 1 . Stage II . Stage IIIA Figure 3. Ovcrall survival and survival by stagc in clderly patients undergoing surgery for lung cancer. Adjuvant treatments for stage IIIA disease did not improve the survival (p 0.4, log-rank test). No patients with stage II Jung cancer reached a 3-year survival. Female sex (p = 0.03 log-rank test; p = 0.009 Mann-Whitney U test) (Figure 4), and the cancer in the elderly extent of resection (p 0.02, Kruskal-Wallis) significantly influenced the survival. Elderly patients subjected to lobectomies had a better survival than those with pneumonecto­ mies (p = 0.03), the latter presenting with a better outcome than those with Iesser resections (p = 0.03). Opera ti ve mortality, defined as death occurr­ing within 30 days from surgery, was 2.4 % ( 4 patients). Two patients in the younger age group with stage IIIA epidermoid carcinoma died after lobectomy (1) and pneumonectomy (1). In addition, two patients in the middle age group with epidermoid (1) and adenocarcinoma . (1) died in the postoperative period. They had undergone a lobectomy and a pneumonectomy for stage I and IIIA disease, respectively. Adult respiratory distress syndrome was the cause of death in ali but one patient who died from post-pneumonectomy pulmonary embolism. By multivariate analysis, the extent of resec­ tion was the only variable indipendently influencing survival (p = 0.007). Overall mean and median disease-free survi­ vals were 21.6 ± 2.1 (SEM) and 24 months, respcctively. Female sex (p = 0.03, Iog-rank test; p 0.08, Mann Whitney U test), and lobectomy (p = 0.02), were significantly related to disease­frec interval. The former was the only indipen­dent predictor of discase-free survival by multi­variate analysis (p = 0.02). Discussion The indications for resection of lung cancer in Japanese octogenarians have been expanded, based on the remarkable life expectancy in these patients (6.9 and 8.7 years for men and women, rcspcctively).1 Osaki and coworkers have also observed a 5-year survival rate of 25.4 % for resected pa­ 1 tients aged 70-79 ycars. Although thc agc groups in our study are different, we found a comparable overall 24 % 5-year survival with a 26 % 5-year survival if ages from 70 to 77 years wcre considered together. Our 13 % 5-year Rocco G et al. survival for patients older than 77 years is worse than the reported 32.2 % 5-year survival 1 for octogenarians.Considering the high num­ber of patients !ost to follow-up that were included among the censored cases, we believe that our figure could be underestimated. Our overall 2.4 % operative mortality com­plies with the one reported by Yellin. 2 Signifi­cantly, no deaths occurred in the group of patients older than 77 years. Rigid functional selection criteria are crucial for obtaining low mortality rates since cardiorespiratory compli­cations are reported to be a major prognostic factor of long-term survival. 1 1 As noted by others, elderly patients with IIIA disease have acceptable 3-year survivals. In this setting, no statistically significant diffe­rence was noted in our series ad to T3 vs. N2 subsets. However, after the 3-year limit, there is a worse prognostic trend for patients with N2 disease. Surprisingly, there is no 3-year survival for stage II disease in our series. Although the number of stage II patients is too small to draw definitive conclusions, the lack of an adequate mediastinal staging might have played a role in underestimating N2 disease. In our series, those patients undergoing lo­bectomy had a better survival than patients subjected to other surgical procedures. This finding is most probably attributable to the greater functional impairment caused by pneu­monectomy, and to the overall likelihood of recurrence with lesser resections than with lo­bectomies. 5 Surgery for lung cancer in the elderly can be performed with acceptable long-term results in terms of overall and disease-free survival.1 The resort to pneumonectomy should be carefully weighed against potential postoperative risks.2 Even in the elderly patient, resections lesser than lobectomy, although functionally more ac­ceptable, are associated with greater risk of local recurrences. References l. Osaki T, Shirakusa T, Kodate M, Nakanishi R, Mitsudomi T, Ueda H. Surgical treatment of Jung cancer in the octogenarian. Ann Thorac Surg 1994; 57: 188-93. 2. Yellin A. Invited commentary on Osaki.1 Idem. 3. Glantz SA. Statistica per discipline biomediche. 2'"1 Ed. McGraw-Hill Italia srl, 1988. 4. Cox DR. Regression models and lifc-tables. IR Stat Soc B 1972; 34: 187-220. 5. Ginsberg RJ, Rubenstein L, for the Lung Canccr Study Group. Paticnts with Tl NO 11011-SCLC Jung canccr. Lung Cancer 1991; 7S: 83. Radio/ Onco/ 1994; 28: 373-5. Carcinosarcoma of the lung Claudio Della Pona, Gaetano Rocco, Fabio Massera, Mario Robustellini, Gerolamo Rossi, Adriano Rizzi Division of Thoracic Surgery, E. Morelli Regional Hospital, Sandalo, Italy Between .Tanuary 1979, and December 1990, 1242 patients with primary lung cancers were observed in our Division. Among these, 3 (0.2 % -1 male and 2 females; mean age 53 yrs.; range 40-75 yrs.) had a postoperative diagnosis of carcinosarcoma of the lung. Stage I was found in 2 patients, and stage II disease (NJ) in l. The resection included lobectomy in 2 patients and wedge resection in 1. Postoperatively, the patients did not receive any adjuvant treatment. As of December 1993, only one patient (pT2NO) is alive and without evidence of disease at 36 months from surgery, with the other two patients showing short postoperative survivales (3 and 7 months). Key words: lung neoplasms-surgery; carcinosarcoma lntroduction Among rare primary tumors of the lung, carci­nosarcoma occurs in about 1 % . 1 Three patients recently observed and treated in our Division are the subject of the following report and the reason for our review of the literature. Patients and methods Between January 1979, and December 1990, 1242 patients with primary lung cancers were observed in our Division. Among these, 3 (0.2 % . 1 male and 2 females; mean age 53 Correspondcncc to: Claudio Della Pona, M.O., Divi­sion of Thoracic Surgery, E. Morclli Rcgional Hospi­tal, Sondalo, Italy. UDC: 616.24-006.68-089 yrs.; range 40-75 yrs.) had a postoperative diag­nosis of carcinosarcoma (fibrosarcomatous va­riant) of the lung. Retrospective analysis of patients, records, revealed clinical stage I disease in 2 patients, and stage II disease (Nl) in 1 patient. In ali cases, clinical diagnosis was made by conventional preoperative workup including chest and brain CT evaluation, and abdominal ultrasound. Endoscopic findings were unremarkable. Cytology was positive for unde­fined epithelial carcinoma in only one patient. Based on CT findings preoperative mediasti­nal exploration was not indicated. The resection included lobectomy in 2 pa­tients and wedge resection in l. In ali patients, intraoperative frozen section yielded undefined carcinoma of the lung. Careful mediastinal no­dal sampling was done. No significant morbidity was reported. Postoperatively, the patients did not receive any adjuvai1t treatment. Della Pona C et al. Results In ali cases, postoperative histologic diagnosis revealed a spindle-cell fibrosarcomatous com­ponent (Figure 1). In one patient, the epithelial Figure l. Histologic specimcn with cpithelial and sarcomatous components. Hematoxylin-eosin original magnification x 40. component was adenocarcinoma, whereas in the other two it was squamous celi carcinoma. A special kit for immunohistochemical studies was used to detect, among other antigens, ke­ratin and vimentin in both the epithelial and 2 sarcomatous components of the tumor.1' As of December 1993, only one patient (pT2NO) is alive and without evidence of di­sease at 36 months from surgery, with the other two patients showing short postoperative survi­vals (3 and 7 months) due to disseminated metastatic disease. Discussion The carcinosarcoma is a rare primary tumor of the lung ( about 1 % ) . 1 It develops in the perip­heral as well as in the central parenchyma.3 Despite the overall poor prognosis (with a median survival of one year),3 there are some favorable indicators such as prevalent endo­bronchial growth, 1· 3 and the histological finding of a fibromatous sarcoma-like component. 1 The detection of a polypoid lesion inside the bron­chus relates to early diagnosis due to bronchial obstruction symptoms. 1 In addition, extrabron­chial tumors denote a higher degree of invasive­ness towards the chest wall. 2 Preoperative diag­nosis is rarely available also in the event of an endobronchial growth which at best yields an 4 undefined epithelial tumor.2• Crucial to the distinction among different histological subsets is the immunohistochemical analysis of a large surgical specimen. 1 Accordingly, the role of transparietal fine needle biopsy in the diagnostic work-up seems to be limited. In addition, histological differen­tial diagnosis is to be made between carcinosar­coma and spindle-cell variant of squamous celi carcinoma. 3 This distinction can be important in a prognostic perspective. 3 Histogenesis is controversial, 1 although the predominant view support the origin of the tumor from a omnipotent stem celi of the Jung tissue.5 Complete surgical removal of the pathologic material is mandatory when both diagnosis and an attempt at cure are to be made. Nodal involvement, detected in up to 25 % of the patients,1 does not seem to worsen the overall prognosis. 6 The impact of adjuvant treatment 3 on survival is unremarkable. 1• In one series of patients subjected to surgery plus chemo/radiotherapy, the longest survivor was alive at 22 months from surgery. 1 In conclusion, despite a dismal prognostic outlook, surgery for carcinosarcoma of the lung is warranted in order to provide diagnosis and tentative cure. Adjuvant treatment is of no additional benefit in improving overall survival. References l. Ishida T, Tateishi M, Kaneko S, Yano T, Mitsu­domi T, Sugimachi K, Hara N, Otha M. Carcino­sarcoma and spindlc celi carcinoma of the Jung. Clinicopathological an immunoistochcmical studies. J Thorac Cardiovasc Surg 1990; 100 (6): 844-52. 2. Rizzi A, Radaclli F, Robustellini M, Rossi G, Rocco G. II carcinosarcoma del polmone. Min Chir 1990; 45 (5): 323-25. 3. Macdc P, Moad J, Fellows D, Adams CW. Carci­nosarcoma of the Jung with hypertrophic pulmonary osteoarthropathy. Ann Thorac Surg 1991; 51 (3): 488-90. Carcinosarcoma of the lung 4. Heremans A, Verbeken E, Deneffe G, Demedts M. Carcinosarcoma of the lung. Report of two cases and rcvicw of thc literature. Acta Ciin Belg 1989; 44 (2): 110-5. 5. Humphrcy PA, Scroggs Mw, Shclburnc JD. Pulmo­nary carcinomas with a sarcomatoid clcmcnts: an immunocytochcmical and ultrastruttural analysis. Hum Pathol 1988; 19: 155--65. 6. Summcrmann E, Huwer H, Scitz G. Carcinosar­coma of the lung, a tumor which has a poor prognosis and is cxtrcmcly rarcly diagnoscd preopc­rativcly. Thorac Cardiovasc Surg 1990; (38) (49): 247-50. Radio[ Oncol 1994; 28: 376-81. The role of radiotherapy in Jung cancer treatment. Report from Slovenia Miha Debevec Institute of Oncology, Ljubljana, Slovenia Background. In order to evaluate the role of radiotherapy in lung cancer treatment in Slovenia, 276 pts treated in 1988 at the Institute of Oncology in Ljubljana were investigated. Patients and methods. There were 253 males and 23 females, aged 31---83 yrs (median 59); 6 pts had clinical St./, 19 St.//, 78 St.1/Ia, 65 St./Ilb and 108 pts had St./V lung cancer. Distant metastatic sites were as follows: generalised in 29 pts, bone in 32, brain in 21, !iver in 9 and other organs in 14 pts. OJ 267 histologically confirmed lung cancers, 126 were squamous, 62 small-cell, 44 large-cell, 23 adenocarcinomas, and 12 others (mixed, unspecified). Performance status (Karnofsky) was assessed as > 70 in 199, 50-70 in 57 and < 50 in 20 pts. Primary therapy was: RT in 189, RT + ChT in 44, OP + postop. RT in 20, OP + ChT in 2, ChT in 14, and solely symptomatic in 7 pts. In 253 pts treated by RT, tumor dose was > 5.000 cGy ( = radical) in 88, palliative in 156, and only initial in 9 pts. RT as the only method of treatment was applied loco-regionally (& supraclaviculary) in 135, to local + distant metastases in 8, only metastases in 43, whereas in 3 pts first to distant metastases and later on to the lung. Results. By the end of 1993, 7/276 (2.5 %) pts were·still alive. One-year survival of ali treated pts was 25 %, anq, two-year 9 %. OJ 75 pts irradiated loco-regionally with radical doses, 49 % survived one year, 17% two years, and 3 % 5 years. There was a significant difference in the survival according to tumour dose (p<0.001) and performance status (p<0.001), but none with reference to clinical stage 1-Illb (p<0.1) and histology (p<0.1). Treatment response was assessed after loco­ regional radiation in 79 %, and after radiation of metastases in 70 % of cases. Conclusions. Radiotherapy has proved beneficiat for the majority of our patients in terms of lije quality and short term survival. Key words: Jung neopJasms-radiotherapy, Jung cancer, radiotherapy; survivaJ anaJysis; indications, Jife quaJity Introduction Radiotherapy is the most common specific met- Correspondence to: Prof. Miha Debevec, M.D., Ph.D., Institute of Oncology, Zaloška 2, 61105 Ljub­ljana, Slovenia. Fax: + 386 61 1314180. UDC: 616.24-006.6:615.849.114 hod of Jung cancer therapy. Unfortunately, not so much owing to its success, but rather due to the Jack of more suitable treatment methods. Radiotherapy couJd be performed in cases of inoperabJe non-small and small-cell cancer with or without, chemotherapy. The aim of radiothe­rapy is to diminish the disease-related probJems caused by a Jung tumour and/or its regionaJ and 1'l1e role of radiotlzerapy in lung cancer treatment. Repon ji-om Slovenia distant metastases, and to prolong the duration of survival, sometimes also to care. Another important reason is that such specific treatment maintains the patient's hope that not everything has been !ost yet, since the patient is not tratecl only symptomatically. In the selection of patients, we had to consid­ der the actual possibilities for radiation: the capacities of radiation machines and hospitaliza­tion possibilities for in-patients. This paper is aimed to present a review of one-year turnover of patients admitted to the Institute of Oncology Ljubljana in 1988: the types of patients managed, the primary treat­ment approaches used and the results obtained. Patients and method In 1988, 795 new cases of lung cancer were registered by the Cancer Registry of Slovenia. The incidence per 100,000 population was 68.4 for males and 12.7 for females. Thoracic surgery was performed at the De­partment for Thoracic Surgery, Univ. Medica) Centre Ljubljana, and at the Thoracic Surgery of the General Hospital of the second greatest Slovenian town, Maribor. Ali radiotherapy is concentrated at the Insti­tute of Oncology Ljubljana. The radiation faci­lities comprise 2 linear accelerators, 2 cobalt unites, 1 conventional x-ray rnachine ant 2 superficial x-ray machines. Chemotherapy was performed at the Institute of Oncology Ljubljana, and at the Institute for Respiratory Diseases Golnik. There were 369 new patients with Jung cancer treated at the Institute of Oncology Ljubljana in 1988, but only 276 of them were evaluable. This represents 35 % of the registered and 75 % of those treated in the year under study. There were 253 males and only 23 females, aged 31-83 years, median 59 years; 220 of these patients were in the age of 50-70 years. Od 267 histologically confirmed Jung cancers, 126 were squamous, 62 small-cell, 44 large-cell, 23 adenocarcinomas and 12 others (mixed, un­specified), whereas 9 were microscopically not confirmed. Clinical stage was established on the basis of clinical examination, chest x-ray, bronchoscopy and abclominal ultrasonography. Other proce­clures such as chest er, bone scan ancl other x-ray examinations were performed only in the case of suspicious symptoms or cloubtful opera­bility. Six patients had stage 1, 19 stage II, 78 stage IIIa, 65 stage IIlb and 108 stage IV lung cancer. Distant metastatic sites were as follows: generalized in 29 patients, bone in 32, brain in 21, )iver in 9 ancl other organs in 14 patients. Performance status (Karnofsky) was assessecl as > 70 in 199, 50-70 in 57 and < 50 in 20 patients. Reliable data on the duration of symptoms were available for 224 patients: < 3 months in 146, 3-6 months in 37, 6-12 months in 23 ancl > 12 months in 18 patients. The leacling symptoms at the beginning of treatment were-due to primary tumour in 161, regional metastases in 12, both in 11, and clistant metastases in 72; 7 patients presentecl with general symptoms while 13 were asympto­matic. Most patients, 222 (80 % ), were referred to the Institute after previous team consultation, 43 (16 % ) patients on the basis of phone agree­ment ancl 11 ( 4 % ) without previous agreement. The primary treatment was as follows in Table l. Table l. Thc primary trcatmcnt of lung canccr pa­ticnts. Trcatmcnt mcthocl No. of pts. Pcrccnt racliothcrapy 189 68% racliothcrapy + chcmothcrapy 44 16% surgcry + racliothcrapy 20 7% surgcry + chcmothcrapy 1% chcmothcrapy 5% symptomatic only 3% Tota! 100% Racliotherapy was appliecl in 253/276 (92 % ) patients. Of these, 88 patients rcceivccl a "racli­cal" tumour dose, i.e. equivalent close > 5.000 cGy in 5 weeks, whereas 156 had a lower palliative close. In 9 patients radiation was started ancl finishecl before an expected pallia­ 378 Debevec M tive tumour dose could be achieved. In 7 pa­tients radiation was planed but was cancelled before even started because of complications. Daily doses were 250-400 cGy, radical radiation was delivered according to split-course regimen. Results By the end of 1993, 7 of 276 (2.5 % ) treated patients were stili alive. One-year survival was 25 % and two-year 9 % (Figure 1). The value 1. O N = 276 0.5 YEARS Figure l. Patients with Jung eancer treatcd in 1988 at The Institute of Oncology of Ljubljana. of radiation could be estimated in patients treated by radiation alone, without chemothe­rapy or previous surgery, either of primary lung tumour and regional metastases or distant me­tastases. Radiotherapy as the only method of treatment was applied loco-regionally in 135, to lung and distant metastases in 8, only meta­stases in 43, whereas in 3 patients was radiothe­rapy applied first to metastases and later on the lung. The survival of radically loco-regionally irra­diated patients is significantly better (p < 0.001) than that of palliatively imidited ones (Figure 2). After radical irradiation one-and two-year 1.0 P < 0.001 TO> 5000 cGy --H1 = 75 palliative RT ············· H2 = 60 ' .... -.............................. '.'-.-------­ YEARS Figure 2. Survival after loco-regional radiation therapy by tumour 5.000 cGy/5 weeks calculated by Kirk and coworkers2 is considered as radical. However, this dose was not found to be curative. Generally, it is very difficult to speak about curative radiotherapy with such a low five-year survival rate. This depends on the selection of patients rather that on the kind of radiotherapy and support therapy. Most articles report results by Smart and Hilton in 1966:3 in 40 selected irradiated patients with lung cancer five-year survival was achieved in 22.5 % . The 32 % five-year survival was reported by Zhang and co-workers 19894 in 44 patients with oper­able non-small celi lung cancer who have refu­sed surgery. In most unselected patients the survival was essentially lower and clid not ex­ceecl 5 %. Better survival of radically irradiated patients ancl patients with higher perfonnance status has confirmecl our criteria for the selection of treat­ment method: unrestricted use of high-dose racliation therapy in patients with Iower perfor­mance status ancl/or progressecl tumour does neither improve the survival nor alleviates the symptoms. Therefore, patients in worse general conclition were irracliated with palliative doses and the resulting survival rates were lower. The survival of patients with small celi cancer after combined chemotherapy ancl racliation of Jung tumour ancl regionall lymph nodes was reportecl to be somewhat better than that ob-tainecl by chemotherapy alone, but above ali, there were Iess loco-regional recurrences.5 Re­peated bronchoscopy after complete regression evident on x-ray often confirms a tumour of bronchial mucosa. Therefore, we consider irra­diation of chest in limited stage small celi Jung cancer despite possible loco-regional regression after chemotherapy inclicated for safety reasons. We clicl not perform routines prophylactic cranial irradiation. The exception were some patients incluclecl in the stucly in whom prophy­lactic cranial irracliaton was orclerecl in protocol. The preoperative racliotherapy was not per­formecl. The postoperative racliation therapy was performed in the cases of nonraclical surge­ry, metastases in mecliastinal lymph nocles ancl recurrence. Five-year survival of those patients was 17%. Conclusion Racliotherapy as a primary treatment moclality for lung cancer was performed in 92 % of our patients. Suitable for racliation were patients with in­operable or nonraclical operated lung cancer, mostly with progressed disease ancl goocl perfor­mance status, with symptoms of primary tu­mour, regional and/or clistant metastases. Treatment response was obtainecl after loco­regional radiation in 79 % ancl after racliation of metastases in 70 % of patients. We believe that the improved quality of life may result in better short term survival. Rcfercnces l. Debevec M. Guidelines for management of lung cancer. Zdrav Var 1992; 31: 166-9. 2. Kirk J, Gray WM, Watson ER. Cumulative radia­tion effect. Part I: Fractionatcd trcatmcnt rcgimcs. Cli11 Radio! 1971; 22: 145-55. 3. Smart J. Can lung canccr bc curcd by irradiation alone? JAMA 1966; 195: 158-9. 4. Zhang 1-IX, Yin WB, Zhang LJ, Yang ZY, Zhang ZX, Wang M, Chcn DF, Gu XZ. Radiuther 011col 1989; 14: 89-94. 5. Johnson BE. Concurrent approachcs to combincd chcmothcrapy and chcst racliothcrapy for the treat­mcnt of paticnts with limited stagc small celi Jung canccr. Lung Cancer 1994; 10 Suppl l: S281-87. Radio! Oncol 1994; 28: 382-5. Experience in preoperative radiotherapy f or non small cell lung cancer (NSCLC) Vladimir Zharkov, Yury Demidchik, Viacheslav Kurchin, Paul Moiseev Cancer Research Institute, Minsk, Belarus Four hundred and thirty lung cancer patients were drawn into randomized study of the induced hyperglycemia (IH) as a radiosensitizer for preoperative large fraction irradiation; 217 patients underwent IH carried out on the 1-st, 3-rd, and 5-th days or radiotherapy by the i. v. infusion of 40 % glucose; 213 patients were enrolled in the control group. There were the following criteria for the patients' enrolment into the protocol: males, age younger 66 year, cTI-3NO-2MO, histologically proved NSCLC, no contraindications for surgery and IH. The preoperative irradiation was carried out by a total dose of 20 Gy, delivered in five 4-Gy fractions per week; using 20 Me V equipment. Surgery was carried out three days after radiotherapy. Patients with pNI-pN2 received additionally postoperative radiotherapy: 30-36 Gy, 2 Gy per fraction. In stage IIIA, patients' survival was significantly higher for IH compared with the control group: the median survival was 2.0 ± 0.3 years in the study group vs 1.1 ± 0.23 in the control group (P< 0.05). The duration of survival differed significantly for the second year of follow up (50.0 ± 6.0 vs 32.3 ± 4.6 %, P < 0.05). For the fifth year the survival rates were: 29.2 ± 5.2 for the study group vs 17.2 ± 4.2 % in the control group; in N2 patients: 23.9±4.3 vs 10.3±3.4% (P<0.05). Key words: non-small celi Jung carcinoma; preoperative radiotherapy therapy- Introduction Preoperative standard radiotherapy ( 40-60 Gy) for operable Jung cancer shows no clear evi­dence of improvement in the survival. Even for marginally resectable cases as superior sulcus cancers, the problem remains open.1 According to L.H. Einhorn2 there were three phase III non randomized studies that revealed no diffe- Corespondence to: Vladimir Zharkov, MD, Depart­ment for Thoracie Surgery, Caneer Researsh Institute, P.O. Lesnoe-2, 223052 Minsk, Belarus. UDC: 616.24-006.6:615.849.114 rence in long-term results. These researches were aiming at gross eradication of the primary tumor, considering the well-known fact that responders have longer survival than other pa­tients. The program usually takes from 4 to 6 weeks, and further resection is attempted 3 to 5 weeks after radiotherapy. The whole period before surgery usually takes 2 or 3 months, and this interim tumor can produce metastases. Another way is to use large-fraction irradia­tion to impact the conditions of ablastics at the tirne of surgery. This treatment is shorter and takes 4 or 5 days. From the data presented by co-operating centres in Eastern Europe it is evident that such a program improves survival Experience in preoperative radiotherapy for NSCLC 383 in comparison with surgery alone. 3 But despite the use of thoracic irradiation, the local recur­rence was an appreciable site of treatment failure. This concurs with the fact that curability of tumors depends on strongly marked devitali­zation of clonogenic tumour cells. The problem is to overcome the tumour hypoxia that may cause a recurrence of cancer after complete control. Obviously it is impossible to escalate the dose of irradiation without running a risk of complications. The present paper reports the results of using preoperative large fraction radiotherapy with or without radiosensitization by induced hyper­glycemia (IH) in operable lung cancer patients. Material and methods Four hundred and thirty NSCLC patients were enrolled in the randomised study (From Decem­ber 1983 till December 1988) of IH as a radio­sensitizer for preoperative radiotherapy. The research was aimed at determing whether radio­sensitising can increase tumour alteration and improve the long term survival. Patient's selection The criteria for patients' enrolment in to the protocol were as follows: 1) males; 2) age under 66 years; 3) cT1­3NO-2MO; 4) histologically proved NSCLC; 5) no contraindications for surgery and IH (low cardio-pulmonary reserve, diabetes mellitus). Randomization Ali the enrolled patients were randomized into two groups depending on the method of pre­operative treatment (Figure 1). The comparison of the baseline features of both groups revealed no statistic ally significant differences (Table 1). Table l. Paticnt's ::, V) u ::ti l.. ....... i:o C "' .2 Shepherd 10 summarized the results of 15 phase "'::, "'"' > .88u, "' o 4;-.\ d. Vc:ic:iZ 0 t, II pilot trials of induction chemotherapy (5 studies C) 3 en 0 >, .... with chemotherapy alone; 4 studies with chemot­ - 2 o herapy and radiotherapy; and 6 studies with ::::1 v rl • >-.. 0 ";; OOtn :;,-C/JOC/JC/J . C: e v (l) (l) ::::1 0 motherapy and radiotherapy induction regimens, ..C: ..C: 8 8 i;; C: o g .2 §: and overall responses were observed in more "' 8 2 1 o '6 than 50% of patients, although the complete <1) U_g rl--!C"l(")NNNN C::00 ::iz < z '°'°oV);;,;;,<:00 o,..., .5 > u 0 clinical remission rate was Iess than 15 % , and "O u o C the pathological complete response rate was ..C: 2 i"' o..Ei usually less than 10 % . 10 After induction therapy, X u (/J .1..01 V) ·X· <:C) \O rr) \O \O . approximately 70% of the patients were eligible -stN rl rl \O t-.C:-, "O > for thoracotomy, and complete resection was z E 0 . :r ro b!) l--!...-jl--!)-o!)-()-o! .cn ui---. l--1 ,-; ,-; 1--! ,-; -----_g t, 8 >, 0 0ro ·-­C possible in 16 % of the patients.10 Two-and l--11--!1--!1 l--!l--ll--l -' C) en-+-> lI l 1 1 1 ' (/J ,-;)-o!...-.()-o!)--1)-o! H.-;,-;,-..... 1 o.. x 2 ><: E-< 0 2 C) three-year survival ranged from 25% to 30%.10 .... v >, >, .,,,..c: From the review of these 15 studies it is clear .o o O) o.. 0 o.. . "' :s ;,;:a8 2 en "' that induction therapy followed by surgical resec­ 0 o8en tion is feasible, and even though toxicity is en­ u "' 'o ....15.. 8 ..C:f-< s-, "'1. 0 00lf)NNN('f"J'tj'-N007\0 i;l-0::: '" countered, in most studies this does not appear -8 o ..C: 000\(")'tj-('f')"tj-(")\O['---.V)(") >, o u tn --1--!-.:f"NN ......-1 _g -. .g to be unacceptable. However, randomised trials "' o.. z o.-5 ·en .... are needed in order to document a beneficial 0 -§ . . >, o.. u ..C: o .... u -. u ·-effect on survival. o 8 <:C)°' 00 \O \O ('f') 1 ­ 0 <:C) 'T \000lr)(I") ><: + o. Unfortunately, only few prior randomised stu­ .1:§.S ,-; E-< 'U -stN U .5 t; dies have adressed the benefit of preoperative 0 f:E ·s chemotherapy, with the first three studies being ::, o § > ·s C te u i-, c,j ::::1 C) ..C: . i-, without a beneficial effect for patients receiving "' ,::l ><: :::i ·a 23 C) :s b[}'"d -chemotherapy.11-13 Among the patients in these - o 2 f-< :I: C (:l.. "tij ..C:: en »..S :;:E;:5. < o+ ·. trials, the three year survival rate ranged from <1) :,:_ 1 g. 'B><:><:XZ:i><;r:i...r:i..r:i.. . c,f--"'"O UE-< rou00 treatment. 11-13 o::: C 1 .§u i:oi:<:: However, a recent study by Rosell et al. 14 ,.; >,>,>, ............ CC i-, >-. >. 1-< :.-. >. >. ;... 0 (l) (l) C) o.. 1n C:: reported a beneficial impact on survival by the < -:i:r.i.i; zu.,. ;, use of a combination chemotherapy regimen con­ ;... -+-> 1 :E Ql (!) (1) C) oJ) oO bl} ;... ;.... ;.... ;; ::, ::, ::, (/J (/J (/J C e_n t:D e_n e,n ::, ::, ::, ::, (/J (/J (/J (/J -01 . u o ::, ::, (/J (/J The role of chemotherapy in non-small celi lung cancer 389 sisting of mitomycin C, ifosfamide and cisplatin every 3 weeks for three courses before surgery. Ali patients had stage llla NSCLC, 30 patients were randomized to chemotherapy + surgery and 30 patients to surgery alone. Both groups of patients received mediastinal radiotherapy 1.8 Gy to 2.0 Gy for 5 days per week until a cumulative dose of 50 Gy. Median survival was 26 months for patients receiving neo-adjuvant chemotherapy compared to 8 months in patients without it (p<0.001), and the median duration of disease-free survival was 20 months compared to 5 months, respectively (p<0.001).14 This study has been somewhat critizied because of a low number of patients randomised, as well as be­cause of a short follow-up period and multiple interim analyses. 15 Recently, another group of investigators from M. D. Anderson Cancer Center have also publis­hed a report on beneficial effect in patients receiving neoadjuvant chemotherapy. 16 Twenty eight patients were randomized to receive 3 cour­ses of cyclophosphamide, etoposide and cisplatin before surgery, followed by 3 more courses after surgery for responding patients, while 32 patients were randomized to surgery alone. Ali patients had potentially resectable clinical stage III a disease. The estimated median survivals were 64 months and 11 months respectively (p<0.008). Even though these results by Rosell et al. 14 and by Roth et al. 16 are encouraging, neo-adjuvant therapy should still be considered experimental, and shoulcl at this point be used solely in control­led investigational protocols. Chemotherapy far advanced disease Single agents A number of single agents are shown to be of moderat activity in advanced NSCLC (Table 2). Table 2. Established single agents in NSCLC. A cumulative response rate of more than 15 % has been encounterecl for vinblastine, 17 ifosfami­cle, 18 cisplatin, 19 vinclesine,20 and mitomycin-C21 (Table 2). The majority of responses are only partial while complete remissions are rare and occur in less than 5 % of patients treatecl. 22 Recently, the literature on new cytostatic drugs in the treatment of patients with NSCLC reportecl in the literature since 1985 has been reviewecl together with new methocls for aclministration of establishecl clrugs. 23 With respect to new methocls for drug administration two well known cytostatic drugs, ifosfamicle ancl etoposicle, have recently been evaluatecl in trials using oral aclministration insteacl of the usual intravenous route. A cumu­lative response rate for oral etoposicle of 17 % was shown in 4 studies including the total number 27 of 104 patients. 24-Also oral aclministration of ifosfamicle yields a cumulative response rate of 18 % when the close intensity is 7 g or more during a 4 week periocl.28 Among the new clrugs tested, the most active are outlinecl in Table 3. Among the most promi­sings secm to be campthothecin-11 (CPT-11), gcmcitabine, vinorelbine, taxol, fotemustine and zeniplatin which have ali shown rcsponse rates above 20 % among previously untreated pa­tients. c3 Taxotere has shown a promising cumula­tive response rate of 32 % among a total of 92 31 patients included in 3 stuclies.29-The semi-syn­thetic campthothecin analogue topotecan, which like CPT-11 also is a topoisomerase I inhibitor, has been evaluated in two stuclies inclucling 19 patients.32-33 Patients were treatcd with O. 5mg/m2 daily for 5 days ancl the cumulative response rate 33 was 26 % . 32•Also, the antimetabolites 10­EDAM ancl trimetrexate ancl the platinum analo­gues carboplatin ancl (glycolate-0.0) cliammine­platinum II are of interest with cumulative re- Agent with No. of pts Cumulative response rate (%) Authors > 15 % activity Vinblastine lfosfamide Cisplatin Vindesine Mitomycin 22 420 568 295 115 27 26 21 18 18 Schulman et al, 198217 Drings P, 198718 Bunn PA, 198919 S0rensen et al, 199320 Kris MG, 198921 390 SJrensen 1 B Table 3. Effectiveness of the most active new drugs in patients with previously untreated non-small celi lung cancer. Cumulative Response Treatment (overall response rate) rate >20% Taxotere (32 % , 3 studies, 92 patients) Campthothecin-11 (32 % , l study, 72 patients) Gemcitabine (26 % , 2 studies, 87 patients) Topotecan (26 % , 2 studies, 19 patients) Vinorelbine (23 % , 4 studies, 381 patients) Taxol (22 % , 2 studies, 49 patients) Fotemustine (21 % , 1 study, 29 patients) Zeniplatin (21 % , 1 study, 28 patients) :::C:15% 10-EDAM (19%, 2 studies, 64 patients) (Glycolate-0,0) diammine-platinum (II) (17 % , 1 study, 38 patients) Carboplatin (16 % , 3 studies, 171 patients) Trimetrexate (16 % , 4 studies, 141 patients) sponse rate above 15 % in previously untreated patients.22 Combination chemotherapy It is generally accepted that combination chemo­therapy is more active than single agent treatment in patients with NSCLC. The response rate in patients with NSCLC according to the number of cytostatic agents employed was examined by Donnadieu et al. 34 The authors perfonned a meta-analysis of 100 studies including a total of 6,247 patients with NSCLC receiving chemothe­rapy. Among these patients, 23% had limited disease and 77 % disseminated disease. An ove­rall 25 % objective response rate was obtained with a 3 % complete response rate. Overall re­sponse rate, partial response rate and complete response rate were all significantly better in pa­tients with limited disease being 34 % , 27 % and 5 % , respectively, compared with 22 % , 19 % and 3 % , respectively in patients with extensive di­sease (p = <0.001).34 The response rate in NSCLC according to the number of cytostatic agents observed in the meta­analysis by Donnadieu et al.34 is shown in Table 4. Twenty-seven studies used single agent che- Table 4. Response rate in NSCLC according to no. of cytostatic agents (Metaanalysis of 100 studies, 6,247 pts). No. of cytostatic agents No. of pts Response rate % 1 914 6 2 1,643 26 3 2,185 28 4 1,070 28 s 226 so 6 180 24 2-6 5,304 28 Modified from Donnadieu et al., Lung Cancer 1991.34 Single agents (27 studies including 914 patients) vs. combination chemotherapy, (73 studies including 5,404 patients) p<0.001. motherapy with an overall response rate of 6 % among a total of 914 patients while 73 studies used combination chemotherapy with two to six cytostatic agents with a 28 % cumulative response rate among the 5,304 patients included (p = <0.001).34 These results point towards a better effect of combination chemotherapy than of the single agent treatments evaluated. A large number of combination chemotherapy regimens has been evaluated in NSCLC. The most frequently used regimens among the 100 studies included in the meta-analysis by Donna­dieu et al. are shown in Table 5, yielding response rates from 22 % to 41 % . 34 A review by Soyez and Kleisbauer35 have suggested that the most active combinations consist of cisplatinum toget­her with either a vinca alkaloid or etoposide with or without mitomycin-C. The cumulative re­sponse rate in 14 trials with a total of 565 patients employing combinations of cisplatin plus vinde­sine were 37%, while the combination of cisplatin and etoposide was evaluated in 32 studies inclu­ding 1,534 patients with a response raate of 27%. A combination of cisplatin, vindesine and mito­mycin-C was evaluated in 8 trials including 485 patients, giving a cumulative response rate of 41 % , and a combination of cisplatin, vinblastine and mitomycin-C was evaluated in 5 trials inclu­ding 301 patients with an overall response rate of 37%.35 The role of chemotherapy in non-small celi lung cancer Table S. Objcctivc rcsponsc ratc with thc most frcquently used chcmothcrapy regimcns in NSCLC. Cumulativc results Regimcn Numbcr of studics % resp. 11 VBL + MMC + COOP 6 41 37 34 VOS + COOP + VP16 3 VOS+ MMC 3 VOS+ COOP 282 8 30 29 MTX + ADR + CTX + CCNU (MACC) 5 CTX + AOR + COOP (CAP) VP16 + COOP 12 722 409 25 24 22 13 CTX + AOR + MTX + PCZ(CAMP) 5 Modificd from Oonnadicu ct al., Lung Cancer 1991.34 COOP = cisplatin; VP16 = ctoposide; AOR adriamycin; CTX = cyclophosphamide; VDS = vindesinc; VBL = vinblastinc; MTX = mcthotrcxatc; PCZ = procarbazinc; MMV mitomycinc. Randomised studies comparing a combination of cisplatin plus vindesine to a combination of cisplatin plus etoposide has recently been revie­wed by Sd>rensen and Hansen.20 These combina­tion chemotherapies were equally active in rando­mised studies, both with respect to response rates and survivaI.20 In the same review, the effect of the addition of 1 to 2 other drugs to these combinations has not yielded an increase in the surviva!,20 and the addition of 1 or more cytostatic agents to a two-drug regimen consisting of either cisplatin plus etoposide or cisplatin with a vinca alkaloid has not demonstrated uniformly increa­sed response rates. 20•34 •35 No combination chemotherapy has at this point been shown to be clearly more active than other commonly used combination chemotherapy regi­mens. Thus, a standard chemotherapy for advan­ced NSCLC has not been established at this point. Chemotherapy compared to supportive care in advanced NSCLC Severa! studies have compared the effect of com­bination chemotherapy with best supportive care in randomized trials. 36· 43 Ali studies showed a trend towards longer survival for patients receiv­ing chemotherapy compared to patients without (Table 6), and this trend was statistically signifi­ 4o,4t ,44 , cant in four studies. 37Souquet et al.45 published a meta-analysis of 7 studies of combination chemotherapy versus 3541 the best supportive care. Endpoints in the meta analysis were the number of deaths at 3-, Table 6. Combination chemotherapy (CT) vs. bcst supportivc carc (BSC) in NSCLC. No. of pts Median survival (wks) Regimcn P-value Authors (reference no.) CT BSC CT BSC 30.5 CCMEP 62 61 MACC 20 19 0.15 Ccllcrino, 198836 Cormicr, 198237 s.o 14.4 0.005 VP EP VP 32 22 31 21 19.9 0.09 Ganz 198938 21.6 27.6 16.0 O.S Kaas;, 198939 Quoix, 199140 0.03 VP 44 or CAP VP 105 32.6 0.01 17.0 Rapp, 19884 1 24.7 o.os 96 27.0 17.0 0.075 Woods, 199042 MACC 175 32.0 20.0 0.01 Buccheri, 199043 CMP 52 50 36.4 17.l <0.0001 Cartei, 199344 CCMEP -cyclophosphamide + CCNU + mcthotrcxate + etoposidc + cisplatin; MACC mcthotrcxate + adriamycin + CCNU + cyclophosphamide; VP -vindesinc or vinblastine + cisplatin; EP -ctoposidc + cisplatin; CAP -cyclophosphamidc + adriamycin + cisplatin; CMP -cyclophosphamide + mitomycin C + cisplatin. Sd,rensen J B 6-, 9-, 12-, and 18 months. The analysis showed a statistically significant reduction in mortality at 3 months (p = <0.001) and 6 months (p = <0.00001).4:' Even though the odds ratio for mortality also was lower at 9-, 12-, and 18 months, this trend was not statistically significant, as p values were p = O. 02, p 0.03, and p 0.15, respectively.45 These results can also be expressed as the risk reduction for patients recei­ving chemotherapy as cornpared to best suppor­tive care, which was found to be 0.65 at 3 months, 0.73 at 6 months, 0.86 at 9 months, 0.91 at 12 months, and 0.96 at 18 months, respective­ly.45 Although small, this increase in survival suggests that combination chemotherapy has a role for patients with non resectable non-small celi lung cancer, though the effect seems to be most pronounced with respect to increase in short term survival. Discussion The recent results of adjuvant combination che­motherapy are somewhat encouraging and sug­gest that patients with apparently completely resected NSCLC may benefit with respect to disease free survival, but at the cost of toxicity. A significant impact on overall survival has as yet not been documented. It is also apparent that some of the widely used regimens employing either etoposide and cisplatin or a vinca alcaloid and cisplatin have not been evaluated in an adjuvant setting. The effect of such regimens should be evaluated together with a careful des­cription of known prognostic factors in order to identify the patient groups which are most prone to benefit from the treatment. With respect to neo-adjuvant chemotherapy, it has repeatedly been shown that such treatment is feasible. Additionally, the two most recent studies, 14• 16 have also reported a beneficial impact on overall survival. However, there is still limited data from randomized studies, and at this point induction chemotherapy followed by surgical in­tervention for locally advanced non-small celi lung cancer must stili be viewed as experimentaL A number of new cytostatic agents have been evaluated within the last few years in NSCLC patients with advanced disease, and severa! of these agents have repeatedly yielded response rates above 20 % . Severa! of these drugs are real innovations, (with quite novel structure), and these agents hold promise of improvement in the therapy of NSCLC. In the planning of new trials, the evaluation of such agents should have a high priority. A large number of combination chemotherapy regimens have been evaluated in patients with advanced NSCLC without any of these regimens being documented as clearly superior. Thus, a standard treatment has as yet not been establis­hed. The impact of combination chemotherapy on the survival of patients with advanced NSCLC has repeatedly been shown to be beneficial with respect to the quantity of life when evaluated in randomized studies against best supportive care. This suggests that the clinical course of NSCLC can be improved by chemotherapy, but at the expense of toxicity. Accordingly, the quality of life in patients receiving chemotherapy must be balanced against the possible gain in the quantity of life. Decisions as to whether chemotherapy for NSCLC should be used in clinical practice, and whether economic resources should be allocated to such issues, are not simply not tecnical que­stions that could be answered by clinicians or health economists. lndividual patients and their relatives must be involved in the decision-making process to address adequately the preferences for the potential benefits, harms and costs associated with chemotherapy. To ensure such decision ma­kings are based on a firm ground we sorely need information on the quality of life with respect to different treatment options. References l. Slack NH. Bronchogenic carcinoma: Nitrogcn mu­stard as surgical adjuvant and factors influencing survival. Cancer 1970; 25: 987-1002. 2. Brunner KW, Marthaler T, Muller W. Effects of long-term adjuvant chemothcrapy with cyclophos­phamide (NSC-26271) for radically rcsected bron­chogenic carcinoma. Cancer Chemother Rep 1973; 4: 125-32. The role of chemotherapy in non-small celi lung cancer 3. Shiclds TW, Humphrcy EW, Eastridgc CE ct at. Adjuvant canccr chcmothcrapy aftcr rcscction of carcinoma of thc lung. Cancer 1977; 40: 2057-62. 4. Shiclcls TW, Higgins GA, Humphrey EW et al. Prolongecl intermittcnt adjuvant chemothcrapy with CCNU and hydroxyurca aftcr rcscction of carci­noma of the Jung. Cancer 1982; 50: 1713-21. 5. Girling OJ, Stott H, Stcphens RJ ct al. Fiftccn year follow-up of ali paticnts in a stucly of postoperativc chcmotherapy for bronchial carcinoma. Br J Cancer I 985; 52: 867-Tl. 6. Holmes EC, Gail M. Surgical ad_juvant therapy for stagc 11 ancl stagc lil aclenocarcinoma ancl large-cell unclifferentiatccl carcinoma. J Clilz Oncol 1986; 4: 710-15. 7. Lacl T, Rubcnstcin L, Saclcghi A. Thc bencfit of acljuvant trcatmcnt for rcscctccl loeally aclvancccl non-small celi lung cancer. J Ciin Oncol 1988; 6: 9-17. 8. Niiranen A, Niitamo-Korhonen S, Kouri M et al. Adjuvant chemotherapy afler radical surgcry for 11011-small celi Jung cancer: A ranclomizccl study. J Ciin Onco/ 1992; 10: 1927-32. 9. Felci R, Rubinstein L, Thomas PA and the Lung Cancer Stucly Group. Adjuvant chemotherapy with cyclophosphamide, doxorubicin, ancl cisplatin in pa­tients with complctely rcsected stagc I 11011-small celi lung canccr. J Natl Cancer l11St 1993; 85: 299­306. 10. Shephercl FA. Incluction chemotherapy for locally advanced non-small celi Jung canccr. Ann Thorac Sw.. 1993; 55: 1585-92. 11. Dautzcnbcrg B, Bcnichou J, Allard P cl al. Failurc of thc pcrioperativc PCV ncoacljuvant polychemol­herapy in resectablc bronchogcnic non-small celi carcinoma: Results from a ranclomized phase II tria!. Cancer 1990; 65: 2435-41. 12. Fossella FV, Ryan B, Dhingra I-1 et al. lnterim report of a prospectivc ranclomizecl tria! of neo­acljuvanl chemothcrapy plus surgery vs surgery alone for IIIA non-small celi Jung cancer. Proc Am Soc Ciin Oncol 1991; 10: 240. 13. Pass HI, Pogrebniak HW, Steinbcrg SM, Mulshine et al. Ranclomizcd tria! of neoadjuvant thcrapy for lung cancer; intcrim analysis. Ann thorac Surg 1992; 53: 992-8. 14. Roscll R, Gomez-Coclina J, Camps C ct al. A ranclomizccl tria! comparing preoperative chcmo­thcrapy plus surgcry with surgery aJonc in paticnts with non-small ccll lung canccr. N Engl J Med 1994; 330: 153-8. 15. McLachlan S-A, Stocklcr M. Lcttcr to thc cditor. N Engl J Med 1994; 330: 1757. 16. Roth JA, Fossclla F, Komaki R ct al. A randomizccl tria! comparing periopcrativc chcmothcrapy and sur­gcry with surgcry alone in rescctablc stagc III A 11011-small-ccll lung canccr. J Natl Cancer Inst 1994; 86: 673-80. 17. Drings P. Jfosfamiclc in thc trcatmcnt of bronchial carcinoma. In: Brade WB, Nagci EA, Sccbcn S eds. lf'o.ljimzide in Tumor Therapy. New York; Kargcr, 1987; 294-318. 18. SchuJman P, Buclman DR, Vincigucrra V, Wcisel­bcrg L, Abrams S, Dcgman T. Phasc II stucly of diviclccl closc vinblastinc in non-small celi broncho­gcnic carcinoma. Cancer Treat Rep 1982; 66: 171. 19. Bunn PA. Thc expanding role of cisplatin in thc trcatmcnt of 11011-small celi lung cancer. Seminar Oncol 1989; 16 (suppl 4): 1()-21. 20. S(Drcnscn JB, Hansen 1-IH. !s thcrc a role for vinclcsinc in thc trcatmcnt of non-small celi Jung canccr? fm,est New Drug.1· 1993; H: 103-33. 21. Kris MG. Mitomycin in combination chcmothcrapy regimens for paticnts with advancccl non-small celi Jung canccr. In: Gralla RJ, Einhorn LI-! ccls. Treat­ment ami prevelllion of small ce/1 lung cancer mul nmz-small ce/1 /ung cmzcer. New York: Royal Socicty of Medicine 1989; 101-9. 22. S(Drcnscn JB, Clcrici M, l-lanscn HH. Singlc agent chemothcrapy for aclvancccl aclcnocarcinoma of thc Jung. A rcvicw. Cancer Chemother Pharmacol 1988; 21: 89-102. 23. S0rcnsen JB. Trcatmcnt of non-small celi lung canccr: New cytostatic agents. Lwzg Cancer 1993; 10: 173-87. 24. Gatzcmeicr U, Ncuhauss R, Kcchmayr M. Singlc agent oral ctoposiclc in advancccl NSCLC (Chronic daily) ancl in cldcrly patients with SCLC. Lung Cancer 1991; 7 (Suppl 102). 25. Saxman S, Loehrcr PJ, Logic K ct al. Phase II tria! of claily oral ctoposiclc in paticnts with aclvancccl non-small celi Jung canccr. lnvest New Drugs 1991; 9: 253-6. 26. Estapc J, Palombo H, Sanchcz-Lloret J ct al. Chro­nic oral etoposiclc in non-small celi Jung carcinoma. Eur J Cancer 1992; 28A: 835-7. 27. Waits TM, Johnson DH, Hainsworth JD ct al. Prolonged administration of oral ctoposiclc in non­small celi Jung cancer: a phase II tria!. J Ciin Oncol 1992; 10: 292-6. 28. Mancgolcl C, Bischoff M, Fischer JR, Pcuckcrt M ct al. Phase 1/ll tria! of oral ifosfamiclc (01)/Mcsna (M) in aclvancccl 11011-small celi Jung canccr (NSCLC). Proc Am Soc Ciin Oncol 1990; 9: 245. 29. Ccrny T, Wanclcrs J, Kaplan S et al. Taxotcrc is an activc drug in 11011 small celi Jung (NSCLC) canccr: A phase II tria! of thc early clinical trials group (ECTG). Proc Am Soc Ciin Onco/ 1993; 12: 331. 30. B urris H, Eckhardt J, Fielcls S et al. Phasc II trials of taxoterc in paticnts with non-small celi Jung canccr. Proc Am Soc Ciin Oncol 1993; 12: 335. 31. Rigas JR, Francis PA, Kris MG ct al. Phasc 11 tria! of taxotcrc in non-small celi lung canccr (NSCLC). Proc Am Soc Ciin Oncol 1993; 12: 336. Sil>rensen J B 32. Rowinsky EK, Grochow LB, Hcndricks CB ct al. Phasc I and phannacologic study of topotccan: A novci topoisomcrasc I inhibitor. J Ciin Oncol 1992; 10: 647-56. 33. Vcrqcij J, Luncl B, Bcyncn J ct al. Clinical studics with topotccan: Thc EORTC cxpcricnce. 7th NCI­EORTC symposium on new drugs in cancer thera­py, Program and Abstracts. 1992; 118. 34. Donnadieu N, Paesmans M, Sculier J-P. Chemo­therapy of non-small celi I ung cancer according to disease extent: a meta-analysis of the literature. Lung Cancer 1991; 7: 243--52. 35. Soyez F, Klcisbauer J-P. Advances of chemotherapy in non small celi lung cancer. In: Navelbine (Vino­relbine). Update ancl new trend.1·. Montrouge: John Libbey Eurotext Ltd., 1991; 69-77. 36. Cellerino R, Tummarcllo D, Guidi F et al. A ranclomiscd tria! of alternating chemothcrapy versus best supportive carc in aclvancecl 11011 small celi lung canccr. J Cin Onco! 1991; 9: 1453--61. 37. Cormicr Y, Bergeron D, Laforgc J ct al. Benefit of polychemotherapy in advancecl non small celi lung bronchogenie carcinoma. Cancer 1982; 50: 845-9. 38. Ganz PA, Figlin RA, Haskell CM et al. Supportive care versus supportive care and continuation che­motherapy in metastatic non small celi lung cancer: cloes chemothcrapy make a differcncc? Cancer 1989; 63: 1272-8. 39. Kaasa A, Luncl E, Thorocl E et al. Symptomatic trcatment vcrsus combination chemotherapy in pa­ tients with non small celi lung cancer, extensive clisease. Cancer 1991; 67: 2443-7. 40. Quoix E, Dieteman C, Charbonneau J et al. La chimiothcrapie comportant du cisplatine est-elle utile clans le cancer bronchique 11011 microcellulaire au stade IV? Rcsultats d'une etude ranclomisce. Buli Cancer 1991; 78: 341--6. 41. Rapp E, Pater J, Willan A et al. Chemotherapy can prolong survival in patients with advanced 11011 small celi lung cancer: report of a Canadian multi­center randomized tria!. J Ciin Onco! 1988; 6: 633-41. 42. Woods RL, Williams CJ, Levi J et al. Is chemo­therapy worthwhile in aclvanced 11011 small celi lung cancer? A prospeetive randomised tria!. Br J Cancer 1990; 61: 608-11. 43. Buccheri GF, Ferrigno D, Curcio A et al. Continua­tion of chemotherapy versus supportive care alone in patients with inoperable 11011-small celi lung can­cer and stablc disease after two or three cycles of MACC. Results of a randomized prospective study. Cancer 1989; 63: 428-32. 44. Cartei G, Cartei F, Cantone A et al. Cisplatin­yc-lophosphamide-mitomycin combination che­motherapy with supportive care versus supportivc care alone for treatment of metastatic 11011-small-cell lung cancer. J Natl Cancer lnst 1993; 85: 794--800. 45. Souquct PJ, Chauvin F, Boisscl JP, Cellerino R et al. Polychemotherapy in advanced 11011 small cel! lung cancer: a meta-analysis. Lancet 1993; 342: 19-21. Radio/ Oncol 1994; 28: 395-7. Cisplatin/etoposide combined with interferon-gamma in non-smaH celi lung cancer Robert Pirker,1 Christian Prior,2 Robert Voves,3 Meinhard Kneussl,4 Stephan Oroszy, 2 Gerhard Krajnik, 1 Sabine Zochbauer, 1 Heinz Huber1 1 Clinicfor Interna! Medicine I and 4 Clinicfor Interna/ Medicine IV, Vienna; 2 Department for Interna! Medicine, Innsbruck; 3 Landeskrankenhaus, Graz; Austria In order to assess the clinical efficacy of inte,feron-gamma (IFN-gamma) in non-small celi lung cancer (NSCLC), JFN-gamma was added to three cycles of chemotherapy (cisplatin, etoposide) and applied as maintenance therapy in patients with advanced NSCLC. So far, 32 patients have been admitted to this tria!. Twenty-nine patients were evaluable with regard to response. Complete remissions, partial remissions and minor responses were observed in O%, 7 % and 28 % of the patients, respectively. Stable disease was seen in 1 (3 %) patient and progressive disease in 18 (62 'ľ,) patients. In the case of progressive disease, the treatment was discontinued. The duration of median survival was 7 months and the 1-year survival rate was 40 %. Only few patients did receive JFN-gamma maintenance therapy. Further fo!!ow-up of the patients is required in order to determine the influence of this combined treatment on long-term survival. Key words: carcinoma, lung cancer, non-small celi lung cancer-drug therapy; interferon type II; chemotherapy, cisplatin, etoposide Introduction A recent meta-analysis has indicatcd that corn­bination chernotherapy for unresectable non­srnall celi Jung cancer (NSCLC) results in pro­longed survival of the patients. 1 However, in order to further improve the clinical outcome of this disease, new cytotoxic drugs and/or biological agents are required. Previously, interferon-gamma (IFN-gamma) Correspondencc to: Robert Pirker, MD, Clinic for Interna! Medicine I, Wiihringcrgiirtcl 18-20, 1090 Vienna, Austria. Phonc: + 43140400 4429; Fax: + 431 404004461. UDC: 616.24-006.6-085 was found to inhibit tumour growth by its antiproliferative propertics and by activating effector cells. 2 J n addition, synergistic or addi­tive effects of interferons on the activity of cisplatin have been reported.3 Thus we initiated a clinical tria] on chemotherapy combined with 1FN-gamma in patients with aclvanced NSCLC in order to assess whether the acldition of IFN-gamma improves the clinical outcome in patients with NSCLC. Here we report on this ongoing tria!. Patients and methods From January 1992 to December 1993, 32 pa­tients (6 females, 26 males, median age 59 11 (34 %) Pirker R et al. years) with unresectable NSCLC were admitted to this study. The inclusion criteria werc histo­logically confirmed cliagnosis of NSCLC, good performance status, no prior chemotherapy, adequate renal as well as bone marrow func­tion, and written informed conscnt. The tumour was stagcd as described.4 Chcmotherapy consisted of cisplatin (25 mg/ m2 per day intravenously on days 8, 10 and 12) and etoposicle ( 100 mg/ 1112 per day intravenous­ly on days 8, 10 and 12). IFN-gamma (Bendcr + Co GmbH, Vienna, Austria) was given sub­cutaneously (l()0.tg per day on days 1, 3, 5, 8, 15, 17 and 19). In case of fevcr associatecl with IFN-gamma application, paracetamol was ad­ded. These cornbincd treatment cycles were rcpeatecl every 3 weeks. After three cycles, the patients procecded to maintenance therapy with IFN-gamrna (100 .tg subcutaneously three-times per week). Any progression of the disease resulted in discontinuation of treatment. Survival analysis The duration of overall survival of the patients was estimated according to Kaplan-Meier.5 The analysis was based on the patients' status as of the end of March 1994. Results The patients' characteristics are summarized in Table 1. Fifty-three percent of the patients Table 1: Palients' characteristics. No. 32 Age median 59 range 41-70 Sex female/male 6/26 presentcd with stage IV disease. Except in patients with progressive clisease, response was evaluated after 3 treatment cycles. So far, 29 patients were evaluable for response (Table 2). Table 2: Response to trealment (after 3 cycles) in 29 cvaluable patients. Rcsponse No. (0/c,) of patients complete remission O (0%) partial remission 2 (7%) minor response 8 (28 % ) stable disease 1 (3%) progressive disease 18(62%) The overall response was poor with complete and partial remissions in O (O%) ancl 2 (7 % ) patients, respectively. Minor responses and sta­ble disease were seen in 8 (28 % ) and l (3 % ) patients, respectively. During treatment, 18 (62 % ) patients presented with progression. The most frequent treatment-related side effects were leukopenia (mostly WHO grade 1 or 2), nausea and vomiting, flu-like symptoms and fever (data not shown). Only five patients have received IFN-gamma maintenance therapy. The reasons for unfeasi­bility of maintenance therapy included progres­sive disease, side effects of IFN-gamma and patients' refusal of further treatment. According to Kaplan-Meier analysis, the me­dian duration of overall survival was 7 months and the 1-year survival rate was 40 % (Figure 1). p R o B A B 1 0,6 ,_ Hislologic type adenocarcinoma 17 (53%) squamous carcinoma ll (34%) 1 large celi carcinoma 4 ( 13 % ) o F 0,2 Stage IIIA o 4 (13 % ) o...-..-..-...-..-..-.. IIIB o 2 4 a a w . M m m m n M . . u MONTHS Prior treatment exploralive thoracotomy 4 ( 13 % ) Figure 1: Survival of paticnts. The overall survival of radiation [ (3 %) patients was calculated accorcling to Kaplan-Meicr:' chemotherapy 0(0%) OS = overall survival. Cisplatinletoposide combined with inte,jeron-gamma in non-small cel! Jung cancer Discussion Although the response rate seen in our study was low as eomparecl to other studies, the duration of overall survival of our patients was similar to the one reported in another reeent study. 6 The low response rate seen in our study might be due to the rather Iow dose of cisplatin (75 mg/m2 per eyele) and due to the faet that the majority of patients had stage IV disease whieh is less responsive to ehemotherapy. Thus the initial analysis of our study suggests that the addition of IFN-gamma to chemotherapy does not improve elinieal outcome in patients with advaneed NSCLC. This is consistent with a recent randomized tria! in whieh the addition of IFN-gamma or IFN-alpha plus IFN-gamma to chemotherapy did not influenee the response rate or survival of the patients.6 Whether higher doses of IFN-gamma might be efficacious, remains unclear. In a recent stucly, 6 however, higher doses of IFN-gamma clid not improve clinical outeome as eompared to chemotherapy alone. The present study was also planned to test the clinical effieacy of maintenance therapy with IFN-gamma, beeause the effects of interfe­rons might be more pronounced in patients with low or even clinically undetectable tumour burclen. Because only few patients proceecled to maintenance therapy, no conclusions with regard to the efficacy of IFN-gamma during maintenance therapy can be clrawn from our stucly. For this purpose, large study populations are requirecl. Acknowledgement R. Pirker and S. Zbchbauer havc been sup­ported by the "Austrian Science Foundation". References 1. Grilli R. Oxman AD, Julian JA. Chcmothcrapy for advanccd non-small-ccll lnng canccr: how much bcncfit is cnough? ./ Ciin Oncol 1993; II: 1866-72. 2. Baron S, Tyring SK, Fleischmann WR. Coppcnha­vcr DI-1, Nicscl DW, Klimpcl GR, Stanton J, Hughes TK. Thc intcrfcrons. Mcchanisms of action and clinical applications. JAMA 1991; 226: 1375-83. 3. Sklarin NT, Chahinian AP. Fcucr EJ, Lahman LA, Szrajcr L, 1-Iolland JF. Augmcntation of activity of cis-Diammincclichloroplatinum (11) and mitomycin C by interferon in human malignant mcsothclioma xcnografts in nudc micc. Cancer Res 1988; 48: 64-7. 4. Mountain CF. A ncw intcrnational staging systcm for Jung canccr. Chest 1986; 89: 225S-33S. 5. Kaplan EL. Mcicr P. Nonpararnctric cstimation from incomplctc obscrvations . .! Arn Stat Assoc 1958; 53: 457-81. 6. Halmc M, Maasilta PK. Pyrhiincn SO, Mattson KV. lntcrfcrons combincd with chcmothcrapy in thc trcatmcnt of stagc lil-IV non-small celi Jung canccr -a randomiscd study. Eur .! Cancer 1994; 30A: 11-5. Radio! Oncol 1994; 28: 398-402. Endobronchial chemotherapy in combined treatment for non-small celi lung cancer Algirdas Jackevicius, Saulius Cicenas, Eduards Aleknavicius Department of Thoracic Surgery and Racliology, Lithuanian Oncological Center, Vilnius, Lithuania From 1990-1993, 73 patients with non-small cell lung cancer (NSCLC) underwent combined treatment using endobronchial bleomycin ( BL) instillations. These patients were divided into three groups as follows: 21 patients of Group 1 unclerwent preoperative endobronchial chemotherapy (EBCH) with BL, ancl radical surgery; 40 patients of Group 2 hacl EBCH ancl conventional split-course racliotherapy (RT); 12 patients of Group 3 unclerwent EBCH and clynamic fractional racliotherapy (DFRT). ln Group (21 patients) Bleomycin was injectecl using a fiber-bronchoscope 2-3 limes before raclical surgery. A cumulative dose of BL was 60-120 mg. Thirteen lobectomies ancl 8 pulmonectomies were performecl. Afier surgery, tumors ancl lymph nocle (lin) specimens were morphologically evaluated in order to determine tumor and response to BL injections. In Group 2, 40 patients with locally aclvancecl NSCLC were treatecl by BL injections and conventional racliotherapy. Cumulative cloes of BL was 120-150 mg. We instillecl BL endobronchially 15 mg twice a week. The first course of racliotherapy (RT) was 40 Gy. During the treatment, tumor volume was followed by X-ray, CT-scanning ancl fiberbronchoscopy (proclucing endophotoimages). Jf the first course of treatment resultecl in partial tumor response (> 50 % tumor mass recluction), RT and EBC/-1 were continued up to 70 Gy and 200-21 O mg of B L. In cases when tumor response was < 50 %, or there was no regression (< 30 %), only RT up to 60-70 Gy was clelivered in claily fractions of 1.8-2.0 Gy. Group 3 (12 patients) receivecl RT in 4-2-2-2-4 Gy weekly regimens with BL appliecl on Days 1 ancl 5. So the claily close of 4 Gy was combined with a 15 mg injection of BL; the cumulative dose of BL was the same as in Group 2. Not alt of our patients tolerated this treatment. In Group 2, partial tumor response was achievecl in 32 cases, 3 patients hacl partial tumor regression, ancl in 5 patients the treatment provecl ineffective. In Groups 3, rapicl tumor response was achivecl in 8 cases, 2 patients hacl no response, while in another two patients the treatment hacl to be cliscontinuecl because of complications (one of them diecl). During EBCH, the following complications were notecl: fever in 16 patients, leukopenia in 2, and death due to bleecling after rapicl tumor necrosis in one patient. Key worcls: carcinoma, non-small celi Jung; combined modality therapy Correspondencc to: A. Jackcvicius, M.D., Depart­Introduction mcnt of Thoracic Surgery, Lithuanian Oncological Center, Santariskiu 1, 2600 Vilnius, Lithuania. The treatment for lung cancer is one of the UDC: 616.24-006.6-085 most difficult problems in oncology. Only 10­ Endobronchial chemotherapy in combined treatment far NSCLC 15 % of ali patients with lung cancer can be treated surgically. 1 It means that the majority of patients with lung cancer must be treated by other methods such as radiotherapy and che­motherapy. There are some reports on endo­bronchial chemotherapy of patients with Jung cancer. At the 4th Congress of Bronchologists Seiguchi N2 reported on bronchial chemothe­rapy for non-small lung cancer. The author claims having injected bleomycin and immuno­modulator OK-132 through a bronchoscope. The patients also received irradiation. The re­sults of this treatment were good: tumor re­sponse was achieved in 23/53 cases. No severe treatment-related complications were observed. Henry P, and Pierson Jr3 injected mitomycin C endobronchially to patients with carcinoid. Thirty-four patients underwent surgical treat­ment; tumor response was achieved in 75 % . Other authors4 tried to treat their patients with laser, but the treatment entailed severe compli­cations such as tumor hemorrhage. Therefore, less aggresive methods of treatment were sugge­sted. Photochemotherapy was used in patients with central Jung cancer.5 This treatment moda­lity was used in patients with advanced lung cancer. Tumor response was ahcieved in 20 cases. In recent years, there has been a ten­dency to use photodynamic therapy for central lung cancer.5 Initial endobronchial injection of bleomycin into the tumor was followed by laser irradiation. According to the authors, the re­sults of treatment were good. The majority of authors have published their immediate treat­ment results, whereas on information is availa­ble on the follow-up results of such treatment. Based on these reports, the indications and contraindications for bronchial chemotherapy remain doubtful. Hopefully, our investigations will help to provide some explanation as to the usefulness of enclobronchial chemotherapy. Material and methods Patients From 1990-1993, 73 patients with NSCLC were treated by a combination of endobronchial che­motherapy (EBCH), B L surgery, conventional racliotherapy (RT) ancl high-fractional radiothe­rapy (HFRT). The patients were diviclecl into three groups as follows: Group 1 consisted of 21 patients treated by EBCH and surgery, Group 2 of 40 patients treatecl by EBCH and RT, ancl Group 3 of 12 patients treated by EBCH and HFRT. The age of paticnts ranged between 34-72 yrs, mean 55 yrs. There were 5 females among them. The distribution of pa­tients by pTNM stage was as follows: Group L: 15 patients pT2NlM0, 5 patients -pT3N0­1M0, and l patient -pT3N2M0. In 18 cases the tumor was endobronchial and in 3 cases peripheral. In 8 cases we performed pulmonec­tomy, in 11 lobectomy, ancl in 2 bilobectomy. Three patients had adenocarcinoma and others epidermoicl cancers. The distribution of patients in Groups 2 and 3 is cvident from Figure 1. The majority of 25 21 ,A 20 1 ') ' 15 ,: /11 I 8 Squamous Ca Adeno Ca Figure 1. Distribution of paticnts in groups 2 and 3 accorcling to morphology and TNM. patients (26) had T3N0-1M0, 18 patients were treated in T2N0-1M0 stage, and 8 patients had T3N2MO stage of the disease. According to clinical type the patients were clistributed as follows: 45 patients had central tumors and 7 peripheral. In 45 cases the pulmonary tumor was epidermoid carcinoma, in 7 adenocarcino­ma. Group 4-controls: 70 patients unclerwcnt only surgical treatment; of these 45 were with T2NlM0, 12 with TlNlM0, 10 with T3N1M0 and 3 with T3N2MO. Ten patients had periphe­ Jackevicius A et a/. ral and 60 endobronchial tumors. Sixty-two patients had epidermoid carcinoma and 8 ade­nocarcinoma. Group 5-controls: 46 patients received 5-FU 0.5 g i.v. half an hour before conventional radiotherapy twice weekly, up to the tolerable dose (5-6 g of drug). Twenty-seven patients had T3N1MX, 5-T3N2M0, and 14 patients T2N1M0 stage. There were 7 peripheral adeno­carcinomas and 39 central epidermoid tumors. Enclobronchial chemotherapy Bronchoscopy was performed under local ane­sthesia. Tumor surface was washed with sodium chloride solution. Bleomycin was injected with a special needle into the tumor area 3-5 mm in depth in 3-4 points. It is necessary to be careful during the injection of bleomycin; it must not be injected into normal tissues. Bioptic samples from tumors were tak en during bronchoscopy. The procedure was completed if there were no signs of hemorrhage. We used a modified COX mathematic model for chemotherapy. Bleomy­cin injection dosage program was calculated to each patient.6· 7 We injected bleomycin using a BFB2 "Olympus" bronchoscope twice a week before radiotherapy. Cumulative dose of bleo­mycin ranged between 90-125-150 mg. Fractio­nation of radiation dosage was 1.8-2 Gy per week by conventional radiotherapy and 4-2-2-2-4 Gy per week by dynamic hyperfractio­nated radiotherapy. Radiotherapy was applied in the split-course mode. Tumor response was followed by endobronchoscopy and X-ray. Biopsies were taken before and after the study to evaluate the tumor mass difference. Results In Group 1, BL 15 mg was injected 2-3 times through a bronchoscope before surgery. Post­operative complications were similar in both groups, i.e. in patients treated with BL (Group 1) and in controls (Group 4). Bronchus stump leakage was observed only after right pulmo­nectomies (3 and 4 respectively in both groups). Morpohological investigations of the tumor and lymph node specimens show a decreasing mito­tic activity of tumor cells (18 cases), and consi­derable degenerative changes in the tumor after 3-5 endobronchial injections of BL. These changes are equivalent to grade IVa by Oh­boshi-Shimosato criteria. Hypertrophy of lymph follicles and strong interstitial reaction were detected in the resected lymph node specimens. The main effect of BL was coagulative necrois with a slight interstitial and submucosal reaction (16 cases). Partial tumour regression was noti­ced in 17 cases. In Group 2 (40 patients) partial tumor re­sponse (tumor mass reduced by more than 50 % ) was achieved in 32 cases; 3 patients had partial tumor regression (less than 50 % ), while 5 patients failed to respond to therapy (less than 30 % ). In Group 2 (EBCH + RT) better results were achieved when the tumor was situated endobronchially. Tumor was sensitive after 5-7 injections of BL. In other cases, to reach a partial tumor response (more than 50 % ) we had to use 60-70 Gy of RT and 120-200 mg of BL. The results of our 3-year follow-up show that the median survival rates of treated patients in Group l and cntrols by stage were as follows: T2N1M0 30 and 31 mos, T2N1M0 -26 and 27 mos, and T3N2M0 24 and 26 mos. The median survival in Group 2, and Group 3 in treated patients vs. controls, was as follows: T2N1M0 24 mos vs. 19 mos, T3NlM0 21 mos vs. 13 mos, and T3N2M0 -15 mos vs. 12.5 mos. These data show no significant difference in survival. Discussiou The use of Iocal chemotherapy for superficially situated malignant tumors such as melanoma and breast cancer has been reported effective. Along with the development of endoscopic sur­gery (laser, intraluminal radiotherapy) we have been trying to use local chemotherapy in the form of transbronchial intratumoral instillation in lung cancer patients. This method is compa­rable with other endobronchial treatment ap­proaches such as laser applications, intraluminal Endobronchial chemotherapy in combined treatment for NSCLC brnchytherapy, intratumoral ethanol injections, ncbulisation etc. By combining BL injcctions and radiotherapy, we try to reducc tumor vo­lumc and thus scnsitizc tumor cells to RT. There are a lot of other substances known to have the same effect. Using BL, we do not only sensitizc tumor cells to RT but also achicvc chemotherapeutical effect, knowing that epicler­moid carcinomas are scnsitivc to bleomycin. Palliation effect is achieved through tumor mass rcduction. Patients fell bctter after the clisappe­rance of atelectasis. In our cases, histopatholo­gical and cndoscopical findings confirmed great effcctiveness of this mcthod, whitch also proved its clinical value by rcducing such as obstructive pneumonia. Buton L. Speiser8 used after-load brachytherapy for local control of endobron­chial caracinoma. In thrce groups of patients treatecl with different closages, curative doses resultcd in 87 % obstucting improvcment and palliative doses in 84 % , where the group of patients with recurrence showed 70 % improve­mcnt. The author also reported the following treatment-relatecl complications: fatal bleedings in 7 % , bronchial stenosis and radiation bron­chitis in 11 % , pneumothorax, cardiac arrhyt­hmia and infection. Nevertheless, the mcthod proved to be an effective palliation treatment for lung cancer; 24 % of his patients were treated by laser destruction. Cox and Byhar reported about hyperfractionated radiation ancl pulmonary toxicity, which was high ancl clcpcn­ dccl on thc ficlds of HFRT. EBCH can be used as adjuvant thcrapy. Group 2 (40 patients) underwent EBCH and conventional RT, while Group 3 (12 patients) received DFRT. The comparison of trcatment results in both groups has shown that more rapicl tumor regression was achievecl in Group 3. In Group 2, better tumor response was obtained when the tumor was situatcd endo­bronchially, and whcn it was smooth, without signs of hemorrhage and superficial necrosis. In the cascs with T3N2M0 ancl T3NlM0 stages, when the tumor was solid ( cartilaginoid), endo­bronchial regression was clubious and more difficult to achieve. The patients in Group 3 showed rapid tumor necrosis, but one of them diecl due to fatal pulmonary bleedding. In com­parison with thcir controls (Group 5), the pa­tients in Group 2 presentcd with more rapid tumor regression. Howcver, the follow-up re­sults in Groups 2 and 5 show no significant difference in the survival of patients in both groups. Thc prcsented method is very simple, and thc fact that it does not require any addi­tional equipmcnt is certainly and advantage, considering our economic situation. The me­thod is not well known and can be associatecl with various problcms, but nevertheless, our rcsults show that it is comparablc with other cndobronchial mcthocls for tumor volume re­duction in opcrablc patients. Low rate of severe complications renders this treatment approach suitable for outpatient use. It should be combi­ned with other methods of lung cancer trcat­ment. In our opinion, endobronchial chemothc­rapy can bc used only as adjuvant therapy. There are two conditions for such treatment: l) the tumor must be situated in the bronchus, and 2) there should bc no evidence of hcmor­rhage in the tumor. Thc drug which is injectcd into the tumor arca can affect it. Thcre is a danger of rapicl tumor necrosis. Conclusions 1) endobronchial intratumoral chemotherapy with bleomycin is technically easy to perform, and should be used in combination with other methocls of treatment in patients with 11011­small-cell lung cancer. 2) Endobronchial chcmotherapy and radio­therapy should be used in patients with locally aclvanced tumors of the lung without clistant metastascs. The occurrence of trcatment-rela­ted complications is low. 3) This method of treatmcnt has proved inef­fective in patients with locally advanced pcri­bronchial tumors. 4) Immediatc results were better in the group of patients who reccived endobronchial chemo­therapy (bleomycin) and dynamic hyper­fractionatecl radiotherapy. 5) The comparison of three-year follow-up Jackevicius A et al. results showed no significant difference in the survival of treated patients and those in the control groups. References l. Jackcvicius A. Plaucil{ vežys (Lung cancer), Vil­nius: Mokslas, 1975. 2. Seikiguchi N. Endoscopic thcrapy in advansed cen­tral type lung cancer (12 pts). 4th World Congress of Bronchology. Rome 1984; 312-20. 3. Henry F, Pierson Jr. Fine necdle aspiration of carcinoid of the lung. Acta Cytologica 1986; 471-3. 4. Fusisawa T, Hangoo H, Yamaguchi Y. Intratumou­ral ethanol injection for malignant tracheobronchial lesions. Endoscopy 1982; 18: 188-91. 5. Haussingcr C, Huber RM. Photodynamic therapy of bronchial cancer. Pulmonologie 1990; 44 (3): 687-93. 6. Swans CW. Role of optimal theory in cancer che­motherapy. Math Biosci 1990; 101: 227-84. 7. Bertussi A and al. Mathematical moclels of the ce!l cycle with a wiew to tumor. Math Biosci 1981; 158-88. 8. Spencer BL. Remote after loading brachytherapy for the local central of endobronchial carcinoma. Radia/ion Oncology 1993; 25 (4): 579-87. Radio! Oncol 1994; 28: 403-7. Brain metastases from Jung cancer Carsten Nieder, Marcus Niewald, Ursula Nestie, Karin Walter, Klaus Schnabel Department of Radiotherapy, University Hospital of the Saarland, Homburg/Saar, Germany Our analysis included 150 patients who were treated by whole-brain irradiation between 1983 and 7993. All of them received JO x 3 Gy over 2 weeks. In 72 cases surgical resection of brain metastases preceded radiotherapy. 49 patients had primary small-cell lung cancer (SCLC), and 81 non-small cel! lung cancer (NSCLC). The others had mixed tumors. The evaluation of their CT-scans showed a significantly dzfj'erent average number of brain metastases (SCLC 4 met., NSCLC 2 met.) and a smaller surrounding edema in the case of SCLC. Dimne/er of metastases and total cerebral tumor vohune were distributed equally. The interval between primary tumor and development of brain metastases, presence ol extracerebral metastases and Karnoj.·ky pe,formance status were also distributed equally, but SCLC patients were significantly younger. The local remission rate was dependent on histology as well as on the diameter of metastases. in 44 % of ali cases a complete or partial remission was founcl (SCLC 60 %, adeno-ca 39 %, squamous-cell-ca 21 %, p 0.007). 1-Jowever, after multivariate analysis, histology was not fou11d to be a prognostic factor for survival. Surgical treatment also had 110 influence on the survival. Patients'age, Karnof.ky-pe,j'ormance status, extracerebral metastases and the diameter of brain metastases were the only prognostic factors. Mean survival was 123 days (median 69 days). The survival of patients with extracerebral metastases and a Karnof.ky-pe1formance status < 70 was very poor. There.f'ore, they should be treated only in the case of disabling symptoms, when application <.l corticosteroids has failed. Far the majority of patients radiotherapy with JO X 3 Gy seems to be appropriate. Key words: Jung neoplasms, lung cancer; brain neoplasms-secondary, brain metastases; radiotherapy Introduction Due to the generally high incidence of the disease and the high frequency of spread, lung cancer is the predominant primary tumor of patients presenting with brain metastases (Ega­wa, Order, 2 Turalba). 3 Corrcspondcncc to: Dr. med. Carstcn Nicdcr, Dcpart­mcnl of Radiothcrapy, Univcrsity Hospital, 66421 Homburg/Saar, Gcrmany. UDC: 616.831-006.6-033.2:6l6.24-006.6 Various autopsy series demonstrated that 17 -50 % of ali lung cancer patients developcd 5 brain metastases (Davis,4 Doyle,Newman6), dependent on the histological tumor type. This retrospective analysis of 150 irradiated cases is targeted to investigate the differences in pattern of spread and treatment results between brain metastases from small-cell lung cancer (SCLC) and non-small-cell lung cancer (NSCLC) and their influence on patients' prognosis. Nieder C et al. Material and methods The data of 150 consecutive patients with brain metastases from histologically proven Jung can­cer, treated between 1983 and 1993, have been analysecl retrospectively. Patients' characteristics Fourty-nine patients (33 % ) hacl SCLC, 34 pa­tients (23 % ) squamous-cell Jung cancer (SQCLC), 34 (23 % ) sufferecl from aclenocarci­noma of the Jung ancl 33 (22 % ) from other ancl mixed tumor types. 87 % of ali patients hacl signs of increasecl intracranial pressure ancl/or other symptoms clue to brain metastases; 13 % were asymptomatic. Table 1 ancl Table 2 show ali other pretreat­ment characteristics. Table 1. Patients' characteristics. Parameter Ali patients median age (years) 60 multiple brain metastases 61.5 % average number of brain met. 2.6 median diameter of brain met. (cm) 2.2 median total cerebral tumor volurne ( crn3) 8.7 extracerebral metastases 49.7% median Karnofsky-index 7 male : fcrnale ratio 9:1 Table 2. Patients' characteristics. Methods Brain metastases were diagnosecl by CT. In each case radionuclide bone scan abdominal ultrasound, chest x-ray and CT were performed for tumor staging. The primary tumor was confirmed histologically by biopsy (mediastino­scopy or bronchoscopy). Twelve patients receiv­ed radiotherapy after surgical rcsection of a solitary brain metastasis. Ali of them had pre­operative MRI. In 31 cases chemotherapy was administered during the course of the disease, but always before detection of brain metastases and never with curative intent. Ali patients received radiotherapy 5 days a week. Either Co-60-gamma-rays or photons of a linac were used. In ali cases a whole-brain-irradiation of 10 x 3 Gy over 2 weeks was given. The close specifications refcr to a reference point in the middle of the brain. The target volume was encloscd by the 80 %-isoclosc. Nincty-eight % of ali patients received dexamethasone through­out thc course of radiotherapy. The daily dose was < 10 mg in 15 % , 10 ­20 mg in 46 % and > 20 mg in 29 % , depending median interval ( days )* on thc symptoms and the clinical course. In 8.5 % the exact dose was unknown. On the last day of radiotherapy, further CT­scans of the brain and clinical examinations were performed, and later on repeated at 3­month intervals. Survival tirne was calculated from thc beginning of radiotherapy. Survival curves were obtainecl using the Kaplan-Meier­method. Local remission was evaluated accord­ing to the UICC criteria. Ali patients who failed to complete their planned treatment course were also included in the analysis. average number of brain met. surrounding 2.0 0.02 Results edema Ali different pretreatment characteristics of SCLC and NSCLC patients are presented in Tables 1 and 2. Twenty-three patients (15 % ) failed to completc their planned radiotherapy * (betwecn primary diagnosis and development of course due to progressive deterioration of their brain metastascs) SCLC = small-ccll lung cancer, general condition. Karnofsky performance sta­SQCLC = squamous-cell lung canccr tus before therapy was the only parameter Brain metastases _fi-om lung cancer which correlated with incomplete therapy (p < 0.001). In 129 cases, CT-scans were performecl by the encl of therapy. Ali 12 operatecl patients (9 % ) had their metastases completely removed and were without recurrence. The results of other cases were as follows: complete remission in 17 cases ( 13 % ) , partial remission in 40 cases (31 % ), no change in 50 cases (39 % ), progres­sion in 3 cases (2 '1/o ); in 7 cases the course of multiple metastases was different (5 % ). The remission rate was 60 % for SCLC, 39 .3 % for acleno-ca and 21.4 % for SQCLC, p = 0.007. The estimatecl probabilities and the 95 %­confidence intervals of complete ancl partial remission were 56 % ( 43-69 % ) for SCLC, 41 % (32-50 % ) for acleno-ca and 34 % (24-45 % ) for SQCLC. The remission rate was dependent on histology as well as on the diameter of metasta­ses (> 2.5 cm versus ::; 2.5 cm) (Cox proportio­nal hazarcls model). A symptomatic relief was obtained in 48 % of ali cases. Additionally, 13 % of patients were asymptomatic before irradiation and remained Time in months Figure l. Mcan Karnofsky-pcrforrnancc status (karn) and standard dcviation bcforc and aftcr radiothcrapy (n = 150). .9 . .8 > C.. . 7 ::, - "' .6 o .5 t' .4 •rl .o .3 0.05) . .: \ .8 > C.. . 7 ·. ::, "' -.6 1 ': o . .5 >­ ..., . •rl .o 0.05). (NC). The mean values were 294 clays (CR), 142 days (PR) ancl 103 days (NC), p 0.003 Nieder C et al. (CR vs NC), 0.04 (CR vs PR), > 0.05 (PR vs NC). In multivariate analysis (Cox proportional hazards model) Karnofsky performance status (p Qo, ··,.,. "!· 1 .,,.,'O<'I p· .,, Radio/ Oncol 1994; 28: 432. Instructions to authors The journal Radiology and Oncology publishes ori­ginal scientific papers, profcssional papers, review ar­ticles, case reports and varia (reviews, short communi­cations, profcssional information, ect.) pertinent to diagnostic and interventional radiology, computerised tomography, magnetic resonance, nuclear medicine, radiotherapy, clinical and experimental oncology, ra­diobiology, radiophysics and radiation protection. Submission of manuscript to Editorial Board implies that the paper has not been published or submitted for publication elsewhere: the authors are responsible for ali statements in their papers. Acccptcd articlcs bccomc thc property of thc journal and thcrcforc cannot bc publishcd clscwhcrc without writtcn pcrmis­sion from thc Editorial Board. Manuscripts writtcn in English shoukl be sent to thc Editorial Office: Radiology and Oncology, Institute of Oncology, Vrazov trg 4, 61000 Ljubljana, Slovenia; Phone: + 386611320 068, Fax: + 386611314 180. Radiology and Oncology will consider manuscripts prepared according to the Vancouver Agrcemcnt (N Engl J Med 19 91; 324: 42 4-8.; BMJ 19 91; 30 2: 6772.). Ali articles are subjccted to cditorial review and revicw by two independcnt rcfcrces sclcctcd by the Editorial Board. Manuscripts which do not comply with thc technical rcquiremcnts statcd herc will be rcturncd to thc authors for correction beforc thc rcvicw of the rcfcrccs. Rcjccted manuscripts are gcne­rally returncd to authors, howcvcr, thc journal cannot bc held rcsponsiblc for their loss. The Editorial Board rcscrves thc right to rcquirc from thc authors to make appropriate changcs in the content as well as gramma­tical and stylistic corrcctions whcn ncccssary. Thc cxpenscs of additional cditorial work and requcsts for rcprints will bc chargcd to the authors. General instructions: Type the manuscript doublc spa­ccd on one sidc with a 4cm margin at the top and left hand side of thc shect. Write thc papcr in grammati­cally and stylistically correct language. Avoid abbrevia­tions unlcss prcviously explained. Thc technical dala should confirm to thc SI system. Thc manuscript, including thc references may not cxcccd 15 typewritten pages, and the number of figurcs and tables is limited to 4. If appropriate, organise the text so that it includes: Introduction, Material and methods, Results and Discussion. Exccptionally, thc results and discus­sion can be combined in a singlc section. Start each section on a new pagc and numbcr thcsc consecutivcly with Arabic numerals. Authors are encouragcd to submit thcir contributions bcsides threc typewritten copies also on diskcttcs (5 1/4") in standard ASCII format. First page: name and family name of ali authors, a brief and specific title avoiding abbreviations and colloquialisms, -complete aclclrcss of institution for each author, -in thc abstract of not more than 200 worcls covcr the main factual points of thc article, and illustrate them with thc most relevant data, so that thc rcadcr may quiekly obtain a general view of the material. lntroduction is a bricf and concisc section stating thc purposc of thc articlc in rclation to other alrcady published papcrs on the same subjccts. Do not presen! extcnsive rcviews of thc literature. Material and methods should provide cnough informa­tion to cnablc experimcnts to bc repcated. Writc thc Results clearly and conciscly and avoid rcpcating the data in the tables and figures. Discussion should explain the results, and not simply repeat them, interpret thcir significance and clraw conclusions. Graphic material (figures and tables). Each itcm should be scnt in triplicate, one of them markcd original for publication. Only high-contrast glossy prints will be acceptecl. Line drawings, graphs and charts should bc done profcssionally in Indian ink. Ali lettering must be lcgiblc aftcr rcduction to column size. In photo­graphs mask the iclentities of patients. Labe] the figures in pencil on the back indicating author's name, thc first fcw words of thc title and figure number: inclicate the top with and arrow. Write legend to figures and illustrations on a scparate shcet of paper. Omit vertical lines in tables ancl write thc ncxt to tablcs ovcrhead. Labe! thc tablcs on their rcvcrse side. References should be taped in accorclancc with Van­couver style, clouble spaccd on a separatc shcct of' paper. N umbcr thc refcrences in the order in which they appear in the text and quote thcir corrcsponding numbers in the text. Following are some examples of rcfcrenccs from articles, books and book chapters: l. Dent RG, Cole P. In vitro maturation of monocy­tcs in squamous carcinoma of thc lung. Br J Cancer 1981; 43: 486-95. 2. Chapman S, Nakielny R. A guide to radiological procedures. London: Baillicre Tindall, 1986 . 3. Evans R, Alexander P. Mechanisms of cxtracellu­ lar killing of nucleatcd mammalian cells by macropha­ges. In: Nelson DS ecl. Immunobiology of macrophage. New York: Acaclemic Press, 1976: 45 -74. For reprint information in Norih America Contact: lntemarional reprint Co,poration 968 Admiral Callag­lwn Lane, # 268 P. O. Box 12004, Vallejo, CA 94590, Tei: / 707) 553 9230, Fax: (707) 552 9524. Systemic ali timicrobial agent for I\Oth oral and parenteral use orally, 2 x 200 to 400 mg iv,-chronic salmonel!a carriers 4 x 250 mg orally; acute gonorrhea a single dose oj 500 mg orally; in severe injections the oral dosecan be increased to3 x 750mg, and intravenous close to 3 x 400 mg daily; close is reduced in elderly patients and in severe renal dysjunction. Duration of therap y: pyelonephritis at least JO days, peritonilis (in combination with metronidazole) at least 14 days, osteomyelitis at least 6 weeks, other infeclions at least 3 days after disappearance oj clinical signs. Supply: tablets: 250 mg JO tabs; 500 mg 10 tabs; injection: 100 mg/JO ml 5 ampoules. Ali additional injormatlon can be obtainedfrom tile manufactu.rer. ..•KRK. SLOVENIA • 1!.!!1111!'7.,_ Reaches high concentrations in body fluids, tissues and intracellular space.