Research Presentation Session: Head and Neck

RPS 1908 - Head and neck: advances in tumour imaging

March 2, 12:15 - 13:45 CET

7 min
The prediction of biological features using MR imaging in head and neck squamous cell carcinoma: a systematic review and meta-analysis
Hedda Joanne van der Hulst, Amsterdam / Netherlands
Author Block: H. J. van der Hulst, R. W. Jansen, C. Vens, P. Bos, R. Martens, Z. Bodalal, R. G. H. Beets-Tan, M. van den Brekel, J. A. Castelijns; Amsterdam/NL
Purpose: Magnetic resonance imaging (MRI) is an indispensable, routine technique, providing morphological and functional imaging sequences. MRI can potentially capture tumour biology, allowing for longitudinal evaluations of such biology. This systematic review and meta-analysis evaluates the ability of MRI to predict tumour biology in primary head and neck squamous cell carcinoma (HNSCC).
Methods or Background: Studies were screened, selected, and assessed for quality using appropriate tools according to the PRISMA criteria. Fifty-eight articles were analysed, examining the relationship between (functional) MRI parameters and biological features and genetics.
Results or Findings: Most studies focused on associations between individual MRI features and HPV status, revealing a consistent pattern wherein HPV-positive tumours exhibit significantly lower ADCmean (SMD: 0.82; p<0.001) and ADCminimum (SMD: 0.56; p<0.001) values. Lower ADCmean values are also associated with high Ki-67 levels, consistent with a diffusion restriction caused by high cellularity. Several perfusion parameters of the vascular compartment were found to be significantly associated with HIF-1α. Analysis of other biological factors (vascular endothelial growth factor [VEGF], epidermal growth factor receptor [EGFR], tumour cell count, p53, and micro vessel density [MVD]) yielded inconclusive results. This review showed that larger datasets with homogenous acquisition are required to develop and test radiomic-based prediction models capable of capturing different aspects of the underlying tumour biology.
Conclusion: Overall, this meta-analysis demonstrates that rapid and non-invasive characterisation of tumour biology features by MR-imaging is feasible and could enhance clinical outcome predictions and personalised patient management for HNSCC.
Limitations: While comprehensive, the review of MRI parameter associations with biological features is limited by the large variety of methods used in the included articles for calculating biological and MRI parameters.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: This was a review article.
7 min
Comparison of diffusion-weighted MRI using imaging split acquisition of fast spin echo signal (SPLICE) and single-shot echo-planar imaging (SS-EPI) in tumours of the head and neck
Hedda Joanne van der Hulst, Amsterdam / Netherlands
Author Block: H. J. van der Hulst, R. Martens, B. Westerink, L. Braun, L. C. ter Beek, R. Tissier, R. G. H. Beets-Tan, M. van den Brekel, J. A. Castelijns; Amsterdam/NL
Purpose: Traditionally, diffusion-weighted imaging (DWI) uses single-shot echo planar imaging (SS-EPI) for its fast read-out but struggles with distortion, predominantly near bone-air interfaces. Turbo Spin Echo (TSE) DWI, like split acquisition of fast spin echo signal for diffusion imaging (SPLICE), may reduce this issue, especially in the head and neck area. This study compares DW-SPLICE with DW-SS-EPI for head and neck lesions (HN-lesions).
Methods or Background: Between August 2020 and January 2022, 57 potential MRI HN-lesion exams were conducted using both SS-EPI and SPLICE techniques on a 3.0T MRI system. Lesions >0.5cm³ were delineated on both DWI-MRIs. The study assessed differences in whole lesion ADC-values through interclass correlation (ICC), Bland-Altman plots, and mixed-effect model regression. Two blinded radiologists evaluated image quality and distortion levels using a standardised questionnaire.
Results or Findings: Preliminary findings:
The eligible 41 MRI exams of 41 patients included 40 benign and 20 malignant HN-lesions, visible on both DWI techniques. Strong interclass correlation coefficients were found for mean ADC values (ICC agreement 0.864, 95% CI 0.782-0.917), consistency 0.862 (95% CI 0.780-0.915). The Bland-Altman regression showed discrepancies between SPLICE and EPI for higher average ADC values (ß=0.16, p=0.02), but not when assessing benign and malignant lesions separately. Radiologist evaluations of visual distortion showed a trend towards less distortion and better image quality for SPLICE-DWI.
Conclusion: High agreement and consistency in mean ADC values were seen for HN-lesions measured with both techniques. Qualitative analyses of radiologists resulted in a trend towards better distortion and image quality scores for SPLICE-DWI. While a potential ADC value discrepancy was observed in the Bland-Altman analyses, its clinical impact may be minimal if DWI-SPLICE can effectively replace SS-EPI.
Limitations: Identified limitations were: (1) the cohort consists of a mixed group of HN-lesions; (2) data is currently still preliminary.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the IRBd22-270, registered at 28.10.2022.
7 min
Does radiotherapy influence the diagnostic performance of ultrasound-guided fine-needle aspiration cytology of neck nodes in head and neck squamous cell carcinoma?
Sophie Neveu, Geneva / Switzerland
Author Block: S. Neveu, A. Escriva, M. Becker; Geneva/CH
Purpose: The study aims to evaluate and compare the diagnostic performance of ultrasound-guided fine-needle aspiration cytology (US-FNAC) in neck nodes of head and neck squamous cell carcinoma (HNSCC) patients with and without previous radiotherapy (RT+ versus RT-).
Methods or Background: Retrospective study comprising a series of 99 consecutive HNSCC patients (66 RT- and 33 RT+, 25 females, 74 males, mean age = 67±12 years). A total of 136 nodal aspirations (92 RT- nodes, 44 RT+ nodes) were analysed. US-FNAC results were compared with the standard of reference (neck dissection and follow-up).
Results or Findings: The minimum/maximum axial diameter of biopsied RT- versus the RT+ nodes was similar (10.97±7.99 mm/15.9±10.9 mm versus 9.07± 5.9 mm/13.6±6.7 mm), t(134)=1.42/1.266, p<0.05. Likewise, nodal level distribution and US imaging characteristics of benign and malignant nodes were similar in both the groups (p>0.05). For diagnostic US-FNAC probes, the sensitivity, specificity, accuracy, positive and negative predictive values (PPV and NPV) of the US-FNAC in the RT- versus the RT+ group were 91.3%, 95.5%, 93.3%, 95.5%, 91.3% versus 94.7%, 88.9%, 91.8%, 89.5%, 94.4%, respectively (p>0.05). However, the proportion of non-diagnostic cytology results was significantly higher (7/44 = 15.9%) in the RT+ versus the RT- group (2/92 = 2.2%), p=0.005.
Conclusion: Although, irradiation significantly increases the proportion of non-diagnostic samples, it has no influence on the diagnostic performance of US-FNAC in neck nodes of HNSCC patients.
Limitations: Limitations of the current study include its retrospective design and a potential selection bias as only patients with US-FNAC were included in the current study.
Funding for this study: Funding was received from the institution.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This retrospective study was approved by the local Ethics Committee and was done in accordance to the Helsinki declaration (CER 11-092R). Informed consent was waived.
7 min
Detection of recurrent nodal disease in head and neck squamous cell carcinoma with multiparametric contrast-enhanced FDG-PET/DWIMRI
Sophie Neveu, Geneva / Switzerland
Author Block: S. Neveu1, A. Varoquaux2, O. Rager1, C. De Vito1, N. Dulguerov1, C. Combescure1, H. Zaidi1, M. Becker1; 1Geneva/CH, 2Marseille/FR
Purpose: The purpose of this study was to determine the diagnostic performance of contrast-enhanced FDG-PET/DWIMRI to detect recurrent nodal disease in head and neck squamous cell carcinoma (HNSCC) treated with radio(chemo)therapy (RCT).
Methods or Background: This prospective study included 47 patients treated with RCT ± surgery for advanced HNSCC who underwent 67 contrast-enhanced FDG-PET/DWIMRI examinations as part of their surveillance protocol. Lymph nodes (LNs) were evaluated prospectively using a 5-point Likert scale by experienced readers. Lymph node size, ADCmean/ADCmin and SUVmean/SUVmax were measured for all detected LNs. The standard of reference was histology of neck dissection specimens in 34% and follow-up ≥ 5years in 66% of patients. The diagnostic performance was assessed both in terms of sensitivity, specificity and accuracy and using Receiver Operating Analysis (ROC) in 67 examinations, 134 neck sides, 938 LN levels and 220 LNs.
Results or Findings: The prevalence of LN metastases depended on the type of analysis performed (50% per neck side, 5% per LN level and 44% per LN). Mean minimum axial diameter was 6.5±4.0 mm in benign and 7.0±5.7mm in malignant LNs (p>0.05) and only 11% of all nodes had a diameter >10 mm. ADC mean/ADC min values were 1.14±0.31|0.62±0.29x10-3 mm2/s and 1.17±0.29|0.78±0.29x10-3 mm2/s in benign and malignant LNs, respectively (p>0.05) while SUV mean/SUV max values were 1.87±0.84|2.38±1.15 in benign and 3.40±2.38|4.79±3.45 in malignant LNs (p<0.05). FDG-PET/DWI-MRI had a sensitivity/specificity/accuracy of 90%/64%/72% per examination, 88%/82%/83% per neck side, 74.5%/97%/96% per LN level, and 49.5%/76.4%/64.5% per LN. The areas under the curve (AUC) were 0.926 per examination, 0.933 per neck side, 0.874 per LN group and 0.661 per LN.
Conclusion: Analysis type heavily influences results. As most recurrent metastatic LNs are <1 cm in size, the sensitivity of contrast-enhanced FDG-PET/DWIMRI is limited in the per level and per LN analysis.
Limitations: The limitation was that it was a single study center.
Funding for this study: Funding was received from the Swiss National Science Foundation (SNSF) (grant number: N0 320030_173091/1).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This prospective clinical study was approved by the Swissethics committee (Project ID 2017-00748) and performed in accordance with the guidelines of the Helsinki II declaration. Written informed consent was obtained from all subjects.
7 min
Predicting response to exclusive combined radio-chemotherapy in naso-oropharyngeal cancer: the role of texture analysis
Eleonora Bicci, Florence / Italy
Author Block: E. Bicci, L. Calamandrei, M. Pietragalla, S. Busoni, S. Paolucci, F. Mungai, C. Nardi, L. Bonasera, V. Miele; Florence/IT
Purpose: The aim of the study was to identifying MRI texture features able to predict response to radiochemotherapy (RT-CHT) in patients with naso-oropharyngeal carcinoma before treatment to help clinical decision making.
Methods or Background: 41 patients with naso-oropharyngeal carcinoma (NPC-OPC) were enrolled in the study and were divided into 2 groups (responders/non-responders) according to MRI, 18F-FDG-PET/CT performed 3-4 and 12 months after RT-CHT, and biopsy.
Pre-RT-CHT lesions were segmented and radiomic features were extracted from ADC-maps and post-gadolinium T1 MRI images acquired on a single MRI scanner.
A radiomic model differentiating responders from non-responders was built using the LASSO algorithm on all the features extracted, which allows at the same time to make feature reduction and selection.
Receiver Operating Characteristic (ROC) curves and Area Under the Curve (AUC) values were generated a 95% confidence interval (CI) was reported.
Results or Findings: LASSO model selected five statistically significant features: Maximum 2D Diameter Slice, Dependence Non-uniformity, Gray Level Non-Uniformity, Small Area high gray level emphasis from post-gadolinium T1 MRI images and 10 Percentile from ADC maps AUC associated with the ROC was: 0.978 with 95% CI: 0.937-1.
Conclusion: Texture analysis on post-gadolinium T1-images and ADC-maps could potentially predict response to therapy in patients with NPC-OPC who will undergo exclusive treatment with RT-CHT, being, therefore, a useful tool in clinical decision-making.
Limitations: The limitations of this study were that (1)The use of an MRI-unit only is an advantage to make the sample as homogeneous as possible, but currently a disadvantage for the lack of generalizability of results. (2) The relative low sample size. Nevertheless, most studies on texture analysis of OPC and NPC are based on Computed Tomoghrapy (CT) and not MRI examinations.(3) The small number of patients with residual cancer for the well-known excellent response to RT-CHT of OPC (especially HPV+) and NPC.
Funding for this study: No funding was received for the study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the Ethics Committee of our Institution (study protocol n: 21800)
7 min
Early intra-treatment T1 rho imaging predicts treatment response to induction chemotherapy in patients with advanced nasopharyngeal carcinoma
Qi Yong AI, Hong Kong / China
Author Block: Q. Y. AI, Y. M. Tsang, Z. Yu, B. B. Ma, E. P. Hui, W. Chen, A. D. King; Hong Kong/HK
Purpose: The aim of the study was early prediction of treatment response to induction chemotherapy (IC) in patients with advanced nasopharyngeal carcinoma (NPC) that could help clinicians to adjust treatment plans. This study evaluates whether quantitative T1 rho imaging, a pulse sequence that is newly applied to research in cancer, can predict treatment response to IC at early intra-treatment stage in patients with advanced NPC.
Methods or Background: Twenty-five eligible patients who had advanced NPC planned for treatment with 3-cycle of IC + concurrent chemoradiation therapy were prospectively recruited. T1rho MRI scans of the primary tumour were performed pre-treatment and early intra-treatment (2-week after IC started) and the T1rho mean value was obtained. Treatment response to IC was evaluated at the third cycle of IC based on the RECIST 1.1. T1rho value of pre-treatment (T1rhopre) and percentage change in T1rho between pre- and intra-treatment scans (%Δ) =[(T1rhointra-T1rhopre)/T1rhopre *100%] were compared between responders (complete response and partial response) and non-responders (stable disease and progression disease) using Mann-Whitney U test. Median (interquartile) values were reported.
Results or Findings: Comparing responders (n= 15) and non-responders ( n=10), responders showed higher pre-treatment T1 rhopre [66.6 (66.2, 74.3) x msec vs. 63.8 (61.7, 69.5) x msec, p=0.03] and greater reduction in T1 rho [ %Δ decrease of -9.0% (-12.1%, 3.9%) vs. 3.2% (-4.9%, 6.6%), (p=0.03)].
Conclusion: Pre-treatment and early intra-treatment change in the T1 rho mean values show potential for predicting treatment response to IC in patients with advanced NPC.
Limitations: Preliminary results were reported. Further evaluation is needed.
Funding for this study: This study was funded by Research Grant Committee of The Hong Kong S.A.R. Government (Ref. 14100419)
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the Joint Chinese University of Hong Kong and New Territories East Cluster Clinical Research Ethics Committee.
7 min
Radiologic predictors of response to neoadjuvant chemotherapy in patients with advanced oral cavity tumours
Iris Burck, Frankfurt a. Main / Germany
Author Block: I. Burck, A. Gleich, E. Herrmann, J-E. Scholtz, J. Gruen, P. Thönissen, M. Fleischmann, R. Winkelmann, D. Pinto Dos Santos; Frankfurt a. Main/DE
Purpose: The aim of this study was to explore whether radiological imaging features could predict response to neoadjuvant radiochemotherapy in patients with locally advanced oral-cavity tumour.
Methods or Background: We included 28 Patients (14 women, median age 59+-9.5) with stage IVa oral cavity carcinoma who underwent neoadjuvant radiochemotherapy at our institution between February 2012 and August 2022. All patients received MRI scans before treatment, after 15 days of treatment and preoperatively. MRI images were judged qualitatively based on a Lickert scale with respect to their overall intensity and change in intensity and tumour extent between the scans. Additionally, scans were analysed quantitatively by measuring the absolute signal intensity (SI) of the tumour in ADC, DWI and T2 and normalising the values to the signal intensity of the spinal cord. Differences and quotients of normalised ADC-, DWI- and T2-SIs were calculated. Patients with pathological T1 or T0 stage were classified as responders, while all others were classified as non-responders.
Results or Findings: We found significant differences between responders and non-responders for normalized ADC-SI at both follow-up timepoints (p = 0.014 and p = 0.024, respectively), as well as for the quotient between normalized DWI-SI at baseline and preoperatively (p = 0.011) and the difference between normalized ADC-SI at baseline and preoperatively (p = 0.035).
Conclusion: Diffusion weighted imaging parameters (normalized ACD-SI and normalized DWI-SI) could predict response to neoadjuvant radiochemotherapy in jaw carcinoma and could potentially be used to guide treatment or extent of surgery in these patients.
Limitations: Limitations of this study include small sample size, single-centre and retrospective study design.
Funding for this study: No funding was obtained for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the Institutional Review Board (approvals number 208/12 )
7 min
Image-guided intraoperative assessment of surgical margins in oral cavity squamous cell cancer: a diagnostic test accuracy review
Giorgia Carnicelli, Rome / Italy
Author Block: G. Carnicelli, S. Lusi, G. M. Frigerio, L. Disconzi, G. Mercante, L. Balzarini, M. Francone, C. Giannitto; Milan/IT
Purpose: The objective of this study was to assess the imaging guidance of oral cavity squamous cell cancer. Incomplete primary resection of oral cavity squamous cell cancer has a dramatic impact on prognosis and re-treatment. Imaging guidance may improve the rate of negative margin resection (currently around 45%). A diagnostic test accuracy review was designed to assess the performance of intraoperative ultrasound and magnetic resonance imaging (MRI) in determining free margin status.
Methods or Background: Two independent reviewers systematically screened articles from EMBASE, Cochrane and MEDLINE databases. The study protocol was designed according to standards of the Cochrane handbook for diagnostic test accuracy (DTA) reviews. Among inclusion criteria were MRI and ultrasound-based studies (2016-2023), a free-margin resection cut-off of 5 mm, frequency of US probes ≥ 15 MHz. Negative predictive value (NPV), sensitivity and specificity of intraoperative imaging were calculated for selected studies. The software used for the review process was Review Manager 5.4 .
Results or Findings: Ten articles were included in the final analysis, four used for data extraction (Figure 1). Overall NPV of both techniques for margins < 5 mm ranged from 0.50 to 0.91, with higher diagnostic yield of ultrasound compared to MRI. Sensitivity of imaging ranged from 0.07 to 0.75; specificity ranged from 0.81 to 1.00 (Figure 2). Overall, the mean rate of free margin resection was 55.8% ± 28 . In all studies with control group, imaging guidance always provided a benefit, with increase in free-margin resection of 35% ± 13; the accuracy decreased for stages ≥ T1-2 and with increasing depth of invasion (DOI) values .
Conclusion: NPV of intraoperative ultrasound in guiding oral cavity cancer resection remains suboptimal with 40-50% margins being missed, and varies with DOI and T stage. Nevertheless, it provides a mean 35% improvement in free margin resection rate.
Limitations: Heterogeneity of the included studies limits the scope of this study.
Funding for this study: No funding was obtained for this study.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: There was no ethical committee approval for reasons related to the nature of the study.
7 min
Reproducibility and prognostic value of NI-RADS score in early post-operative CT for oral cavity cancer
Giovanni Mazzon, Padua / Italy
Author Block: M. Panfili, G. Mazzon, S. Longo, S. Settimi, F. Pastore, D. A. Mele, M. Massaccesi, R. Calandrelli, S. Gaudino; Rome/IT
Purpose: This retrospective study aims to assess the inter-observer concordance of the NI-RADS score applied both for primary site and regional lymph-node levels in early post-operative CT scans in patients with surgically treated oral cavity cancer and its correlation with outcomes. The Neck Imaging reporting and Data System (NI-RADS) is a valuable tool for post-treatment imaging in head and neck cancers with good capability to predict disease recurrence. Yet, it remains underexplored its specific utility and prognostic value in early post-operative computed tomography (CT) scans in surgically treated patients with oral cavity cancer (OC) and eligible for adjuvant therapy.
Methods or Background: Two experienced radiologists independently reviewed CT scans performed less than 90 days from surgery and assigned NI-RADS scores both for T site and N site. NI-RADS scores ranged from 1 to 3. The Cohen's kappa was used to test the interobserver agreement and the Kaplan-Meier survival analysis was used to evaluate correlation between NI-RADS scores and patients overall survival.
Results or Findings: Forty-two CT scans were reviewed with a median post-surgery time of 40 days. The Cohen's kappa statistic indicated substantial inter-observer agreement for the primary tumour (kappa = 0.91, 95% CI 0.79-1.00) and regional lymph nodes (kappa = 0.92, 95% CI 0.77-1.00). Patients with a NI-RADS score of 3 at the primary tumour level had a shorter median OS of 7 months compared to those with NI-RADS scores of 1, 2a, and 2b, who had a median OS of 68 months. This difference was statistically significant with a p-value of 0.006.
Conclusion: The study demonstrates that NI-RADS score in early post-operative CT has good reproducibility and could have prognostic implications.
Limitations: Retrospective nature and small number of patients.
Funding for this study: No funding was obtained for the study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: No information provided by the submitter.
7 min
Diagnostic accuracy of MRI in detecting perineural spread of head and neck tumours: a systematic review and meta-analysis
Umida Bafoevna Abdullaeva, Tashkent / Uzbekistan
Author Block: U. B. Abdullaeva1, B. Pape2, J. Hirvonen3; 1Tashkent/UZ, 2Vaasa/FI, 3Tampere/FI
Purpose: The aim of this study was to review the diagnostic accuracy of MRI in detecting perineural spread (PNS) of head and neck tumours using histopathological or surgical evidence from the afflicted nerve as the reference standard.
Methods or Background: Previous studies in the English language published in the last 30 years were searched from PubMed and Embase databases. We included studies that used MRI (with and without contrast enhancement) to detect PNS, histological or surgical confirmation of PNS, and reported exact numbers of patients required for assessing diagnostic accuracy. The outcome measures were sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Heterogeneity was assessed with the Higgins inconsistency test (I2). P values smaller than 0.05 were considered statistically significant.
Results or Findings: A total of 11 retrospective studies were found reporting 319 nerve samples from 245 patients. Sensitivity ranged from 0.46 to 1.00, and specificity from 0.83 to 1.00, with median values of 0.96 and 0.88, respectively. Meta-analytic estimates and their 95% confidence intervals were: 0.85 (0.70-0.95), specificity 0.85 (0.80-0.89), PPV 0.86 (0.70-0.94), and NPV 0.85 (0.71-0.93). We found statistically significant heterogeneity for sensitivity (I2=72%, p=0.003) and PPV (I2=70%, p=0.038), but not for NPV (I2=65%, p=0.119) or specificity (I2=12%, p=0.842). The most frequent MRI features of PNS were nerve enlargement and enhancement. Squamous cell carcinoma and adenoid cystic carcinoma were the most common tumour types, and the facial and trigeminal nerves were the most commonly affected nerves in PNS. Only a few studies provided examples of false MRI diagnoses.
Conclusion: MRI has high diagnostic accuracy in depicting PNS of cranial nerves, yet this statement is based on scarce and heterogeneous evidence.
Limitations: The study is based on a limited number of retrospective studies with some studies subject to sampling bias.
Funding for this study: This study was financially supported by the Sigrid Jusélius Foundation.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: Since this is a systematic review, Institutional Review Board approval was not necessary.

This session will not be streamed, nor will it be available on-demand!