External validation of deep learning-derived 18F-FDG PET/CT delta biomarkers for loco-regional control in head and neck cancer

Authors

  • David Gergely Kovacs Department of Clinical Physiology and Nuclear Medicine, Rigshospitalet, Copenhagen, Denmark; Department of Clinical Medicine, University of Copenhagen, Copenhagen, Denmark https://orcid.org/0000-0002-0383-1446
  • Marianne Aznar Division of Cancer Sciences, University of Manchester, Manchester, the United Kingdom of Great Britain and Northern Ireland; The Christie NHS Foundation Trust, Manchester, the United Kingdom of Great Britain and Northern Ireland
  • Marcel van Herk Division of Cancer Sciences, University of Manchester, Manchester, the United Kingdom of Great Britain and Northern Ireland; The Christie NHS Foundation Trust, Manchester, the United Kingdom of Great Britain and Northern Ireland
  • Iskandar Mohamed The Christie NHS Foundation Trust, Manchester, the United Kingdom of Great Britain and Northern Ireland
  • James Price The Christie NHS Foundation Trust, Manchester, the United Kingdom of Great Britain and Northern Ireland
  • Claes Nøhr Ladefoged Department of Clinical Physiology and Nuclear Medicine, Rigshospitalet, Copenhagen, Denmark; Department of Applied Mathematics and Computer Science, Technical University of Denmark, Lyngby, Denmark
  • Barbara Malene Fischer Department of Clinical Physiology and Nuclear Medicine, Rigshospitalet, Copenhagen, Denmark; Department of Clinical Medicine, University of Copenhagen, Copenhagen, Denmark
  • Flemming Littrup Andersen Department of Clinical Physiology and Nuclear Medicine, Rigshospitalet, Copenhagen, Denmark; Department of Clinical Medicine, University of Copenhagen, Copenhagen, Denmark
  • Andrew McPartlin Princess Margaret Cancer Centre, University of Toronto, Toronto, Canada
  • Eliana M. Vasquez Osorio Division of Cancer Sciences, University of Manchester, Manchester, the United Kingdom of Great Britain and Northern Ireland; The Christie NHS Foundation Trust, Manchester, the United Kingdom of Great Britain and Northern Ireland https://orcid.org/0000-0003-0741-994X
  • Azadeh Abravan Division of Cancer Sciences, University of Manchester, Manchester, the United Kingdom of Great Britain and Northern Ireland; The Christie NHS Foundation Trust, Manchester, the United Kingdom of Great Britain and Northern Ireland https://orcid.org/0000-0003-4839-6705

DOI:

https://doi.org/10.2340/1651-226X.2025.43977

Keywords:

Head and Neck Neoplasms, Positron Emission Tomography Computed Tomography, Neural Networks (Computer), Radiotherapy, Biomarkers

Abstract

Background and purpose: Delta biomarkers that reflect changes in tumour burden over time can support personalised follow-up in head and neck cancer. However, their clinical use can be limited by the need for manual image segmentation. This study externally evaluates a deep learning model for automatic determination of volume change from serial 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography/computed tomography (PET/CT) scans to stratify patients by loco-regional outcome.

Patient/material and methods: An externally developed deep learning algorithm for tumour segmentation was applied to pre- and post-radiotherapy (RT, with or without concomitant chemoradiotherapy) PET/CT scans of 50 consecutive head and neck cancer patients from The Christie NHS Foundation Trust, UK. The model, originally trained on pre-treatment scans from a different institution, was deployed to derive tumour volumes at both time points. The AI-derived change in tumour volume (ΔPET-Gross tumour volume (GTV)) was calculated for each patient. Kaplan-Meier analysis assessed loco-regional control based on ΔPET-GTV, dichotomised at the cohort median. In a separate secondary analysis confined to the pre‑treatment scans, a radiation oncologist qualitatively evaluated the AI‑generated PET‑GTV contours.

Results: Patients with higher ΔPET-GTV (i.e. greater tumour shrinkage) had significantly improved loco-regional control (log-rank p = 0.02). At 2 years, control was 94.1% (95% CI: 83.6–100%) vs. 53.6% (95% CI: 32.2–89.1%). Only one of nine failures occurred in the high ΔPET-GTV group. Clinician review found AI volumes acceptable for planning in 78% of cases. In two cases, the algorithm identified oropharyngeal primaries on pre-treatment PET-CT before clinical identification.

Interpretation: Deep learning-derived ΔPET-GTV may support clinically meaningful assessment of post-treatment disease status and risk stratification, offering a scalable alternative to manual segmentation in PET/CT follow-up.

Downloads

Download data is not yet available.

References

Marur S, Forastiere AA. Head and neck squamous cell carcinoma: update on epidemiology, diagnosis, and treatment. Mayo Clin Proc. 2016;91(3):386–96.

https://doi.org/10.1016/j.mayocp.2015.12.017 DOI: https://doi.org/10.1016/j.mayocp.2015.12.017

Hillner BE, Siegel BA, Liu D, Shields AF, Gareen IF, Hanna L, et al. Impact of positron emission tomography/computed tomography and positron emission tomography (PET) alone on expected management of patients with cancer: initial results from the National Oncologic PET Registry. J Clin Oncol. 2008;26(13):2155–161.

https://doi.org/10.1200/JCO.2007.14.5631 DOI: https://doi.org/10.1200/JCO.2007.14.5631

Jensen K, Friborg J, Hansen CR, Johansen J, Kristensen CA, Andersen E, et al. The Danish Head and Neck Cancer Group (DAHANCA) 2020 radiotherapy guidelines. Radiother Oncol. 2020;151:149–51.

https://doi.org/10.1016/j.radonc.2020.07.037 DOI: https://doi.org/10.1016/j.radonc.2020.07.037

Rasmussen JH, Vogelius IR, Fischer BM, Friborg J, Aznar MC, Persson GF, et al. Prognostic value of 18F-fludeoxyglucose uptake in 287 patients with head and neck squamous cell carcinoma. Head Neck. 2015;37(9):1274–81.

https://doi.org/10.1002/hed.23745 DOI: https://doi.org/10.1002/hed.23745

Conway DI, Petticrew M, Marlborough H, Berthiller J, Hashibe M, Macpherson LM. Socioeconomic inequalities and oral cancer risk: a systematic review and meta-analysis of case-control studies. Int J Cancer. 2008;122(12):2811–9.

https://doi.org/10.1002/ijc.23430 DOI: https://doi.org/10.1002/ijc.23430

De Felice F, Lei M, Oakley R, Lyons A, Fry A, Jeannon JP, et al. Risk stratified follow-up for head and neck cancer patients – an evidence-based proposal. Oral Oncol. 2021;119:105365.

https://doi.org/10.1016/j.oraloncology.2021.105365 DOI: https://doi.org/10.1016/j.oraloncology.2021.105365

Strauss SB, Aiken AH, Lantos JE, Phillips CD. Best practices: application of NI-RADS for posttreatment surveillance imaging of head and neck cancer. Am J Roentgenol. 2021;216(6):1438–51.

https://doi.org/10.2214/AJR.20.23841 DOI: https://doi.org/10.2214/AJR.20.23841

Eberly HW, Sciscent BY, Lorenz FJ, Rettig EM, Goyal N. Current and emerging diagnostic, prognostic, and predictive biomarkers in head and neck cancer. Biomedicines. 2024;12(2):415.

https://doi.org/10.3390/biomedicines12020415 DOI: https://doi.org/10.3390/biomedicines12020415

Paidpally V, Chirindel A, Lam S, Agrawal N, Quon H, Subramaniam RM. FDG-PET/CT imaging biomarkers in head and neck squamous cell carcinoma. Imaging Med. 2012;4(6):633–47.

https://doi.org/10.2217/iim.12.60 DOI: https://doi.org/10.2217/iim.12.60

Leclerc M, Lartigau E, Lacornerie T, Daisne JF, Kramar A, Grégoire V. Primary tumor delineation based on 18F-FDG PET for locally advanced head and neck cancer treated by chemo-radiotherapy. Radiother Oncol. 2015;116(1):87–93.

https://doi.org/10.1016/j.radonc.2015.06.007 DOI: https://doi.org/10.1016/j.radonc.2015.06.007

Håkansson K, Rasmussen JH, Rasmussen GB, Friborg J, Gerds TA, Fischer BM, et al. A failure-type specific risk prediction tool for selection of head-and-neck cancer patients for experimental treatments. Oral Oncol. 2017;74:77–82.

https://doi.org/10.1016/j.oraloncology.2017.09.018 DOI: https://doi.org/10.1016/j.oraloncology.2017.09.018

Kim SA, Roh JL, Kim JS, Lee JH, Lee SH, Choi SH, et al. 18F-FDG PET/CT surveillance for the detection of recurrence in patients with head and neck cancer. Eur J Cancer. 2017;72:62–70.

https://doi.org/10.1016/j.ejca.2016.11.009 DOI: https://doi.org/10.1016/j.ejca.2016.11.009

Abgral R, Querellou S, Potard G, Le Roux PY, Le Duc-Pennec A, Marianovski R, et al. Does 18F-FDG PET/CT improve the detection of posttreatment recurrence of head and neck squamous cell carcinoma in patients negative for disease on clinical follow-up? J Nucl Med. 2009;50:24–9.

https://doi.org/10.2967/jnumed.108.055806 DOI: https://doi.org/10.2967/jnumed.108.055806

van der Veen J, Gulyban A, Nuyts S. Interobserver variability in delineation of target volumes in head and neck cancer. Radiother Oncol. 2019;137:9–15.

https://doi.org/10.1016/j.radonc.2019.04.006 DOI: https://doi.org/10.1016/j.radonc.2019.04.006

Litjens G, Kooi T, Bejnordi BE, Setio AAA, Ciompi F, Ghafoorian M, et al. A survey on deep learning in medical image analysis. Med Image Anal. 2017;42:60–88.

https://doi.org/10.1016/j.media.2017.07.005 DOI: https://doi.org/10.1016/j.media.2017.07.005

Liu X, Faes L, Kale AU, Wagner SK, Fu DJ, Bruynseels A, et al. A comparison of deep learning performance against health-care professionals in detecting diseases from medical imaging: a systematic review and meta-analysis. Lancet Digit Health 2019;1(6):e271–97.

https://doi.org/10.1016/S2589-7500(19)30123-2 DOI: https://doi.org/10.1016/S2589-7500(19)30123-2

Isensee F, Jaeger PF, Kohl SAA, Petersen J, Maier-Hein KH. nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nat Methods. 2021;18(2):203–11.

https://doi.org/10.1038/s41592-020-01008-z DOI: https://doi.org/10.1038/s41592-020-01008-z

Ren J, Eriksen JG, Nijkamp J, Korreman SS. Comparing different CT, PET and MRI multi-modality image combinations for deep learning-based head and neck tumor segmentation. Acta Oncol. 2021;60(11):1399–406.

https://doi.org/10.1080/0284186X.2021.1949034 DOI: https://doi.org/10.1080/0284186X.2021.1949034

Kovacs DG, Ladefoged CN, Andersen KF, Brittain JM, Christensen CB, Dejanovic D, et al. Clinical evaluation of deep learning for tumor delineation on 18F-FDG PET/CT of head and neck cancer. J Nucl Med. 2024;65(4):623–9.

https://doi.org/10.2967/jnumed.123.266574 DOI: https://doi.org/10.2967/jnumed.123.266574

Andrearczyk V, Oreiller V, Abobakr M, Akhavanallaf A, Balermpas P, Boughdad S, et al. Overview of the HECKTOR Challenge at MICCAI 2022: automatic head and neck tumor segmentation and outcome prediction in PET/CT. Head Neck Tumor Challenge 2022. 2023;13626:1–30.

https://doi.org/10.1007/978-3-031-27420-6_1 DOI: https://doi.org/10.1007/978-3-031-27420-6_1

Moe YM, Groendahl AR, Tomic O, Dale E, Malinen E, Futsaether CM. Deep learning-based auto-delineation of gross tumour volumes and involved nodes in PET/CT images of head and neck cancer patients. Eur J Nucl Med Mol Imaging. 2021;48(9):2782–92.

https://doi.org/10.1007/s00259-020-05125-x DOI: https://doi.org/10.1007/s00259-020-05125-x

Thompson RF, Valdes G, Fuller CD, Carpenter CM, Morin O, Aneja S, et al. Artificial intelligence in radiation oncology: a specialty-wide disruptive transformation? Radiother Oncol. 2018;129(3):421–6.

https://doi.org/10.1016/j.radonc.2018.05.030 DOI: https://doi.org/10.1016/j.radonc.2018.05.030

Marschner SN, Jensen C, Tylén P, Al-Farra G, Jakobsen JB, Jensen AR, et al. Risk stratification using 18F-FDG PET/CT and artificial neural networks in head and neck cancer patients undergoing radiotherapy. Cancers (Basel). 2021;13(18):4576. DOI: https://doi.org/10.3390/diagnostics11091581

https://doi.org/10.3390/cancers13184576 DOI: https://doi.org/10.3390/cancers13184576

Collins GS, Moons KGM, Dhiman P, Riley RD, Beam AL, Van Calster B, et al. TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods. BMJ. 2024;385:e078378.

https://doi.org/10.1136/bmj-2023-078378 DOI: https://doi.org/10.1136/bmj-2023-078378

Kovacs DG, Ladefoged CN, Rosenkjær JB, Mawassi F, Hingelberg LA, Laursen AOA, et al. Is the clinical implementation of in-house artificial intelligence-developed algorithms happening? J Nucl Med. 2025;66(2):183–5.

https://doi.org/10.2967/jnumed.124.268156 DOI: https://doi.org/10.2967/jnumed.124.268156

Ronneberger O, Fischer P, Brox T. U-Net: convolutional networks for biomedical image segmentation. In: Navab N, Hornegger J, Wells WM, Frangi AF, editors. Medical image computing and computer-assisted intervention – MICCAI 2015. Cham: Springer; 2015. p. 234–41. DOI: https://doi.org/10.1007/978-3-319-24574-4_28

Kamnitsas K, Ledig C, Newcombe VFJ, Simpson JP, Kane AD, Menon DK, et al. Efficient multi-scale 3D CNN with fully connected CRF for accurate brain lesion segmentation. Med Image Anal. 2017;36:61–78.

https://doi.org/10.1016/j.media.2016.10.004 DOI: https://doi.org/10.1016/j.media.2016.10.004

Oktay O, Nanavati J, Schwaighofer A, Carter D, Bristow M, Tanno R, et al. Evaluation of deep learning to augment image-guided radiotherapy for head and neck and prostate cancers. JAMA Netw Open. 2020;3:e2027426.

https://doi.org/10.1001/jamanetworkopen.2020.27426 DOI: https://doi.org/10.1001/jamanetworkopen.2020.27426

Hatamizadeh A, Nath V, Tang Y, Yang D, Roth H, Xu D. Swin UNETR: swin transformers for semantic segmentation of brain tumors in MRI images. arXiv. 2022. DOI: https://doi.org/10.1007/978-3-031-08999-2_22

https://doi.org/10.48550/arXiv.2201.01266

Turečková A, Tureček T, Komínková Oplatková Z, Rodríguez-Sánchez A. Improving CT image tumor segmentation through deep supervision and attentional gates. Front Robot AI. 2020;7:106.

https://doi.org/10.3389/frobt.2020.00106 DOI: https://doi.org/10.3389/frobt.2020.00106

Oken MM, Creech RH, Tormey DC, Horton J, Davis TE, McFadden ET, et al. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol. 1982;5:649–55.

https://doi.org/10.1097/00000421-198212000-00014 DOI: https://doi.org/10.1097/00000421-198212000-00014

Piccirillo JF, Tierney RM, Costas I, Grove L, Spitznagel EL Jr, et al. Prognostic importance of comorbidity in a hospital-based c­ancer registry. JAMA. 2004;291:2441–7.

https://doi.org/10.1001/jama.291.20.2441 DOI: https://doi.org/10.1001/jama.291.20.2441

Thomson D, Yang H, Baines H, Miles E, Bolton S, West C, et al. NIMRAD – a phase III trial to investigate the use of nimorazole hypoxia modification with intensity-modulated radiotherapy in head and neck cancer. Clin Oncol (R Coll Radiol). 2014;26:344–7.

https://doi.org/10.1016/j.clon.2014.03.003 DOI: https://doi.org/10.1016/j.clon.2014.03.003

Marcu DC, Grava C, Marcu LG. Current role of delta radiomics in head and neck oncology. Int J Mol Sci. 2023;24(3):2214.

https://doi.org/10.3390/ijms24032214 DOI: https://doi.org/10.3390/ijms24032214

van Timmeren JE, Bussink J, Koopmans P, Smeenk RJ, Monshouwer R. Longitudinal image data for outcome modeling. Clin Oncol (R Coll Radiol). 2025;38:103610.

https://doi.org/10.1016/j.clon.2024.06.053 DOI: https://doi.org/10.1016/j.clon.2024.06.053

Rokuss M, Kirchhoff Y, Akbal S, Kovacs B, Roy S, Ulrich C, et al. LesionLocator: zero-shot universal tumor segmentation and tracking in 3D whole-body imaging [Internet]. arXiv preprint arXiv:2502.20985 [cs.CV]; 2025. Available from: https://arxiv.org/abs/2502.20985 [Cited date: 23 July 2025] DOI: https://doi.org/10.1109/CVPR52734.2025.02875

Kelly CJ, Karthikesalingam A, Suleyman M, Corrado G, King D. Key challenges for delivering clinical impact with artificial intelligence. BMC Med. 2019;17:195.

https://doi.org/10.1186/s12916-019-1426-2 DOI: https://doi.org/10.1186/s12916-019-1426-2

Downloads

Additional Files

Published

2025-08-30

How to Cite

Kovacs, D. G., Aznar, M., van Herk, M., Mohamed, I., Price, J., Ladefoged, C. N., … Abravan, A. (2025). External validation of deep learning-derived 18F-FDG PET/CT delta biomarkers for loco-regional control in head and neck cancer. Acta Oncologica, 64, 1143–1151. https://doi.org/10.2340/1651-226X.2025.43977

Issue

Vol. 64 (2025): Acta Oncologica

Section

Original article

Categories

License

Copyright (c) 2025 David Gergely Kovacs, Marianne Aznar, Marcel van Herk, Iskandar Mohamed, James Price, Claes Nøhr Ladefoged, Barbara Malene Fischer, Flemming Littrup Andersen, Andrew McPartlin, Eliana M. Vasquez Osorio, Azadeh Abravan

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.