Radiol Oncol 2024; 58(3): 366-375. doi: 10.2478/raon-2024-0044
366
research article 
Predictive potential of dynamic contrast-
enhanced MRI and plasma-derived angiogenic 
factors for response to concurrent 
chemoradiotherapy in human papillomavirus-
negative oropharyngeal cancer
Alja Longo1,2, Petra Hudler 3, Primoz Strojan2,4, Gaber Plavc2,4, Lan Umek5,  
Katarina Surlan Popovic1,2
1 Institute of Radiology, University Medical Centre Ljubljana, Ljubljana, Slovenia
2 Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia
3 Institute of Biochemistry and Molecular Genetics, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia
4 Institute of Oncology, Ljubljana, Slovenia
5 Faculty of Public Administration, University of Ljubljana, Ljubljana, Slovenia
Radiol Oncol 2024; 58(3): 366-375.
Received 24 May 2024 
Accepted 22 July 2024
Correspondence to: Alja Longo, M.D., Institute of Radiology, University Medical Centre Ljubljana, Zaloška cesta 2, SI-1000 Ljubljana, 
Slovenia. E-mail: alja.longo@kclj.si 
Disclosure: No potential conflicts of interest were disclosed. 
This is an open access article distributed under the terms of the CC-BY license (https://creativecommons.org/licenses/by/4.0/).
Background. Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) can assess tumour vascularity, 
which depends on the process of angiogenesis and affects tumour response to treatment. Our study explored the 
associations between DCE-MRI parameters and the expression of plasma angiogenic factors in human papilloma 
virus (HPV)-negative oropharyngeal cancer, as well as their predictive value for response to concurrent chemoradio-
therapy (cCRT).
Patients and methods. Twenty-five patients with locally advanced HPV-negative oropharyngeal carcinoma were 
prospectively enrolled in the study. DCE-MRI and blood plasma sampling were conducted before cCRT, after receiv-
ing a radiation dose of 20 Gy, and after the completion of cCRT. Perfusion parameters ktrans, kep, Ve, initial area under 
the curve (iAUC) and plasma expression levels of angiogenic factors (vascular endothelial growth factor [VEGF], 
connective tissue growth factor [CTGF], platelet-derived growth factor [PDGF]-AB, angiogenin [ANG], endostatin 
[END] and thrombospondin-1 [THBS1]) were measured at each time-point. Patients were stratified into responders and 
non-responders based on clinical evaluation. Differences and correlations between measures were used to generate 
prognostic models for response prediction.
Results. Higher perfusion parameter ktrans and higher plasma VEGF levels successfully discriminated responders from 
non-responders across all measured time-points, whereas higher iAUC and higher plasma PDGF-AB levels were also 
discriminative at selected time points. Using early intra-treatment measurements of ktrans and VEGF, a predictive model 
was created with cut-off values of 0.259 min-1 for ktrans and 62.5 pg/mL for plasma VEGF.
Conclusions. Early intra-treatment DCE-MRI parameter ktrans and plasma VEGF levels may be valuable early predic-
tors of response to cCRT in HPV-negative oropharyngeal cancer.
Key words: DCE-MRI; angiogenesis; VEGF; concurrent chemoradiotherapy; response; head and neck squamous cell 
carcinoma
Radiol Oncol 2024; 58(3): 366-375.
Longo A et al. / DCE-MRI and angiogenesis measures for response prediction in HPV-negative HNSCC 367
Introduction
Head and neck cancers are the seventh most com-
mon cancer worldwide, contributing to about 4.5% 
of global cancer deaths.1 The most recent (8th) edi-
tion of the Union for International Cancer Control 
(UICC) staging system distinguishes human pap-
illoma virus (HPV) positive and HPV-negative 
head and neck squamous cell carcinoma (HNSCC) 
as distinct entities, with the latter typically asso-
ciated with poorer prognoses.2 In recent decades, 
non-surgical treatment strategies for locally ad-
vanced HPV-negative HNSCC have remained 
largely unchanged, relying primarily on concur-
rent chemoradiotherapy (cCRT). The identification 
of reliable biomarkers for predicting responses to 
cCRT could aid in stratifying patients for potential 
individualised treatment adjustments. 
Vascular-rich HNSCCs are thought to be more 
responsive to treatment and show greater sen-
sitivity to radiation compared to less vascular 
HNSCCs due to enhanced delivery of chemothera-
peutic agents and improved oxygenation status.3 
Vascular characteristics of HNSCC can be inves-
tigated using functional MR imaging, specifically 
dynamic contrast-enhanced MRI (DCE-MRI). In 
recent years, both pre-treatment and early intra-
treatment functional MR imaging have shown 
promising results in predicting cCRT outcomes for 
HNSCC.4,5 Perfusion parameter ktrans, a measure of 
capillary permeability, is particularly promising 
for its potential clinical application in predicting 
survival and could be used to guide treatment de-
cisions.5 
Tumour vascularity depends on angiogenesis, 
a process regulated by various angiogenic factors 
and cytokines within the tumour microenviron-
ment. Key players in this regulation include vas-
cular endothelial growth factors (VEGFs), which 
are potent inducers of angiogenesis and vascular 
permeability. In addition, several other angiogenic 
factors have been implicated in promoting angio-
genesis in HNSCC, including connective tissue 
growth factor (CTGF), platelet-derived growth 
factor (PDGF) and angiogenin (ANG), whereas 
endostatin (END) and thrombospondin-1 (THBS1) 
have been investigated as endogenous angiogen-
esis inhibitors.6–9
The objective of this study was to investigate 
differences in DCE-MRI perfusion parameters and 
the levels of specific plasma angiogenic factors in 
HPV-negative oropharyngeal squamous cell carci-
noma (OPSCC) patients, distinguishing between 
responders and non-responders to cCRT. Our goal 
was to identify measures that could improve pa-
tient stratification and enhance the accuracy of 
outcome prediction.
Patients and methods 
This was a single-centre prospective study con-
ducted at the Institute of Radiology, University 
Medical Centre Ljubljana, in collaboration with 
the Institute of Oncology Ljubljana and the 
Institute of Biochemistry and Molecular Genetics, 
Faculty of Medicine, University of Ljubljana. The 
study was approved by the National Medical 
Ethics Committee of the Republic of Slovenia (No. 
0120-247/2019/4, 12 June 2019) and the Committee 
for Medical Ethics of the Institute of Oncology 
Ljubljana (OI: 28.5.2019, ERIDEK-0064/2019). 
Written informed consent was obtained from all 
patients.
Patient selection
We consecutively included patients with previ-
ously untreated locally advanced HPV-negative 
OPSCC who were scheduled for curative cCRT. 
Patients were treated with the intensity-modulated 
radiotherapy (RT) technique to a radiation dose of 
70 Gy in 35 fractions over 7 weeks. Cisplatin 40 mg/
m2/week IV or carboplatin 1.5 AUC/week were ad-
ministered concurrently with RT. HPV status was 
determined by p16 immunohistochemistry and in 
situ hybridization studies. Patients with concur-
rent malignancies and those unable to undergo 
MRI examination were excluded from the study.
Study design
Patients underwent DCE-MRI before treatment 
initiation, after receiving 20 Gy of irradiation, and 
10−12 weeks after cCRT completion. Peripheral 
venous blood samples were obtained at each MRI 
timepoint.
MR imaging
MRI scans were conducted using a 3.0T 
MAGNETOM Trio, A Tim System scanner 
(Siemens Healthineers®, Forchheim, Germany), 
equipped with a head and neck receive coil. The 
diagnostic imaging protocol included an axial 
T2-weighted sequence with short tau inversion 
recovery (STIR) (TR/TE 5010/71 ms, TI 170 ms, ma-
trix size 256 x 256, slice thickness 3 mm, 10% gap, 
Radiol Oncol 2024; 58(3): 366-375.
Longo A et al. / DCE-MRI and angiogenesis measures for response prediction in HPV-negative HNSCC368
and field of view (FOV) 18 x 18 cm) and a contrast-
enhanced (CE) axial T1-weighted volumetric inter-
polated breath-hold examination (VIBE) sequence 
(TR/TE 3.26/1.26 ms, voxel size 1.1 x 0.9 x 1.5 mm, 
matrix size 288 x 288, and FOV 250 cm²). DCE-MRI 
was performed using a 3D fast low-angle shot 
(FLASH) sequence optimized for spatial and tem-
poral resolution (TR/TE 5/1.16 ms, matrix size 220 
cm², slice thickness 4 mm, temporal resolution 9 s, 
total acquisition time 6 min). T1 mapping was uti-
lized to convert signal intensities into gadolinium 
concentration. The T1 map was derived from two 
pre-contrast flip angle images (6° and 15°). Baseline 
images were acquired for 27 seconds. Intravenous 
administration of the paramagnetic contrast me-
dium gadobutrol (Gadovist®, Bayer HealthCare 
Pharmaceuticals) was performed at a dose of 0.1 
mmol/kg body weight with a flow rate of 3.5 mL/s, 
followed by a 20 mL saline flush. The diagnos-
tic imaging protocol (axial STIR and CE axial T1 
VIBE) was employed for pre-treatment tumour 
evaluation and radiation treatment planning.
Treatment response
Patients were stratified into two groups based on 
clinical evaluation. Responders demonstrated com-
plete disappearance of the primary tumour within 
10−12 weeks following cCRT. Non-responders en-
compassed all other patients, including those with 
partial response, stable disease, or progressive dis-
ease.
DCE-MRI perfusion analysis
We conducted the post-processing of DCE 3D 
datasets using a commercially available Tissue 
4D workflow within the syngo.via medical imag-
ing software (Siemens Healthineers, Forchheim, 
Germany). We utilized a population-based arterial 
input function with an intermediate pre-set, and 
motion correction was automatically conducted 
using the software algorithm.
Regions of interest (ROI) delineation was veri-
fied by two experienced head and neck radiolo-
gists (A. L., K. Š. P.). ROI were delineated across 
VIBE multiple images, resulting in the generation 
of volumes of interest (VOI). VOI delineation was 
conducted twice for each examination. Initially, the 
entire tumour volume was delineated, excluding 
larger vessels and necrotic regions. These were as-
sessed visually based on their appearance and en-
hancement characteristics, with large necrotic are-
as showing little to no enhancement. Subsequently, 
a smaller tumour volume was delineated, charac-
terized by homogeneous or near-homogeneous 
enhancement, with a substantial margin from sur-
rounding heterogeneities. The Tofts two-compart-
ment pharmacokinetic model was utilised to derive 
a set of perfusion parameters, which included the 
volume transfer constant between blood plasma 
and extracellular extravascular space (ktrans), the to-
tal extracellular extravascular space (EES) volume 
fraction (Ve), the reflux constant (kep) and the initial 
area under the curve (iAUC) for the selected VOI. 
The software generated parameter maps depicted 
as colour-coded images. The mean values of per-
fusion parameters from the two delineated VOIs 
were used for analysis.
Biochemical analysis of plasma 
angiogenic factors
Blood samples underwent processing to yield 
0.5 ml plasma samples, which were immediately 
stored in the Protein LoBind® Tubes (Eppendorf, 
Hamburg, Germany) at -80°C. Plasma concentra-
tions of VEGF, CTGF, PDGF-AB, ANG, END, and 
THBS1 were measured at selected time-points 
using the enzyme-linked immunosorbent as-
say (ELISA) technique. The following ELISA 
Kits were used for the analysis: Human VEGF 
ELISA Kit (KHG0111), Human PDGF-AB ELISA 
Kit (EHPDGFAB), Human Angiogenin ELISA 
Kit (EHANG), Human Endostatin ELISA Kit 
(EHCOL18A1), Human Thrombospondin 1 ELISA 
Kit (EHTHBS1), all Invitrogen (Life Technologies 
Corporation, Carlsbad, CA, USA); and Human 
CTGF Mini ABTS ELISA Development Kit Catalog# 
900-M317 Range (PEPROTech, EC Ltd., London, 
VB, part of Thermo Fisher Scientific Inc., Waltham, 
MA, USA).
All steps were performed following the manu-
facturers’ protocols, and each assay was conduct-
ed in duplicate or triplicate to ensure accuracy 
and reliability. To determine the levels of ANG, 
END, PDGF-AB, and THBS1, the plasma samples 
underwent dilution at ratios of 8000-, 100-, 2-, and 
400-fold using 1x Assay Diluent A or D. The ab-
sorbance measurements were conducted using 
Synergy H4 (Biotek) at 450 or 405 nm with wave-
length correction at 650 nm. Standard curves were 
calculated in GraphPad Prism V. 10 for Windows 
(GraphPad Software, Boston, MA, USA, www.
graphpad.com) using curve fitting analyses, fol-
lowed by the calculation of results for plasma sam-
ples. Sample concentrations were expressed in ng/
mL or pg/mL.
Radiol Oncol 2024; 58(3): 366-375.
Longo A et al. / DCE-MRI and angiogenesis measures for response prediction in HPV-negative HNSCC 369
Statistical methods
Shapiro-Wilk tests were used to assess the normal-
ity of data distributions. Depending on the results, 
either Student’s t-tests for independent samples or 
Mann-Whitney U tests were employed to investi-
gate disparities in DCE-MRI perfusion parameters 
and plasma levels of angiogenic factors between 
responders and non-responders. P-value below 
0.05 was considered significant, while a p-value 
between 0.05 and 0.10 was regarded as indicative 
of a trend. Spearman correlation analysis was per-
formed to analyse correlations between DCE-MRI 
perfusion parameters and biochemical param-
eters. Predictive models for treatment outcomes 
were developed using logistic regression analysis 
and a classification tree model. These models were 
validated using the leave-one-out strategy. The da-
ta analysis was carried out using Python extension 
libraries, including Orange10, Pandas11, NumPy11, 
Matplotlib12, and Scipy13.
Results
Clinical characteristics
Twenty-five patients met the inclusion criteria. 
Initial DCE-MRI was performed on average 11 
days (SD 4.2; range 3−18) prior to cCRT initiation, 
and the intra-treatment DCE-MRI was performed 
after receiving an average of 21 Gy (SD 2.9; range 
18−34) of irradiation. Venous samples were collect-
ed an average of 11 days (SD 6.4; range 2−19) prior 
to cCRT initiation, and the intra-treatment samples 
were obtained after receiving an average of 23 Gy 
(SD 5.7; range 18−40) of irradiation. Patient and tu-
mour characteristics are summarised in Table 1.
Data exclusion and missing values in 
analysis
Significant discrepancies were noted in the DCE-
MRI parameter values for one patient, indicating a 
possible technical issue. As a result, we decided to 
exclude this patient with stage III OPSCC, leaving 
us with 24 patients for analysis. Additionally, there 
was missing data in the biochemical analysis, with 
two patients failing to attend the follow-up visit 
at 10−12-week post-cCRT. Furthermore, some data 
were missing due to values falling below the de-
tection limit, including 8 samples for ANG and 1 
sample for THBS1.
Treatment outcome
The median follow-up for 24 patients was 30 
months (range 17–53 months). At the last follow-up 
visit, 17 patients were alive, 3 of them with active 
malignant disease. 7 patients died, with index can-
cer and infection being the cause of death in 5 and 
2 patients, respectively. The latter 2 patients were 
free of treated cancer at the time of death. Among 
the participants, 15 were classified as responders 
(complete response), whereas 9 were categorised as 
non-responders (incomplete response). All non-re-
sponders experienced disease progression during 
the follow-up period. Of these, one individual was 
eligible for salvage surgery and remained recur-
rence-free throughout the follow-up period, while 
two others developed distant metastases, involv-
ing the lung and skin, respectively. Conversely, 
among the responders, one patient experienced 
local recurrence and subsequently underwent sur-
gery, while another developed distant metastasis 
involving the bone and lung. The remaining 13 
responders remained disease-free throughout the 
follow-up period.
Correlation between perfusion 
parameters and plasma angiogenic 
factors
Spearman correlation analysis revealed no strong 
correlations between DCE-MRI perfusion param-
eters and biochemical parameters. However, mod-
erate positive correlations were observed between 
kep and both VEGF and ANG before treatment, as 
well as between VEGF and both ktrans and iAUC af-
ter receiving 20 Gy of RT. On the other hand, mod-
erate negative correlation was observed between 
ktrans and THBS1 and also between iAUC and CTGF 
after receiving 20 Gy of RT. The remaining param-
TABLE 1. Summary of patient, tumour and treatment 
characteristics
Number of patients 25
Age, median (range) 60 (45−70)
Male gender (%) 25 (100)
Primary location (%)
    Oropharynx 25 (100)
Stage (%)
    III 5 (20)
    IVa 8 (32)
    IVb 12 (48)
Radiotherapy dose (Gy), median (range) 70 (68−70)
Concomitant chemotherapy, median no. 
of cycles (range) 6 (1−7)
Radiol Oncol 2024; 58(3): 366-375.
Longo A et al. / DCE-MRI and angiogenesis measures for response prediction in HPV-negative HNSCC370
eters exhibited weak correlations, including ktrans 
and VEGF before treatment (Spearman rho = 0,166) 
(Table 2, Figure 1). 
Representative co-registered anatomic MR im-
ages and colour-coded pharmacokinetic para-
metric maps of a responder and a non-responder 
patient before and after 20 Gy RT are shown in 
Figures 2 and 3, respectively.
Dynamic contrast-enhanced (DCE) 
parameters
Compared to non-responders, responders exhib-
ited higher ktrans values before treatment (0.270, SD 
0.087 min-1 vs. 0.169, SD 0.062 min-1, p = 0.006) as 
well as after receiving 20 Gy of RT (0.289, SD 0.067 
min-1 vs. 0.215, SD 0.027 min-1, p = 0.007) (Table 3). 
After receiving 20 Gy of RT, an increase in ktrans 
was observed in both groups. iAUC was larger in 
responders compared to non-responders before 
treatment (0.334, SD 0.109 vs. 0.220, SD 0.092, p = 
0.015) but not after receiving 20 Gy of RT (0.361, 
SD 0.070 vs. 0.336, SD 0.099, p = 0.468). Before treat-
ment, responders exhibited a larger Ve compared 
to non-responders (0.306, SD 0.082 vs. 0.237, SD 
0.108, p = 0.092); however, this difference was only 
observed as a trend. After receiving 20 Gy of RT, 
there was an increase in Ve in all patients, but the 
fractional increase in Ve between responders and 
non-responders was insignificant. 
Responders exhibited a trend of higher kep val-
ues compared to non-responders after receiving 
20 Gy of RT (0.819, SD 0.229 vs. 0.709, SD 0.086, p 
= 0.065). Before treatment, kep values did not dis-
tinguish between responders and non-responders. 
Biochemical analysis
Plasma levels of VEGF were consistently higher 
in responders compared to non-responders across 
all three time points – before treatment (p = 0.011), 
after receiving 20 Gy of RT (p < 0.001) as well as 
10−12 weeks following the completion of cCRT (p 
= 0.010) (Table 4). There were no significant differ-
ences observed in the dynamics of plasma VEGF 
levels between the two groups (p = 0.462).
Plasma levels of PDGF-AB showed a decreasing 
trend across the three time points for all patients, 
TABLE 2. Correlations between dynamic contrast-enhanced magnetic resonance 
imaging (DCE-MRI) perfusion parameters and biochemical parameters, along 
with their respective Spearman’s rho and p-values
Time-point DCE-MRI parameter
Biochemical 
parameter
Spearman  
rho p
Before 
treatment
kep VEGF 0.433 0.034*
kep ANG 0.420 0.058
After 20 Gy 
ktrans THBS1 -0.453 0.026*
ktrans VEGF 0.362 0.083
iAUC VEGF 0.391 0.056
iAUC CTFG -0.315 0.134
ANG = angiogenin; CTGF = connective tissue growth factor; iAUC = initial area under the 
curve; THBS1 = thrombospondin-1; VEGF = vascular endothelial growth factor
FIGURE 1. A correlation heatmap visualising perfusion parameters alongside plasma concentrations of angiogenic factors before concurrent 
chemoradiotherapy (cCRT) initiation (left) and after receiving 20 Gy radiotherapy (right).
Radiol Oncol 2024; 58(3): 366-375.
Longo A et al. / DCE-MRI and angiogenesis measures for response prediction in HPV-negative HNSCC 371
with no significant difference observed in the 
dynamics between responders and non-respond-
ers. After receiving 20 Gy of RT, plasma levels of 
PDGF-AB were higher in responders compared to 
non-responders (p = 0.021), whereas no significant 
difference was observed between the two groups 
before treatment (p = 0.431) and 10−12 weeks after 
CRT completion (p = 0.919).
Apart from several values falling below the 
detection limit, there was considerable variation 
in the measured levels of plasma ANG among all 
patients. No notable dynamics or significant dif-
ferences were observed between responders and 
non-responders. Plasma levels of THBS1 remained 
stable across the three time points, with no dis-
cernible differences observed between respond-
ers and non-responders. An insignificant increase 
in plasma levels of END was observed across the 
three time points in all patients, with no noticeable 
differences between responders and non-respond-
ers. Plasma levels of CTGF displayed considerable 
variability, with no notable differences observed 
between responders and non-responders (Table 4).
Predicting treatment response
Based on Spearman correlation analysis and de-
scriptive statistics for DCE-MRI and biochemical 
parameters between responders and non-respond-
ers, we selected two variables for inclusion in sta-
tistical modelling to predict treatment outcomes, 
namely ktrans and VEGF as measured after receiv-
ing 20 Gy of RT. The multivariate logistic regres-
sion (LR) analysis, which incorporated two predic-
tor variables, ktrans and plasma VEGF after receiv-
ing 20 Gy of RT, yielded a classification accuracy 
(CA) of 0.708 and area under the ROC curve (AUC) 
of 0.859. Notably, the variable VEGF achieved sta-
tistical significance (p = 0.033), whereas the vari-
able ktrans approached significance (p = 0.056). 
A univariate LR analysis utilizing only the vari-
able ktrans, measured after receiving 20 Gy of RT, 
exhibited a higher CA of 0.792, but a significant-
ly lower AUC of 0.625. Notably, the variable ktrans 
demonstrated statistical significance (p = 0.019). 
The classification tree model, using the same 
two variables, achieved CA of 0.833 and AUC of 
0.881. The classification tree (Figure 4) structure 
consists of one terminal node on the first branch 
and two terminal nodes on the second. Variable 
ktrans was found to be the most influential in deter-
mining treatment response with a cut off value of 
0.259 min-1, followed by variable VEGF with a cut 
off value of 62.5 pg/mL.
Discussion
In the present study we confirmed the efficacy of 
DCE-MRI in combination with plasma angiogenic 
factor analysis before and early during cCRT in 
predicting response to cCRT in patients with HPV-
negative OPSCC. 
The key DCE-MRI perfusion parameter ktrans re-
flects microvascular permeability and blood flow 
within the primary tumour.10 In our study, we 
showed that responders to cCRT had significant-
ly higher pre-treatment ktrans values compared to 
non-responders. This finding is in line with several 
previous reports on HNSCC and other malignan-
cies that linked higher pre-treatment ktrans values 
to longer survival and better local control.10-16 This 
supports the premise that higher perfusion within 
tumours is associated with increased oxygenation 
levels, which in turn enhances radiosensitivity 
and improves chemotherapeutic drug delivery.3 A 
previous study exploring intra-treatment dynam-
ics of DCE-MRI parameters in HNSCC has shown 
a larger fractional increase in primary tumour 
ktrans and Ve in responders versus non-responders.17 
Similarly, patients achieving locoregional control 
TABLE 3. A summary of tumour perfusion characteristics for responders and non-
responders: mean values, standard deviations (SD), and corresponding p-values
DCE-MRI parameter Responders  (n = 15)
Non-responders 
(n = 9) p
ktrans (SD)
   Before treatment 0.270 (0.087) 0.169 (0.062) 0.006**
   After 20 Gy 0.289 (0.067) 0.215 (0.027) 0.007**
   Δ 0.019 (0.060) 0.047 (0.078) 0.326
kep (SD)
   Before treatment 0.924 (0.231) 0.817 (0.258) 0.144
   After 20 Gy 0.819 (0.229) 0.709 (0.086) 0.065
   Δ -0.105 (0.274) -0.108 (0.245) 0.980
Ve (SD)
   Before treatment 0.306 (0.082) 0.237 (0.108) 0.092
   After 20 Gy 0.366 (0.086) 0.347 (0.107) 0.636
   Δ 0.059 (0.093) 0.109 (0.196) 0.721
iAUC (SD)
   Before treatment 0.334 (0.109) 0.220 (0.092) 0.015*
   After 20 Gy 0.361 (0.070) 0.336 (0.099) 0.468
   Δ 0.027 (0.120) 0.116 (0.183) 0.676
iAUC = initial area under the curve
Radiol Oncol 2024; 58(3): 366-375.
Longo A et al. / DCE-MRI and angiogenesis measures for response prediction in HPV-negative HNSCC372
demonstrated persistently elevated or increasing 
levels of intra-treatment blood flow (BF), volume 
(BV), and permeability surface (PS), whereas one 
study reported a decrease in ktrans after 2 weeks 
of cCRT, which was linked to significantly re-
duced survival.4,15 Interestingly, we observed an 
increase in ktrans values among both responders 
and non-responders after receiving 20 Gy RT. This 
observation aligns with the concept that low-dose 
radiation induces proangiogenic signalling, pro-
motes angiogenesis, improves tumour perfusion, 
enhances vascular permeability, reduces hypoxia, 
and subsequently leads to vascular normalisa-
tion.18 Our results suggest that this effect exists 
independently of the final treatment outcome. 
iAUC is influenced by permeability, blood flow, 
and washout, providing an alternative parameter 
to ktrans in clinical trials.19,20 Our study showed that 
pre-treatment iAUC was higher in responders 
compared to non-responders, mirroring the ktrans 
results. However, unlike ktrans, this distinction dis-
appeared after receiving 20 Gy of RT. 
In addition to the increase in blood flow and 
vascular permeability, radiation-induced cellular 
degradation leads to the expansion of interstitial 
space.17 DCE-MRI parameter Ve reflects the EES, 
which consists of interstitial fluid and connective 
tissue and differs significantly in neoplastic tissues 
from most normal tissues.21 Earlier research has in-
dicated a mostly positive correlation between el-
evated pre-treatment Ve and favourable outcomes, 
but these findings were inconsistent.12-14 One 
study reported a greater fractional increase in Ve 
among responders compared to non-responders.17 
Similarly, in the present study we observed a ten-
dency for a larger pre-treatment Ve in responders 
compared to non-responders. We observed a rise 
in Ve across all patients after receiving 20 Gy of RT. 
However, both the fractional increase of Ve and the 
absolute difference in Ve values after receiving 20 
Gy of RT between responders and non-responders 
were found to be insignificant. We assume that the 
differences in the EES of tumours that may exist 
before treatment between responders and non-re-
sponders are diminished by the shared effects of 
radiotherapy.
VEGF is a main driver of angiogenesis, known 
to be upregulated in the plasma of various cancer 
patients.22-24 Several studies have investigated the 
relationship between VEGF expression in tissue 
specimens and DCE-MRI parameters, especially 
ktrans, resulting in conflicting findings.25-28 VEGF 
has been investigated as a potential biomarker in 
various cancer types, including HNSCC.29,30 High 
FIGURE 3. Co-registered volumes of interest (VOI) and color-coded ktrans maps, 
together with concentration curves alongside corresponding ktrans values before 
treatment (upper section) and intra-treatment after receiving 20 Gy radiotherapy 
(bottom section) in a representative non-responder patient.
FIGURE 2. Co-registered volumes of interest (VOI) and color-coded ktrans maps, 
together with concentration curves alongside corresponding ktrans values before 
treatment (upper section) and intra-treatment after receiving 20 Gy radiotherapy 
(bottom section) in a representative responder patient.
FIGURE 4. Graphical presentation of the classification tree model.
cCRT = concurrent chemoradiotherapy; VEGF = vascular endothelial growth factor
Radiol Oncol 2024; 58(3): 366-375.
Longo A et al. / DCE-MRI and angiogenesis measures for response prediction in HPV-negative HNSCC 373
VEGF expression levels in serum have been linked 
to increased tumour growth, invasiveness, and a 
higher risk of metastasis.31 To our knowledge, no 
studies to date have explored the combination of 
DCE-MRI with plasma angiogenic factors as poten-
tial prognostic biomarkers. According to literature, 
the average reported healthy plasma VEGF value 
is 42 pg/mL (SD 20, range 9−126 pg/mL), whereas in 
cancer it is 74 pg/mL (SD 116, range 19−730 pg/mL), 
the latter displaying considerable variability.23 A 
pre-therapeutic plasma VEGF threshold below 26 
ng/L was reported to correlate with a more favour-
able outcome in HNSCC.30 Our results, on the oth-
er hand, indicate a different trend - plasma VEGF 
levels were consistently higher in responders when 
compared to non-responders across all three time 
points – before treatment, after receiving 20 Gy of 
RT, as well as 10−12 weeks following the comple-
tion of cCRT. Explaining the rationale behind this 
observation is challenging, as the factors affect-
ing plasma VEGF levels are highly multifactorial. 
Compared to the total amount of VEGF in the hu-
man body, tumour contributes to a relatively small 
percentage of VEGF. Besides the tumour, skeletal 
muscles seem to serve as a significant reservoir for 
VEGF in the body. Moreover, the renal clearance of 
VEGF should be considered as a contributing fac-
tor to plasma VEGF levels.23 Although it may be 
overly ambitious to assume a direct relationship 
between tumour blood flow, permeability and cir-
culating plasma VEGF levels, the observed higher 
plasma VEGF levels in responders merit considera-
tion and further investigation. Since lower-doses 
of radiation induce proangiogenic signalling and 
TABLE 4. Expression of plasma angiogenic factors for responders and non-responders: mean values, 95% confidence interval 
(CI) and corresponding p-values
Plasma Angiogenic Factor Responders (n = 15) Non-responders (n = 9) p
VEGF (95% CI) pg/mL
   Before treatment 66.9 (58.1–75.7) 50.2 (41.8–58.5) 0.011*
   After 20 Gy 68.4 (61.0–75.7) 48.0 (40.4–55.5) 0.0006**
   10−12 weeks after cCRT completion 74.7 (50.8–98.6) 49.0 (38.5–59.5) 0.010*
PDGF-AB (95% CI) pg/mL
   Before treatment 2826.6 (1909.3–3743.9) 2430.6 (1179.1–3682.1) 0.431
   After 20 Gy 2496.0 (2105.8–2886.2) 1799.1 (1319.3–22789) 0.021*
   10−12 weeks after cCRT completion 1513.5 (902.7–2124.3) 1468.5 (779.1–2157.9) 0.919
ANG (95% CI) pg/mL
   Before treatment 2140.8 (1588.2–2693.4) 1218.0 (223.0–2213.0) 0.060
   After 20 Gy 2124.1 (1653.4–2594.9) 1438.0 (405.6–2470.4) 0.131
   10−12 weeks after cCRT completion 1898.1 (1248.7–2547.6) 1384.8 (255.7–2513.9) 0.347
THBS1 (95% CI) ng/mL
   Before treatment 1113.2 (1069.3–1157.2) 1115.6 (1055.8–1175.5) 0.943
   After 20 Gy 1112.3 (1051.5–1173.1) 1162.7 (1064.2–1261.2) 0.257
   10−12 weeks after cCRT completion 1061.2 (1008.8–1113.8) 1136.8 (1002.9–1270.7) 0.210
END (95% CI) pg/mL
   Before treatment 64792.6 (52731.7–76853.4) 59350.6 (47299.1–71402.1) 0.521
   After 20 Gy 71269.2 (55039.7–87498.7) 59636.7 (48124.0–71149.5) 0.283
   10−12 weeks after cCRT completion 84744.3 (65061.5–104427.2) 74754.7 (54514.5–94994.9) 0.475
CTGF (95% CI) pg/mL
   Before treatment 32.6 (13.9–51.2) 24.4 (14.4–34.5) 0.976
   After 20 Gy 31.2 (24.1–122.8) 23.5 (12.3–34.8) 0.721
   10−12 weeks after cCRT completion 26.1 (15.7–36.6) 23.2 (15.1–31.3) 1
ANG = angiogenin; cCRT = concurrent chemoradiotherapy; CTGF = connective tissue growth factor; END = endostatin; PDGF-AB = platelet-derived 
growth factor AB; THBS1 = thrombospondin-1; VEGF = vascular endothelial growth factor
Radiol Oncol 2024; 58(3): 366-375.
Longo A et al. / DCE-MRI and angiogenesis measures for response prediction in HPV-negative HNSCC374
promote angiogenesis, elevated levels of VEGF, as 
a crucial player in angiogenesis, may manifest in 
the circulating plasma as a reflection of this pro-
cess.18 This could explain why the correlation be-
tween VEGF and ktrans/iAUC appears only after the 
start of the treatment. It could also provide insight 
into why these two parameters performed the best 
when conducting statistical modelling. 
PDGF promotes tumour-associated angiogen-
esis by up-regulating VEGF production.31,32 A 
sole notable distinction between responders and 
non-responders emerged at the intra-treatment 
time-point for plasma PDGF-AB, which displayed 
higher levels among responders. This observation 
emerged in the same intra-treatment time point 
where elevated plasma VEGF levels appeared as 
effective predictors of complete response. This 
further suggests that physiological processes oc-
curring early during cCRT may be reflected in 
the plasma levels of these two pro-angiogenic fac-
tors. Apart from VEGF and PDGF-AB, none of the 
analysed angiogenic factors exhibited significant 
promise for predicting response.
The most effective multivariate predictive mod-
els incorporated intra-treatment plasma VEGF and 
ktrans values. Both logistic regression and classifi-
cation tree models outperformed the univariate 
models that used only the ktrans variable, indicating 
that plasma VEGF levels provide significant addi-
tional predictive value in determining response. 
Based on our analysis, the classification tree mod-
el utilising intra-treatment plasma VEGF and ktrans 
emerged as the best-performing model.
While this study has several limitations, nota-
bly its small sample size and the lack of healthy 
controls for angiogenic factors, our study cohort 
stands out for its high homogeneity in tumour 
stage and HPV status, making it unique compared 
to similar studies. Validating our findings would 
necessitate a larger study, including healthy con-
trols. Additionally, it is important to note that ab-
solute values of perfusion parameters can vary 
depending on the post-processing software used 
for pharmacokinetic modelling in DCE-MRI. 
Therefore, caution is needed when comparing 
these values across different studies.
In conclusion, the early intra-treatment DCE-
MRI parameter ktrans and plasma VEGF levels 
emerged as the two most robust predictors for 
treatment response to cCRT in HPV-negative oro-
pharyngeal cancer, yielding effective predictive 
models with cut-off values of 0.259 min⁻¹ for ktrans 
and 62.5 pg/mL for plasma VEGF.
Acknowledgments
This research was funded by the Slovenian 
Research and Innovation Agency (ARIS), grant 
number P3-0307.
References
1. Barsouk A, Aluru JS, Rawla P, Saginala K, Barsouk A. Epidemiology, risk fac-
tors, and prevention of head and neck squamous cell carcinoma. Med Sci 
2023; 11: 42. doi: 10.3390/medsci11020042
2. Shinomiya H, Nibu KI. Etiology, diagnosis, treatment, and prevention of hu-
man papilloma virus-associated oropharyngeal squamous cell carcinoma. 
Int J Clin Oncol 2023; 28: 975-81. doi: 10.1007/s10147-023-02336-8
3. Chen Z, Han F, Du Y, Shi H, Zhou W. Hypoxic microenvironment in cancer: 
molecular mechanisms and therapeutic interventions. Signal Transduct 
Target Ther 2023; 8: 70. doi: 10.1038/s41392-023-01332-8
4. Martens RM, Noij DP, Ali M, Koopman T, Marcus JT, Vergeer MR, et al. 
Functional imaging early during (chemo)radiotherapy for response predic-
tion in head and neck squamous cell carcinoma; a systematic review. Oral 
Oncol 2019; 88: 75-83. doi: 10.1016/j.oraloncology.2018.11.005
5. Bos P, van der Hulst HJ, van den Brekel MWM, Schats W, Jasperse B, Beets-
Tan RGH, et al. Prognostic functional MR imaging parameters in head and 
neck squamous cell carcinoma: a systematic review. Eur J Radiol 2021;144: 
109952. doi: 10.1016/j.ejrad.2021.109952
6. Kikuchi R, Kikuchi Y, Tsuda H, Maekawa H, Kozaki K, Imoto I, et al. Expression 
and clinical significance of connective tissue growth factor in advanced head 
and neck squamous cell cancer. Hum Cell 2014; 27: 121-8. doi: 10.1007/
s13577-014-0092-0
7. Bran B, Bran G, Hörmann K, Riedel F. The platelet-derived growth factor 
receptor as a target for vascular endothelial growth factor-mediated anti-
angiogenetic therapy in head and neck cancer. Int J Oncol 2009; 34: 255-61. 
doi: 10.3892/ijo_00000147
8. Homer JJ, Greenman J, Stafford ND. Circulating angiogenic cytokines as 
tumour markers and prognostic factors in head and neck squamous cell 
carcinoma. Clin Otolaryngol Allied Sci 2002; 27: 32-7. doi: 10.1046/j.0307-
7772.2001.00519.x
9. Huang C, Zhou X, Li Z, Liu H, He Y, Ye G, et al. Downregulation of thrombos-
pondin-1 by DNA hypermethylation is associated with tumor progression in 
laryngeal squamous cell carcinoma. Mol Med Rep 2016; 14: 2489-96. doi: 
10.3892/mmr.2016.5580
10. Demšar J, Curk T, Erjavec A, Gorup Č, Hočevar T, Milutinovič M et al. Orange: 
Data Mining Toolbox in Python. J Mach Learn Res. 2013; 14: 2349−2353. 
11. McKinney W. Data structures for statistical computing in Python. 
Proceedings of the 9th Python in Science Conference, 2010; 51-56.
12. Hunter J D Matplotlib: A 2D Graphics Environment. Comput Sci Eng 2007; 
9(3), 90–95. 
13. Virtanen P, Gommers R, Oliphant T E, Haberland M, Reddy T, Cournapeau D, 
et. al. SciPy 1.0: Fundamental algorithms for scientific computing in Python. 
Nature Methods 2020; 17(3), 261-272.
14. Bernstein JM, Homer JJ, West CM. Dynamic contrast-enhanced mag-
netic resonance imaging biomarkers in head and neck cancer: potential to 
guide treatment? A systematic review. Oral Oncol 2014; 50: 963-70. doi: 
10.1016/j.oraloncology.2014.07.011
15. Dijkhoff RAP, Beets-Tan RGH, Lambregts DMJ, Beets GL, Maas M. Value of 
DCE-MRI for staging and response evaluation in rectal cancer: a systematic 
review. Eur J Radiol 2017; 95: 155-68. doi: 10.1016/j.ejrad.2017.08.009
16. Ng SH, Liao CT, Lin CY, Chan SC, Lin YC, Yen TC, et al. Dynamic contrast-en-
hanced MRI, diffusion-weighted MRI and 18F-FDG PET/CT for the prediction 
of survival in oropharyngeal or hypopharyngeal squamous cell carcinoma 
treated with chemoradiation. Eur Radiol 2016; 26: 4162-72. doi: 10.1007/
s00330-016-4276-8
Radiol Oncol 2024; 58(3): 366-375.
Longo A et al. / DCE-MRI and angiogenesis measures for response prediction in HPV-negative HNSCC 375
17. Chan SC, Cheng NM, Hsieh CH, Ng SH, Lin CY, Yen TC, et al. Multiparametric 
imaging using 18F-FDG PET/CT heterogeneity parameters and functional 
MRI techniques: prognostic significance in patients with primary advanced 
oropharyngeal or hypopharyngeal squamous cell carcinoma treated with 
chemoradiotherapy. Oncotarget 2017; 8: 62606-21. doi: 10.18632/onco-
target.15904 
18. Ng SH, Lin CY, Chan SC, Yen TC, Liao CT, Chang JT, et al. Dynamic contrast-en-
hanced MR imaging predicts local control in oropharyngeal or hypopharyn-
geal squamous cell carcinoma treated with chemoradiotherapy. PLoS One 
2013; 8: e72230. doi: 10.1371/journal.pone.0072230
19. Baer AH, Hoff BA, Srinivasan A, Galbán CJ, Mukherji SK. Feasibility analysis 
of the parametric response map as an early predictor of treatment efficacy 
in head and neck cancer. Am J Neuroradiol 2015; 36: 757-62. doi: 10.3174/
ajnr.A4296
20. Wong KH, Panek R, Dunlop A, Mcquaid D, Riddell A, Welsh LC, et al. Changes 
in multimodality functional imaging parameters early during chemoradia-
tion predict treatment response in patients with locally advanced head and 
neck cancer. Eur J Nucl Med Mol Imaging 2018; 45: 759-67. doi: 10.1007/
s00259-017-3890-2
21. Yamazaki T, Young KH. Effects of radiation on tumor vasculature. Mol 
Carcinog 2022; 61: 165-72. doi: 10.1002/mc.23360
22. Bisdas S, Seitz O, Middendorp M, Chambron-Pinho N, Bisdas T, Vogl TJ, et 
al. An exploratory pilot study into the association between microcirculatory 
parameters derived by MRI-based pharmacokinetic analysis and glucose 
utilization estimated by PET-CT imaging in head and neck cancer. Eur Radiol 
2010; 20: 2358-66. doi: 10.1007/s00330-010-1803-x
23. Galbraith SM. MR in oncology drug development. NMR Biomed 2006; 19: 
681-9. doi: 10.1002/nbm.1093
24. Lowry M, Zelhof B, Liney GP, Gibbs P, Pickles MD, Turnbull LW. Analysis 
of prostate DCE-MRI: comparison of fast exchange limit and fast ex-
change regimen pharmacokinetic models in the discrimination of malig-
nant from normal tissue. Invest Radiol 2009; 44: 577-84. doi: 10.1097/
RLI.0b013e3181b4c1fe
25. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 
2011; 144: 646-74. doi: 10.1016/j.cell.2011.02.013
26. Kut C, Mac Gabhann F, Popel AS. Where is VEGF in the body? A meta-
analysis of VEGF distribution in cancer. Br J Cancer 2007; 97: 978-85. doi: 
10.1038/sj.bjc.6603923
27. Macedo F, Ladeira K, Longatto-Filho A, Martins SF. Gastric cancer and angio-
genesis: is VEGF a useful biomarker to assess progression and remission? J 
Gastric Cancer 2017; 17: 1-10. doi: 10.5230/jgc.2017.17.e1
28. Liu B, Gao H, Zhou F, Zhao W, Yang Y. Dynamic contrast-enhanced magnetic 
resonance imaging in cervical cancer: correlation between quantitative pa-
rameters and molecular markers hypoxia-inducible factors-1-alpha, vascular 
endothelial growth factor, and Ki-67. Clin Radiol 2024; 79: e826-e33. doi: 
10.1016/J.CRAD.2024.01.039
29. Di N, Cheng W, Jiang X, Liu X, Zhou J, Xie Q, et al. Can dynamic contrast-
enhanced MRI evaluate VEGF expression in brain glioma? An MRI-guided 
stereotactic biopsy study. J Neuroradiol 2019; 46: 186-92. doi: 10.1016/j.
neurad.2018.04.008
30. Lindgren A, Anttila M, Arponen O, Rautiainen S, Könönen M, Vanninen R, et 
al. Prognostic value of preoperative dynamic contra st-enhanced magnetic 
resonance imaging in epithelial ovarian cancer. Eur J Radiol 2019; 115: 66-
73. doi: 10.1016/j.ejrad.2019.03.023
31. Keil VC, Pintea B, Gielen GH, Hittatiya K, Datsi A, Simon M, et al. Meningioma 
assessment: kinetic parameters in dynamic contrast-enhanced MRI ap-
pear independent from microvascular anatomy and VEGF expression. J 
Neuroradiol 2018; 45: 242-8. doi: 10.1016/j.neurad.2018.01.050
32. Kulapaditharom B, Boonkitticharoen V, Sritara C. Plasma vascular en-
dothelial growth factor dysregulation in defining aggressiveness of head 
and neck squamous cell carcinoma. J Oncol 2012; 2012: 687934. doi: 
10.1155/2012/687934
33. Siemert J, Wald T, Kolb M, Pettinella I, Böhm U, Pirlich M, et al. Pre-
therapeutic VEGF level in plasma is a prognostic bio-marker in head and 
neck squamous cell carcinoma (HNSCC). Cancers 2021; 13: 3781. doi: 
10.3390/cancers13153781
34. Dumitru CS, Raica M. Vascular endothelial growth factor family and head 
and neck squamous cell carcinoma. Anticancer Res 2023; 43: 4315-26. doi: 
10.21873/anticanres.16626
35. Laschke MW, Elitzsch A, Vollmar B, Vajkoczy P, Menger MD. Combined 
inhibition of vascular endothelial growth factor (VEGF), fibroblast growth 
factor and platelet-derived growth factor, but not inhibition of VEGF alone, 
effectively suppresses angiogenesis and vessel maturation in endometriotic 
lesions. Hum Reprod 2006; 21: 262-8. doi: 10.1093/humrep/dei308