DOI QR코드

DOI QR Code

Development and Testing of a Machine Learning Model Using 18F-Fluorodeoxyglucose PET/CT-Derived Metabolic Parameters to Classify Human Papillomavirus Status in Oropharyngeal Squamous Carcinoma

  • Changsoo Woo (Department of Radiology and Research Institute of Radiological Science and Center for Clinical Imaging Data Science, Severance Hospital, Yonsei University College of Medicine) ;
  • Kwan Hyeong Jo (Department of Nuclear Medicine, Korea University Guro Hospital) ;
  • Beomseok Sohn (Department of Radiology and Research Institute of Radiological Science and Center for Clinical Imaging Data Science, Severance Hospital, Yonsei University College of Medicine) ;
  • Kisung Park (Department of Radiology and Research Institute of Radiological Science and Center for Clinical Imaging Data Science, Severance Hospital, Yonsei University College of Medicine) ;
  • Hojin Cho (Department of Nuclear Medicine, Severance Hospital, Yonsei University College of Medicine) ;
  • Won Jun Kang (Department of Nuclear Medicine, Severance Hospital, Yonsei University College of Medicine) ;
  • Jinna Kim (Department of Radiology and Research Institute of Radiological Science and Center for Clinical Imaging Data Science, Severance Hospital, Yonsei University College of Medicine) ;
  • Seung-Koo Lee (Department of Radiology and Research Institute of Radiological Science and Center for Clinical Imaging Data Science, Severance Hospital, Yonsei University College of Medicine)
  • Received : 2021.11.21
  • Accepted : 2022.10.31
  • Published : 2023.01.01

Abstract

Objective: To develop and test a machine learning model for classifying human papillomavirus (HPV) status of patients with oropharyngeal squamous cell carcinoma (OPSCC) using 18F-fluorodeoxyglucose (18F-FDG) PET-derived parameters in derived parameters and an appropriate combination of machine learning methods in patients with OPSCC. Materials and Methods: This retrospective study enrolled 126 patients (118 male; mean age, 60 years) with newly diagnosed, pathologically confirmed OPSCC, that underwent 18F-FDG PET-computed tomography (CT) between January 2012 and February 2020. Patients were randomly assigned to training and internal validation sets in a 7:3 ratio. An external test set of 19 patients (16 male; mean age, 65.3 years) was recruited sequentially from two other tertiary hospitals. Model 1 used only PET parameters, Model 2 used only clinical features, and Model 3 used both PET and clinical parameters. Multiple feature transforms, feature selection, oversampling, and training models are all investigated. The external test set was used to test the three models that performed best in the internal validation set. The values for area under the receiver operating characteristic curve (AUC) were compared between models. Results: In the external test set, ExtraTrees-based Model 3, which uses two PET-derived parameters and three clinical features, with a combination of MinMaxScaler, mutual information selection, and adaptive synthetic sampling approach, showed the best performance (AUC = 0.78; 95% confidence interval, 0.46-1). Model 3 outperformed Model 1 using PET parameters alone (AUC = 0.48, p = 0.047) and Model 2 using clinical parameters alone (AUC = 0.52, p = 0.142) in predicting HPV status. Conclusion: Using oversampling and mutual information selection, an ExtraTree-based HPV status classifier was developed by combining metabolic parameters derived from 18F-FDG PET/CT and clinical parameters in OPSCC, which exhibited higher performance than the models using either PET or clinical parameters alone.

Keywords

Acknowledgement

This study was supported by a faculty research grant of Yonsei University College of Medicine for (6-2021-0151).

References

  1. McDermott JD, Bowles DW. Epidemiology of head and neck squamous cell carcinomas: impact on staging and prevention strategies. Curr Treat Options Oncol 2019;20:43
  2. Tumban E. A current update on human papillomavirus-associated head and neck cancers. Viruses 2019;11:922
  3. Johnson DE, Burtness B, Leemans CR, Lui VWY, Bauman JE, Grandis JR. Head and neck squamous cell carcinoma. Nat Rev Dis Primers 2020;6:92
  4. Taberna M, Mena M, Pavon MA, Alemany L, Gillison ML, Mesia R. Human papillomavirus-related oropharyngeal cancer. Ann Oncol 2017;28:2386-2398 https://doi.org/10.1093/annonc/mdx304
  5. Amin MB, Greene FL, Edge SB, Compton CC, Gershenwald JE, Brookland RK, et al. AJCC cancer staging manual, 8th ed. New York: Springer, 2017
  6. Solomon B, Young RJ, Rischin D. Head and neck squamous cell carcinoma: genomics and emerging biomarkers for immunomodulatory cancer treatments. Semin Cancer Biol 2018;52(Pt 2):228-240 https://doi.org/10.1016/j.semcancer.2018.01.008
  7. You EL, Henry M, Zeitouni AG. Human papillomavirus-associated oropharyngeal cancer: review of current evidence and management. Curr Oncol 2019;26:119-123 https://doi.org/10.3747/co.26.4819
  8. Benson E, Li R, Eisele D, Fakhry C. The clinical impact of HPV tumor status upon head and neck squamous cell carcinomas. Oral Oncol 2014;50:565-574 https://doi.org/10.1016/j.oraloncology.2013.09.008
  9. Spadarella G, Ugga L, Calareso G, Villa R, D'Aniello S, Cuocolo R. The impact of radiomics for human papillomavirus status prediction in oropharyngeal cancer: systematic review and radiomics quality score assessment. Neuroradiology 2022;64:1639-1647 https://doi.org/10.1007/s00234-022-02959-0
  10. Ravanelli M, Grammatica A, Tononcelli E, Morello R, Leali M, Battocchio S, et al. Correlation between human papillomavirus status and quantitative MR imaging parameters including diffusion-weighted imaging and texture features in oropharyngeal carcinoma. AJNR Am J Neuroradiol 2018;39:1878-1883 https://doi.org/10.3174/ajnr.A5792
  11. Sohn B, Choi YS, Ahn SS, Kim H, Han K, Lee SK, et al. Machine learning based radiomic HPV phenotyping of oropharyngeal SCC: a feasibility study using MRI. Laryngoscope 2021;131:E851-E856 https://doi.org/10.1002/lary.28889
  12. Suh CH, Lee KH, Choi YJ, Chung SR, Baek JH, Lee JH, et al. Oropharyngeal squamous cell carcinoma: radiomic machine-learning classifiers from multiparametric MR images for determination of HPV infection status. Sci Rep 2020;10:17525
  13. Haider SP, Mahajan A, Zeevi T, Baumeister P, Reichel C, Sharaf K, et al. PET/CT radiomics signature of human papilloma virus association in oropharyngeal squamous cell carcinoma. Eur J Nucl Med Mol Imaging 2020;47:2978-2991 https://doi.org/10.1007/s00259-020-04839-2
  14. Gillies RJ, Kinahan PE, Hricak H. Radiomics: images are more than pictures, they are data. Radiology 2016;278:563-577 https://doi.org/10.1148/radiol.2015151169
  15. Boellaard R, Delgado-Bolton R, Oyen WJ, Giammarile F, Tatsch K, Eschner W, et al. FDG PET/CT: EANM procedure guidelines for tumour imaging: version 2.0. Eur J Nucl Med Mol Imaging 2015;42:328-354 https://doi.org/10.1007/s00259-014-2961-x
  16. Begum S, Cao D, Gillison M, Zahurak M, Westra WH. Tissue distribution of human papillomavirus 16 DNA integration in patients with tonsillar carcinoma. Clin Cancer Res 2005;11:5694-5699 https://doi.org/10.1158/1078-0432.CCR-05-0587
  17. Pope P, Webster J. The use of an F-statistic in stepwise regression procedures. Technometrics 1972;14:327-340 https://doi.org/10.2307/1267425
  18. Lunardon N, Menardi G, Torelli N. ROSE: a package for binary imbalanced learning. R Journal 2014;6:79-89 https://doi.org/10.32614/RJ-2014-008
  19. He H, Bai Y, Garcia EA, Li S. ADASYN: adaptive synthetic sampling approach for imbalanced learning. Proceedings of the International Joint Conference on Neural Networks, IJCNN 2008, part of the IEEE World Congress on Computational Intelligence, WCCI 2008; 2008 Jun 1-6; Hong Kong: IEEE; 2008:1322-1328
  20. Han H, Wang WY, Mao BH. Borderline-SMOTE: a new oversampling method in imbalanced data sets learning. In: Huang DS, Zhang XP, Huang GB, eds. Advances in intelligent computing. ICIC 2005. Lecture notes in computer science, vol 3644. Berlin: Springer, 2005:878-887
  21. Chawla NV, Bowyer KW, Hall LO, Kegelmeyer WP. SMOTE: synthetic minority over-sampling technique. J Artif Intell Res 2002;16:321-357 https://doi.org/10.1613/jair.953
  22. Simm J, De Abril IM, Sugiyama M. Tree-based ensemble multitask learning method for classification and regression. IEICE Trans Inf Syst 2014;97:1677-1681 https://doi.org/10.1587/transinf.E97.D.1677
  23. Schapire RE. Explaining AdaBoost. In: Scholkopf B, Luo Z, Vovk V, eds. Empirical inference. Berlin, Heidelberg: Springer, 2013
  24. Ke G, Meng Q, Finley T, Wang T, Chen W, Ma W, et al. LightGBM: a highly efficient gradient boosting decision tree. Proceedings of the 31st International Conference on Neural Information Processing Systems; 2017 Dec 4-9; Long Beach, CA, USA: NIPS; 2017:3149-3157
  25. Chen T, Guestrin C. Xgboost: a scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining; 2016 Aug 13-17; San Francisco, CA, USA: KDD; 2016:785-794
  26. Lundberg SM, Lee SI. A unified approach to interpreting model predictions. Proceedings of the 31st International Conference on Neural Information Processing Systems; 2017 Dec 4-9; Long Beach, CA, USA: NIPS; 2017:4768-4777
  27. D'Souza G, Kreimer AR, Viscidi R, Pawlita M, Fakhry C, Koch WM, et al. Case-control study of human papillomavirus and oropharyngeal cancer. N Engl J Med 2007;356:1944-1956 https://doi.org/10.1056/NEJMoa065497
  28. Lechner M, Liu J, Masterson L, Fenton TR. HPV-associated oropharyngeal cancer: epidemiology, molecular biology and clinical management. Nat Rev Clin Oncol 2022;19:306-327 https://doi.org/10.1038/s41571-022-00603-7
  29. Schache AG, Powell NG, Cuschieri KS, Robinson M, Leary S, Mehanna H, et al. HPV-related oropharynx cancer in the United Kingdom: an evolution in the understanding of disease etiology. Cancer Res 2016;76:6598-6606 https://doi.org/10.1158/0008-5472.CAN-16-0633
  30. Chaturvedi AK, Zumsteg ZS. A snapshot of the evolving epidemiology of oropharynx cancers. Cancer 2018;124:2893-2896 https://doi.org/10.1002/cncr.31383
  31. Zumsteg ZS, Cook-Wiens G, Yoshida E, Shiao SL, Lee NY, Mita A, et al. Incidence of oropharyngeal cancer among elderly patients in the United States. JAMA Oncol 2016;2:1617-1623 https://doi.org/10.1001/jamaoncol.2016.1804
  32. Haeggblom L, Ramqvist T, Tommasino M, Dalianis T, Nasman A. Time to change perspectives on HPV in oropharyngeal cancer. A systematic review of HPV prevalence per oropharyngeal subsite the last 3 years. Papillomavirus Res 2017;4:1-11 https://doi.org/10.1016/j.pvr.2017.05.002
  33. Pfister DG, Ang KK, Brizel DM, Burtness BA, Busse PM, Caudell JJ, et al. Head and neck cancers, version 2.2013. Featured updates to the NCCN guidelines. J Natl Compr Canc Netw 2013;11:917-923 https://doi.org/10.6004/jnccn.2013.0113
  34. Sohn B, Koh YW, Kang WJ, Lee JH, Shin NY, Kim J. Is there an additive value of 18 F-FDG PET-CT to CT/MRI for detecting nodal metastasis in oropharyngeal squamous cell carcinoma patients with palpably negative neck? Acta Radiol 2016;57:1352-1359 https://doi.org/10.1177/0284185115587544
  35. Kim SJ, Pak K, Kim K. Diagnostic accuracy of F-18 FDG PET or PET/CT for detection of lymph node metastasis in clinically node negative head and neck cancer patients; a systematic review and meta-analysis. Am J Otolaryngol 2019;40:297-305 https://doi.org/10.1016/j.amjoto.2018.10.013
  36. Freihat O, Toth Z, Pinter T, Kedves A, Sipos D, Cselik Z, et al. Pre-treatment PET/MRI based FDG and DWI imaging parameters for predicting HPV status and tumor response to chemoradiotherapy in primary oropharyngeal squamous cell carcinoma (OPSCC). Oral Oncol 2021;116:105239
  37. Bose P, Brockton NT, Dort JC. Head and neck cancer: from anatomy to biology. Int J Cancer 2013;133:2013-2023 https://doi.org/10.1002/ijc.28112
  38. Wijsman R, Kaanders JH, Oyen WJ, Bussink J. Hypoxia and tumor metabolism in radiation oncology: targets visualized by positron emission tomography. Q J Nucl Med Mol Imaging 2013;57:244-256
  39. Harshani JM, Yeluri S, Guttikonda VR. Glut-1 as a prognostic biomarker in oral squamous cell carcinoma. J Oral Maxillofac Pathol 2014;18:372-378
  40. Botha H, Farah CS, Koo K, Cirillo N, McCullough M, Paolini R, et al. The role of glucose transporters in oral squamous cell carcinoma. Biomolecules 2021;11:1070
  41. Baschnagel AM, Wobb JL, Dilworth JT, Williams L, Eskandari M, Wu D, et al. The association of (18)F-FDG PET and glucose metabolism biomarkers GLUT1 and HK2 in p16 positive and negative head and neck squamous cell carcinomas. Radiother Oncol 2015;117:118-124
  42. Krupar R, Robold K, Gaag D, Spanier G, Kreutz M, Renner K, et al. Immunologic and metabolic characteristics of HPV-negative and HPV-positive head and neck squamous cell carcinomas are strikingly different. Virchows Arch 2014;465:299-312 https://doi.org/10.1007/s00428-014-1630-6
  43. Ribbat-Idel J, Perner S, Kuppler P, Klapper L, Krupar R, Watermann C, et al. Immunologic "cold" squamous cell carcinomas of the head and neck are associated with an unfavorable prognosis. Front Med (Lausanne) 2021;8:622330
  44. Keramida G, Dizdarevic S, Bush J, Peters AM. Quantification of tumour (18) F-FDG uptake: normalise to blood glucose or scale to liver uptake? Eur Radiol 2015;25:2701-2708 https://doi.org/10.1007/s00330-015-3659-6
  45. Chaturvedi A, Engels E, Pfeiffer R, Hernandez B, Xiao W, Kim E, et al. Human papillomavirus (HPV) and rising oropharyngeal cancer incidence and survival in the United States. J Clin Oncol 2011;29:5529
  46. Damgacioglu H, Sonawane K, Zhu Y, Li R, Balasubramanian BA, Lairson DR, et al. Oropharyngeal cancer incidence and mortality trends in all 50 states in the US, 2001-2017. JAMA Otolaryngol Head Neck Surg 2022;148:155-165 https://doi.org/10.1001/jamaoto.2021.3567
  47. Shinn JR, Davis SJ, Lang-Kuhs KA, Rohde S, Wang X, Liu P, et al. Oropharyngeal squamous cell carcinoma with discordant p16 and HPV mRNA results: incidence and characterization in a large, contemporary United States cohort. Am J Surg Pathol 2021;45:951-961 https://doi.org/10.1097/PAS.0000000000001685
  48. Arsa L, Siripoon T, Trachu N, Foyhirun S, Pangpunyakulchai D, Sanpapant S, et al. Discrepancy in p16 expression in patients with HPV-associated head and neck squamous cell carcinoma in Thailand: clinical characteristics and survival outcomes. BMC Cancer 2021;21:504