Klinicna studija / Clinical study
Zgodnje odkrivanje metastaz v bezgavkah v vratu pri bolnicah z rakom dojke: vloga ultrazvoka in igelne biopsije
Early detection of neck lymph node metastases in breast cancer patients: the role of ultrasound and fine needle
biopsy
Avtor / Author	Janez Rebol1,2 Peter Balon1 Peter Kokol3 Boštjan Lanišnik1 Iztok
Takač2,4 Andraž Dovnik4 Nina Fokter Dovnik5 Monika Sobočan2,4 Darja Arko24
Ustanova / Institute	"'Univerzitetni klinični center Maribor, Klinika za otorinolaringologijo, cervikalno in
maksilofacialno kirurgijo, Maribor, Slovenija; 2Univerza v Mariboru, Medicinska fakulteta, Maribor, Slovenija; 3Univerza v Mariboru, Fakulteta za elektrotehniko, računalništvo in informatiko, Maribor, Slovenija; 4 Univerzitetni klinični center Maribor , Klinika za ginekologijo in perinatologijo, Maribor, Slovenija; 5 Univerzitetni klinični center Maribor, Oddelek za onkologijo, Maribor, Slovenija;
1 University Medical Center Maribor, Clinic for Otorhinolaryngology & Maxillofacial Surgery, Maribor, Slovenia; 2 University of Maribor, Faculty of Medicine, Maribor, Slovenia; 3 University of Maribor, Faculty of Electrical Engineering and Computer Science, Maribor, Slovenia; 4 University Medical Center Maribor, Clinic for Gynecology and Perinatology, Maribor, Slovenia; 5 University Medical Center Maribor, Department of Oncology, Maribor, Slovenia;
Ključne besede:
rak dojke, metastaze, metastaze v vratu, dejavniki tveganja
Key words:
breast cancer, metastases, neck metastasis, risk factors
Članek prispel / Received
8. 2. 2018
Članek sprejet / Accepted
13. 2. 2019
Naslov za dopisovanje / Correspondence
Monika Sobocan, dr. med., Univerzitetni klinični center Maribor, Ljubljanska ul. 5, 2000 Maribor E-pošta: monika.sobocan@gmail.com
Izvleček
Namen: Namen raziskave je bil pridobiti razumevanje porazdelitve odkritih metastaz v vratu pri bolnicah z rakom dojk.
Metode: Skozi retrospektivno analizo smo identificirali 41 bolnic z rakom dojke ter analizirali klinično-patološke značilnosti bolnic in značilnosti potrjenih metastaz v vratu. Pregledali smo podatke bolnic, obravnavanih v naši ustanovi med leti 2010 do 2016, povzete iz medicinske dokumentacije. Diagnozo metastaz v vratu smo potrdili s tankoigelno biopsijo. Glede na prizade-
Abstract
Purpose: To analyze the distribution and pattern of neck metastasis in breast cancer (BC) patients. Methods: The clinicopathological characteristics of 41 patients with BC and confirmed neck metastasis (NM) treated at our institution between 2010 and 2016 were retrospectively analyzed by reviewing medical records. The diagnosis of NM was confirmed by fine needle aspiration cytology, and node localization was classified into six levels according to Robbins' classification. Statistical analysis was performed using
acta medico-biotechnica i2?
2019; 12 (2): 21-28
Klinicna studija / Clinical study
tost bezgavk so bile metastaze klasificirane v 6 ravni glede na Robbinsonovo klasifikacijo. Statistično analizo smo opravili s pomočjo programa SPSS. Za analizo smo uporabili deskriptivno statistiko, Pearsonovo korelacijo in multidimenzionalno korespondenčno statistiko.
Rezultati: S pomočjo ultrazvoka vratu smo metastaze odkrili pri 34 bolnicah (85 %). 38 % bolnic je bilo asimptomat-skih. V 17 % ni bilo tipnih nobenih bezgavk. Povprečni čas od diagnoze do razvoja metastaz je bil 5,8 let (razpon od 0 do 26 let). 53,7 % metastaz je bilo ugotovljeno na ravni V; IV. stopnja je bila vključena v 19,5 %. 9,8 % metastaz je bilo prisotnih na obeh ravneh. V 75 % so bile metastaze prisotne samo na eni ravni. Pomembna korelacija (p < 0,01) je bila prisotna med: i) starostjo bolnice in časom od diagnoze do razvoja metastaz v vratu, ii) starostjo in stopnjo bolezni ter iii) stopnjo in videzom konglomeratov.
Zaključek: Metastaze v vratu se pri bolnicah z rakom dojke običajno kažejo poleg položaja v supraklavikularni loži, pa tudi na drugih ravneh. Zaznavanje zgodnjih okultnih meta-statskih bezgavk je mogoče z ultrazvočnim pregledom v kombinaciji s citološko punkcijo.
descriptive statistics, Pearson's correlation and multidimensional correspondence analysis. All statistical analyses were performed using SPSS.
Results: At the time of neck ultrasonography, 34 (85%) patients had developed distant metastases, 38% of the patients were asymptomatic, and 17% had no palpable nodes. The average time from BC diagnosis to the development of NM was 5.8 years (range, 0-26 years). Lymph node metastasis occurred at level V in 53.7% of patients and level IV in 19.5% of patients. Two-level involvement occurred in 9.8% and one-level involvement in 75% of patients. There were significant correlations (p < 0.01) between patient age and time from diagnosis to development of NM, age and stage of the disease, and stage and appearance of conglomerates. Conclusion: The most common site of NM from BC was next to the supraclavicular fossa, although they occurred at other levels. Ultrasonography combined with fine needle aspiration cytology is effective for the detection of early occult me-tastatic lymph nodes.
INTRODUCTION
Breast cancer (BC) is the most common cancer in women worldwide. In Europe the average annual mortality is between 13 and 19/100,000 women (1,2). In Slovenia, the annual mortality rate per year is 17/100,000 women (3). Mortality rates have improved mainly because of advances in therapy rather than early detection (4). The prolonged life expectancy of women with BC is associated with a greater risk of distant metastasis. The most common sites of distant metastasis are the spine and pelvis (77%), liver (39%), and central nervous system (5).
The occurrence of cervical lymph node metastasis in patients with BC is associated with poor prognosis. A study assessing the development of neck metastasis (NM) in the supraclavicular region in 2,658 patients (6) reported that only 4.3% of the patients developed metastasis. At that time in 2001, the factors associated with a higher probability of developing NM were age <40 years, tumor size >3 cm, high histological grade,
lymphovascular invasion, negative estrogen receptor status, synthetic phase fraction >4%, >4 positive axillary nodes, and involvement of axillary nodes at levels II and III.
Infraclavicular primaries commonly metastasize to the supraclavicular region and not to the neck (7). BC metastasis to the ipsilateral supraclavicular fossa is classified as loco-regional metastasis (N3c). Lymph nodes above the supraclavicular region are not mentioned, although they should be considered as distant metastases (7).
A study of NM in BC patients in the United Kingdom (UK) in 2012 showed considerable variation in the approach towards staging as well as treatment (8). The aim of this study was to analyze the distribution and characteristics of NM in patients with BC and explore the value of ultrasonography and fine needle aspiration cytology for the early detection of occult nodes.
2l\ ACTA medico-biotechnica
2019; 12 (2): 21-28
Klinicna studija / Clinical study
METHODS
Patients with BC referred to the Department of Otorhinolaryngology & Maxillofacial Surgery of the University Medical Center Maribor between 2010 and 2016 were scanned for the presence of NM. Patients were referred to our department because of the presence of suspicious nodes on the neck or high levels of the breast carcinoma biomarker cancer antigen 15-3. The study included patients with a confirmed diagnosis of NM by fine needle biopsy. Patients referred to the Department of Otorhinolaryngology & Maxillofacial Surgery were evaluated using ultrasonography, which is a commonly used first-line screening method (9). Partitioning the neck into six levels is a widely used method for determining the exact location of metastasis during head and neck surgery (10), and this technique was used in the present study.
Patients and materials
Over a 7-year period, 1,349 patients with primary BC were treated at our institution. Approximately 3% of all BC patients treated at our institution at a given time were diagnosed with NM during follow-up. Patients were referred for head and neck examinations after undergoing screening at the Center for Diseases of the Breast. The examination was either a regular follow-up examination or an unscheduled examination in patients with deteriorating disease conditions. Ultrasonography and fine needle biopsy were performed in all patients. BC metastases were confirmed by cytological analysis. The Voluson 730 ultrasound system (GE Healthcare; Boston, MA, USA) was used with a 5-10 MHz linear transducer. The location of metastases in the neck was classified according to Robbins (10) as follows: level I, submental and submandibular nodes; level II, upper jugular group of nodes; level III, middle jugular nodes; level IV, caudal jugular nodes; level V, nodes of the posterior neck triangle; and level VI, anterior neck compartment.
The distribution, size, and number of metastatic lymph nodes, the presence of conglomerates, and the presence of contralateral lymph node metastasis were evaluated. These features were determined by
ultrasonography.
Data analysis was performed using the SPSS statistical package version 20.0 (IBM, Armonk, NY, USA). Descriptive analysis and Pearson's correlation were used, and a p-value <0.05 was considered statistically significant. Multidimensional correspondence analysis (MCA) was used to assess the associations between node dimensions and number, the presence of conglomerates, the occurrence of collateral lymph node metastases, the distribution of nodes and stage. MCA is a multivariate statistical and visualization technique that is extremely useful for analyzing multiway contingency tables in cases showing correspondence between columns and rows. It is conceptually similar to principal component analysis, although it is used for categorical rather than continuous data. The MCA results are presented as "bi - plots." The proximity of items in a bi-plot represents the degree of association between them. Items positioned near each other show a greater association than items further away. Each bi-plot is characterized by two dimensions that can be identified during the analysis. The study was approved by the Institutional Review Board (approval No. UKC-MB-KME-46/14).
RESULTS
Patient characteristics
The study included 41 patients with cytologically confirmed NM. The average age of BC patients at the time of NM diagnosis was 54.9 years (range, 38-81 years). The primary BC was invasive ductal carcinoma in 29 (70.7%) patients, invasive lobular carcinoma in 2 (4.9%), invasive ductal and lobular carcinoma in 3 (7.3%), invasive ductal and inflammatory carcinoma in 1 (2.4%), and invasive micropapillary carcinoma in 1 (2.4%) patient. Five (12.2%) patients had no exact histologic diagnosis and were diagnosed by cytological evaluation at an advanced disease stage. In these cases, histological samples were not obtained by surgery. The primary treatment in 31 (75.6%) patients was surgery combined with postoperative chemotherapy,
acta medico-biotechnica i2?
2019; 12 (2): 21-28
Klinicna studija / Clinical study
radiotherapy, and/or hormonal therapy according to established guidelines. Three (7.3%) patients underwent neoadjuvant chemotherapy followed by surgery. Among the surgically treated patients, mastectomy was performed in 22 and breast conserving surgery in 11 women. Axillary lymph node dissection was performed in 26 patients and sentinel node biopsy in two. Four patients (9.8%) were treated with chemotherapy only and two (4.9%) with hormone therapy only. Exact information on primary treatment could not be obtained for one patient who was treated in a different institution.
Metastasis evaluation
At the time of the otorhinolaryngological examination, 34 of 41 patients (85%) had developed distant metastases. Of these patients (n = 34), 29 developed metastasis at multiple sites. The locations of the distant metastases were as follows: bones (73.5%), liver (35.3%), lung (32.4%), pleura (17.6%), central nervous system (14.7%), skin (8.8%), mediastinum (8.8%), and peritoneum (2.9%).
BC neck dissemination
Pathological lymph nodes were characterized as solitary or conglomerate-forming according to ultrasonography. Solitary pathological lymph nodes were round as determined by sonography, hypoechoic, without clear boundaries from the surrounding tissue, and with pathologic peripheral or central vascularization. The average maximal diameter of solitary pathological lymph nodes was 16.2 mm (range, 3-28 mm). The average time from the diagnosis of BC to the diagnosis of NM was 5.8 years (range, 0-26 years). A proportion of 38% of patients were asymptomatic. No nodes were palpable in 17% of patients.
Neck node characteristics
In 43.9% of cases, metastatic nodes were smaller than 10 mm, and 26.8% were larger than 20 mm (Table 1). Conglomerates of metastatic nodes were present in 11 patients (26.8%). One solitary metastatic node was present in 19.5% of patients; two nodes were observed in 26.8%; three in 12.2%; four in 12.5%, and five or more in 29.3% of the patients. The majority of metastatic nodes (53.7%) were in level V, which included supraclavicular metastases; 19.5% of patients had metastatic nodes in level IV and 9.8% had metastases in two levels. One patient had metastasis in level I, which is within the submandibular region (Table 1). Metastases in only one level were detected in 75.6% of the patients, and in two levels in 9.8% of
Table 1: Patient neck node metastasis characteristics
Size of metastatic nodes in breast cancer patients		
Size	No.	%
< 1 cm	18	43.9
1-2 cm	12	29.3
> 2 cm	11	26.8
Metastatically involved neck levels in breast cancer patients		
Level	No. of patients	%
I	1	2.4
IV	8	19.5
V	22	53.7
IV+V	4	9.8
3 levels	2	4.9
4 levels	4	9.8
Number of metastatically involved neck levels in patients with breast cancer		
Number of levels	No. of patients	%
1	31	75.6
2	4	9.8
3	2	4.9
4	4	9.8
24 acta medico-biotechnica
2019; 12 (2): 21-28
Klinicna studija / Clinical study
patients (Table 1). In 12.2% of the patients, metastases were detected on the contralateral side of the neck.
Correlation between patient characteristics and risk foi NM
There was a moderate positive correlation between patient age and onset of metastasis (r = 0.345, p < 0.05). There was a moderate negative correlation between patient age and disease stage (r = -0.358, p < 0.01) and a significant negative correlation between onset of the NM and disease stage (r = -0.538, p < 0.01). There was a small but statistically significant correlation between stage and the appearance of conglomerates (r = 0.037, p < 0.01) and a negative correlation between age and contralateral node involvement (r = -0.108, p < 0.05). Time from first BC diagnosis to NM diagnosis and involvement of contralateral nodes was also statistically significant (r= -0.194, p < 0.01) (Table 2). All other variables were not significantly correlated.
Multidimensional analysis and effect of staging on risk for NM
The associations between node dimensions and number, the presence of conglomerates, and the occurrence of contralateral lymph node metastasis
pC
« *
ji"
V
#
Figure 1. Bl-plot of associations between the dimensions of the nodes (size), their number (number of nodes), presence of conglomerates (conglomerate), appearance of contralateral lymph node metastases (contralateral nodes), the distribution of nodes distribution) and Stage (IA=1, IIA=2, IIB=3, IIIA=4, IIIC=5).
Table 2: Pearson's correlation coefficients (r) between variables in patients with breast cancer (N = 41). *p < 0.05, **p < 0.01. Age - age of the patient at the time of primary disease; Time - time from first treatment to the development of neck metastases; Stage - stage at the first presentation (TNM); No. levels - number of metastatically involved neck levels; Level - certain level in the neck; Size - size of the lymph node; Congl - presence of conglomerates; Contra - involvement of contralateral lymph nodes; No. nodes - number of metastatic lymph nodes in the neck.
	Age	Time	Stage	No. levels	Level	Size	Congl	Conba	No. nodes
Age	1	.345*	-.358**	.222	.204	.162	-.030*	-.108*	.043
Time	.345*	1	-.538**	-.43	.077	.037*	-.110	-.194**	-.178
Stage	-.358**	-.538**	1	.202	-.240	-.107	-.037**	.196	-.152
were analyzed using MCA (optimal scaling). The results of the analysis are shown in a bi-plot in Figure 1.
The results showed that primary disease stage had an effect on patients with the following characteristics:
in patients diagnosed with stage IA BC, the presence of contralateral nodes and node size > 2 cm in regions IV and V were associated with the risk of NM. In stage IIA, NM was strongly associated with the presence of conglomerates. In patients with stage IIB disease, a minimum of two nodes and region V involvement
acta medico-biotechnica i2?
2019; 12 (2): 21-28
Klinicna studija / Clinical study
were associated with an increased risk of NM. Stage IIIA was associated with the presence of NM in the absence of conglomerates, but in patients with nodes 1-2 cm and one or two nodes in region IV or V. Stage IIIB was strongly associated with NM in patients with more than four nodes involving regions IV and V.
DISCUSSION
Tumor cell dissemination occurs shortly after primary tumor vascularization (11). The ability of cells to lie dormant explains why metastases appear many years after the treatment of the primary tumor (12). The incidence of NM of 3% in the present study is comparable to that previously reported at 4% (13). BC can metastasize to different head and neck regions, although the most common site is the supraclavicular fossa (14-16). A recent study (17) explored patterns of metastasis to the head and neck region and demonstrated extreme variability in the clinical and pathological features of BC metastasis. In the present study, the average time from primary diagnosis to the detection of NM was 5.8 years (range, 0-26 years). A study by Gondim et al. (17) reported a period of 10.9 years (range 1-33 years) for the detection of metastasis. This difference can be attributed to differences in methodology, as the present study focused exclusively on NM, whereas the study by Gondim assessed metastasis to both the head and neck regions.
In 44% of patients with BC, metastatic lymph nodes were smaller than 10 mm. This suggests that cases suspicious for metastatic disease need to be evaluated carefully to identify small sonographically detectable nodes. Sonography needs to be considered in patients with BC who are at a high risk for developing NM because metastatic lymph nodes are not always detected by palpation alone. In the present study, palpation failed to detect metastatic nodes in 17% of cases. In patients with BC at a high risk of metastasis, even a 4-mm node needs to be considered suspicious, and ultrasound-guided fine needle aspiration cytology should be performed. Ultrasound alone is only up to 70% accurate in finding occult metastases, and its
2ó\ acta medico-biotechnica
2019; 12 (2): 21-28
accuracy increases to 89% when combined with fine needle aspiration cytology (18). Obtaining sufficient material for aspiration cytology is frequently difficult in small nodes. In these cases, the procedure should be repeated unless the lymph node is negative or in cases in which there is no clinical benefit in obtaining an exact diagnosis.
NM are often thought to arise within the supraclavicular fossa only (1,6). This was not the case in the patients included in this study. Although level V was the most commonly involved region, metastatic nodes were detected in level IV, as well as in two or more levels in a comparatively large number of patients. One patient had metastases in level I. In almost one third of the patients, metastatic nodes appeared in conglomerates, which is a typical sign of malignancy (1,6). However, metastases were solitary in 20% of patients, and 46% of patients had only one or two metastatic nodes. The onset of metastatic disease in the neck was correlated with patient age and disease stage. Higher stages were associated with metastases in regions IV and V. Elderly patients had fewer conglomerates and metastases in the contralateral side of the neck. This is consistent with research findings that distant metastasis (regardless of type) is less frequent in older patients than in young patients (19).
The proximity of stages IIB and IIIA in the bi-plot reflected their similarity (both showed region V involvement and were characterized by the presence of two nodes, and weak associations with the absence of contralateral nodes and region I were detected in both stages). The proximity and similarity between stages IA and IIIB was surprising. Both stages were associated with region V involvement; however, stage IIIB was also strongly associated with the presence of more than four nodes.
By the time they underwent ultrasound examination, 85% of the patients had already developed distant metastases. NM development was associated with the occurrence of other distant metastases. In these patients, early detection of asymptomatic NM is probably not clinically meaningful. However, NM may be the primary site of disease recurrence in some patients, and early diagnosis and treatment may be of value in selected
Klinicna studija / Clinical study
patients in this group.
In the 7th edition of the TNM classification, NM are mentioned as N3c in the ipsilateral supraclavicular fossa (20). Our study showed that in addition to the supraclavicular fossa (part of neck level V), metastases also appeared in other levels (19.5% in neck level IV). These are currently staged as distant metastases in the TNM classification.
STUDY LIMITATIONS
This study had methodological limitations such as its retrospective nature and the restriction of available data to that collected during active patient follow-up. In addition, the patient group analyzed (n = 41) was small, although the number is consistent with the general incidence of NM in BC. To the best of our knowledge, this is the largest analysis of NM in BC patients currently available.
CONCLUSIONS
Metastasis of BC to the neck may involve other regions in addition to the supraclavicular fossa. Nodes can be very small at initial presentation and were solitary in 20% of the patients. Neck palpation should be combined with ultrasonography of the neck and fine needle aspiration cytology in patients at high risk of NM to improve the early detection of occult nodes. Author contributions: JR, BL, and PB evaluated data from the ENT point of view, performed critical data analysis, and critically revised the manuscript draft. PK performed data analysis and interpreted data in terms of the current knowledge. DA, AD, NFD, MS, and IT collected available gynecological and oncological patient data and evaluated the data in terms of the current body of knowledge. All authors participated in the drafting and critical manuscript evaluation.
REFERENCES
1.	Bosetti C, Bertuccio P, Malvezzi M, Levi F, Chat-enoud L, Negri E et al. Cancer mortality in Europe, 2005-2009, and an overview of trends since 1980. Ann Oncol. 2013;24(10):2657-71.
2.	DeSantis C, Ma J, Bryan L, Jemal A. Breast cancer statistics, 2013. CA Cancer J Clin. 2014;64(1):52-62.
3.	Dimitrova N, Znaor A, Agius D, Eser S, Sekerija M, Ryzhov A et al. Breast cancer in South-Eastern European countries since 2000: Rising incidence and decreasing mortality at young and middle ages. Eur J Cancer. 2017;83:43-55.
4.	Becker N. [Development of the incidence and mortality of breast cancer]. Radiologe. 2001 Apr;41(4):337-43.
5.	Bunting JS, Hemsted EH, Kremer JK. The pattern of spread and survival in 596 cases of breast cancer related to clinical staging and histological grade. Clin Radiol. 1976 Jan;27(1):9-15.
6.	Chen SC, Chen MF, Hwang TL, Chao TC, Lo YF, Hsueh S et al. Prediction of supraclavicular lymph
node metastasis in breast carcinoma. Int J Radiat Oncol Biol Phys. 2002 1;52(3):614-9.
7.	Edge SB, Compton CC. The American Joint Committee on Cancer: the 7th Edition of the AJCC Cancer Staging Manual and the Future of TNM. Ann Surg Oncol. 2010 24;17(6):1471-4.
8.	Bisase B, Kerawala C. Survey of UK practice for management of breast cancer metastases to the neck. Ann R Coll Surg Engl. 2012;94(7):484-9.
9.	Šantl Letonja M, Letonja M, Nikolajevic Starčevic J, Petrovič D. Ultrazvočni in klasični dejavniki tveganja karotidne ateroskleroze pri bolnikih s sladkorno boleznijo tipa 2. Acta Med Biotech. 2013;6(1):22-44.
10.	Robbins K, Denys D. The committee for neck dissection classification. American Head and Neck Society's revised classification for neck dissection 2000. In: Johnson J, Shaha A, editors. Proceeings of the Fifth International Conference in Head and Neck Cancer. Omnipress Madison; 2000. p. 36571.
acta medico-biotechnica i2?
2019; 12 (2): 21-28
Klinicna studija / Clinical study
11.	Sherwood LM, Parris EE, Folkman J. Tumor angio-genesis: Therapeutic implications. N Engl J Med. 1971 18;285(21):1182-6.
12.	Sesterhenn AM, Albert US, Barth PJ, Wagner U, Werner JA. The status of neck node metastases in breast cancer-loco-regional or distant? The Breast. 2006;15(2):181-6.
13.	Sproson EL, Herd MK, Spedding A V., Brennan PA, Puxeddu R. Treatment of breast adenocarcinoma metastasis to the neck: Dedifferentiation of the tumor as suggested by hormone markers. Head Neck. 2012;34(8):1095-9.
14.	Al-Benna S, Tzakas E. Submental lymph node metastasis from invasive ductal breast carcinoma. Arch Gynecol Obstet. 2012, 9;285(4):1153-6.
15.	Schuler PJ, Heikaus S, Friebe-Hoffmann U, Hoffmann TK, Greve J, Klenzner T et al. Mammakar-zinommetastasen im Kopf-Hals-Bereich. HNO. 2010, 3;58(8):859-65.
16.	Pata YS, Ünal M, Gülhan S. Internal jugular vein thrombosis due to distant malignancies: two case
reports and literature review. J Laryngol Otol. 2008,14;122(3).
17.	Gondim DD, Chernock R, El-Mofty S, Lewis JS. The great mimicker: Metastatic breast carcinoma to the head and neck with emphasis on unusual clinical and pathologic features. Head Neck Pathol. 2016;11(3):306-13.
18.	van den Brekel MW, Castelijns JA, Stel H V, Luth WJ, Valk J, van der Waal I et al. Occult metastatic neck disease: detection with US and US-guided fine-needle aspiration cytology. Radiology. 1991;180(2):457-61.
19.	Purushotham A, Shamil E, Cariati M, Agbaje O, Muhidin A, Gillett C et al. Age at diagnosis and distant metastasis in breast cancer - A surprising inverse relationship. Eur J Cancer. 2014;50(10):1697-705.
20.	Sobin LH, Gospodarowicz MK, Wittekind C. TNM Classification of Malignant Tumours. 7th ed. Wi-ley-Blackwell; 2009.
acta medico-biotechnica i2?
2019; 12 (2): 21-28