The prognostic value of CT radiomic features from primary tumours and pathological lymph
nodes in head and neck cancer patients
Zhai, Tiantian
DOI:
10.33612/diss.111448998
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2020
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Zhai, T. (2020). The prognostic value of CT radiomic features from primary tumours and pathological lymph nodes in head and neck cancer patients. University of Groningen. https://doi.org/10.33612/diss.111448998
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
CHAPTER 1
1.1 Introduction
Head and neck cancer (HNC) originating from the oral cavity, larynx and pharynx is responsible for about 0.83 million new cancer cases and 0.43 million cancer deaths worldwide every year. These are predominantly squamous cell carcinoma (SCC) [1,2]. Much progress has been made in the treatment of HNC in the last two decades. The introduction of chemotherapy, better understanding of human papillomavirus (HPV)-related oropharynx SCC, and multidisciplinary management, has resulted in improved overall survival (OS) rates [3–5].
Based on data from the Surveillance, Epidemiology and End Results (SEER) program and the Netherlands Cancer Registry, the 5-year OS rate for HNC patients is approximately 60%. For patients with early stage disease, the 5-year OS rate can even reach 70-90% [6]. However, 30%-50% of patients with locally advanced head and neck squamous cell carcinoma (HNSCC) still experience treatment failures, predominantly occurring at the site of the primary tumour, followed by regional failures and distant metastases [7]. The prolonged life expectancy of HNC patients has consequently increased the number of patients with acute and late toxicity following treatment. To enable more personalised treatment approaches for HNC patients, there is a rising demand for adequate prediction of treatment failure, as well as complications [5,8–11].
Normal tissue complication probability (NTCP) models and tumour control probability (TCP) models are used for the prediction of treatment outcome, this can mean treatment-related side effects and/or tumour control. For HNSCC patients, the most common treatment-related side effects are xerostomia, dysphagia and tube feeding dependence, while the main treatment failures are local failure, regional failure, distant metastasis and death. The current NTCP models of the most frequently reported side effects and TCP models of the main treatment failures of HNSCC are mainly based on classic prognostic factors such as tumour stage, performance status, age, baseline toxicity scores, dose-volume parameters, etc. [7,10,12–15]. However, patients with similar prognostic factors may still have different outcomes. In other words, these TCP and NTCP models need to be further improved before they can be used for future personalised medicine [16]. New emerging data in the form of radiomics, also called image biomarkers (IBM), reflect
11
General introduction
1
the intensity, shape and textural heterogeneity of the region of interest derived from medical images. Such imaging data has shown significant associations with survival and complications in HNSCC patients [17,18]. An important question that has remained unanswered in previous publications is about the the extent to which the addition of these radiomic features improves the predictive power of models consisting only of classical prognostic factors, such as TNM staging, performance status and baseline toxicity scores.
The additional role of radiomic features in predicting radiation-induced toxicities for HNSCC patients has been explored in a thesis by LV van Dijk (UMC Groningen, 2018) and associated publications [18–20]. This thesis focused on improving the prediction of treatment failures.
The overarching aim of this thesis was to evaluate the prognostic ability of radiomic features and to test whether the performance of prediction models for different treatment failures could be improved by the addition of radiomics to the classical prognostic factors for HNSCC patients primarily treated with radiotherapy.
1.2 Candidate features
Classic prognostic factors such as clinical and biological factors have been shown to influence treatment response, and new emerging radiomics (image biomarkers) have also demonstrated their prognostic values in a series of studies [21–27]. All of them are included as candidate parameters in the analyses described within thesis.
Clinical features
Currently, clinical factors are most frequently used in routine clinical practice to estimate local control (LC), regional control (RC), distant metastasis-free survival (DMFS), disease-free survival (DFS) and overall survival (OS) rates of individual patients. The clinical factors can be categorised as follows: (a) tumour-related and (b) patient-related prognostic factors.
TNM classification is the most important tumour-related feature and has been used in routine clinical practice to guide treatment decision-making [8,22,28]. Generally, systemic therapy is recommended for patients with advanced TNM stage to improve tumour control. Other similar features such as tumour volume, tumour diameter and overall
stage have also frequently been reported [21,22,29,30]. Besides TNM classification, large variations in OS and LC can be found in patients with HNSCC from different tumour subsites [28]. Despite being centimeters apart, nasopharyngeal and laryngeal cancer have a higher radio-sensitivities than oral cavity and hypopharynx cancer [7,10,25,31]. In addition to tumour-related prognostic factors, patient-related features such as high alcohol intake, active smoking status, presence of co-morbidities, advanced age, poor WHO performance score, instance of weight loss and male gender all showed significant associations with worse LC, RC, DFS and OS [11,22,28,32]. These factors are currently taken into consideration when treatment decisions are made. For example, patients with a poor WHO performance score who are older than 70 years are not considered candidates for systemic treatment, while surgery is not preferred for patients with severe co-morbidities [8].
Histopathological/molecular biological features
Many histopathological and molecular biological markers that are associated with prognosis have been previously studied and reported. Extranodal extension (ENE) and p16 are the most important biomarkers identified for HNSCC. These have therefore been included in the AJCC 8th TNM classification to guide treatment decisions [5,8,33–35]. Additionally, the over-expression of epidermal growth factor receptor (EGFR) showed an association with a poor response to radiotherapy. Also, the concurrent use of anti-EGFR monoclonal antibody cetuximab with radiotherapy has been shown superior to radiotherapy alone, with improved survival [36]. There are many other potential biological factors, such as gene mutations related to hypoxia, intrinsic radio-sensitivity and apoptosis, that are likely to play a role in determining a patient’s radiation response. However, none have yet entered routine clinical practice [22,37].
A main concern with histopathological/molecular biological features is that the tests are not always available and feasible in the clinic. For patients treated with primary non-surgical modalities (the patient population discussed in this thesis), such as (chemo) radiotherapy, only limited pathological information is available. Furthermore, tests based on the limited biopsy specimen might not be reliable and representative for the whole tumour [22,34].
13
General introduction
1
Radiomics, or IBM, refers to the comprehensive quantification of tumour phenotypes based on medical images. A wide variety of medical images is already available for diagnostic, staging and radiotherapy treatment planning purposes. Radiomic features can be extracted from these medical imaging modalities (computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET) or ultrasound) without the need for additional image acquisition [38,39]. For HNSCC patients primarily treated with radiotherapy (RT) at the University Medical Centre Groningen (UMCG), a pre-treatment planning CT scan is always acquired and saved for every patient. Therefore a large dataset is available to study CT-based radiomic features. Compared with the histopathological/molecular biological features based on the limited and invasive needle biopsy specimens, radiomics is able to capture the phenotype of the whole tumour non-invasively [38]. Moreover, radiomic features are objectively calculated by using standard formulas, providing quantitative information regarding intensity, shape and textural characteristics of a region of interest. This transforms three-dimensional morphological tumor information into multi-dimensional and mineable data [16,17,40–44]. Together, the advantages of radiomic features make them very promising candidates for the prediction of patient-specific treatment outcome.
1.3 Study cohort
It is important to study prognostic factors in a well-defined patient group treated with the same modalities [12]. In this thesis, with respect to this criterion, we focused on the HNSCC patients who were primarily treated with definitive radiotherapy, which accounts for two-thirds of new HNSCC patients [45,46]. All studies in this thesis are retrospective analyses, which was composed of 707 consecutive, prospectively collected, non-surgically treated HNSCC patients from University Medical Centre Groningen (UMCG) , the Netherlands, 113 HNSCC patients from department of radiation oncology in Maastricht (MAASTRO), the Netherlands, and 289 nasopharyngeal cancer patients from Shantou University Medical College (SUMC), China with a standardised follow-up programme [47].
The first part of this thesis (Chapter 2 and 3) focussed on improving the prediction of
different prognostic endpoints with CT-based radiomic features.
We tested the hypothesis that the prediction of overall survival (OS) could be improved by adding radiomic features to clinical features in Chapter 2. Furthermore, the ability to
generalise the prognostic value of radiomics for different tumour types was investigated by training a model on nasopharyngeal cancer patients and externally validating this model on other HNC sub-types.
With the knowledge gained in Chapter 2, we subsequently tested the performance of radiomic features in a systematic and thorough analysis using local control (LC), regional control (RC), distant metastasis-free survival (DMFS) and disease-free survival (DFS) rate as endpoints in Chapter 3.
We expected that prognostic models including radiomic and clinical features could be used to identify patients with a high risk of treatment failures prior to treatment, and consequently could support more effective personalised treatment approaches.
The second part of this thesis (Chapter 4 and Chapter 5) focused on advanced HNSCC
patients with pathological lymph nodes (N+). Instead of identifying patients who are at high risk of treatment failures, in this study we tried to estimate the failure risk of each individual pathological lymph node. If we can identify those lymph nodes with high failure risk prior to treatment, intensified radiation schedules could be applied to the specific high-failure risk node to improve nodal control without increasing the dose to normal tissue. Alternatively, pre-treatment lymph node-targeted dissection could be arranged to avoid severe post-operative complications.
Chapter 4 explored the predictive clinical and radiomic features that could be used to
identify pathological lymph nodes that have a large risk of persistence or recurrence. A pre-treatment prediction model for nodal failures was developed and internally validated in Chapter 4. Before this model is introduced into the clinic, a TRIPOD (transparent
reporting of a multivariable prediction model for individual prognosis or diagnosis) type 4 level model validation [48] would be required. Therefore, we validated the model in an external cohort of HNSCC patients from the MAASTRO clinic in Chapter 5.
The findings of this thesis are summarised and future perspectives are discussed in
15
General introduction
1
REFERENCES
1. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018;68:394–424.
2. Baxi S, Fury M, Ganly I, Rao S, Pfister DG. Ten years of progress in head and neck cancers. JNCCN J Natl Compr Cancer Netw 2012;10:806–10.
3. Pignon JP, Maître A le, Maillard E, Bourhis J. Meta-analysis of chemotherapy in head and neck cancer (MACH-NC): An update on 93 randomised trials and 17,346 patients. Radiother Oncol 2009;92:4–14.
4. Blanchard P, Baujat B, Holostenco V, Bourredjem A, Baey C, Bourhis J, et al. Meta-analysis of chemotherapy in head and neck cancer (MACH-NC): A comprehensive analysis by tumour site. Radiother Oncol 2011;100:33–40.
5. Ang KK, Harris J, Wheeler R, Weber R, Rosenthal DI, Nguyen-Tân PF, et al. Human papillomavirus and survival of patients with oropharyngeal cancer. N Engl J Med 2010;363:24–35.
6. Howlader N, Noone AM, Krapcho M, et al. SEER cancer statistics review, 1975-2013, based on November 2015 SEER. http://seer.cancer.gov/csr/1975_2013/.
7. Pagh A, Grau C, Overgaard J. Failure pattern and salvage treatment after radical treatment of head and neck cancer. Acta Oncol (Madr) 2016;55:625–32.
8. Pfister DG, Spencer SA, Brizel DM, Burtness BA, Busse PM, Caudell JJ, et al. Head and neck cancers, version 1.2015 featured updates to the NCCN guidelines. JNCCN J Natl Compr Cancer Netw 2015;13:847–56.
9. Liu MT, Chang TH, Lin JP, Huang CC, Wang AY. Prognostic factors affecting the outcome of nasopharyngeal carcinoma. Jpn J Clin Oncol 2003;33:501–508.
10. Regueiro CA, Aragón G, Millán L, Valcárcel FJ, de la Torre A, Magallón R. Prognostic factors for local control, regional control and survival in oropharyngeal squamous cell carcinoma. Eur J Cancer 1994;30:2060–7.
11. Langius JAE, Bakker S, Rietveld DHF, Kruizenga HM, Langendijk JA, Weijs PJM, et al. Critical weight loss is a major prognostic indicator for disease-specific survival in patients with head and neck cancer receiving radiotherapy. Br J Cancer 2013;109:1093-1099.
12. van den Brekel MWM, Bindels EMJ, Balm AJM. Prognostic factors in head and neck cancer. Eur J Cancer 2002;38(8):1041-1043.
13. Beetz I, Schilstra C, Van Der Schaaf A, Van Den Heuvel ER, Doornaert P, Van Luijk P, et al. NTCP models for patient-rated xerostomia and sticky saliva after treatment with intensity modulated radiotherapy for head and neck cancer: The role of dosimetric and clinical factors. Radiother Oncol 2012;105:101–6.
14. Christianen MEMC, Schilstra C, Beetz I, Muijs CT, Chouvalova O, Burlage FR, et al. Predictive modelling for swallowing dysfunction after primary (chemo)radiation:
Results of a prospective observational study. Radiother Oncol 2012;105:107–14. 15. Wopken K, Bijl HP, Langendijk JA. Prognostic factors for tube feeding dependence after
curative (chemo-) radiation in head and neck cancer: A systematic review of literature. Radiother Oncol 2018;126:56–67.
16. Lambin P, Leijenaar RTH, Deist TM, Peerlings J, De Jong EEC, Van Timmeren J, et al. Radiomics: The bridge between medical imaging and personalized medicine. Nat Rev Clin Oncol 2017;14:749–62.
17. Aerts HJWL, Velazquez ER, Leijenaar RTH, Parmar C, Grossmann P, Cavalho S, et al. Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat Commun 2014;5:4006.
18. van Dijk L V., Brouwer CL, van der Schaaf A, Burgerhof JGM, Beukinga RJ, Langendijk JA, et al. CT image biomarkers to improve patient-specific prediction of radiation-induced xerostomia and sticky saliva. Radiother Oncol 2017;122:185–91.
19. van Dijk L V., Noordzij W, Brouwer CL, Boellaard R, Burgerhof JGM, Langendijk JA, et al. 18F-FDG PET image biomarkers improve prediction of late radiation-induced xerostomia. Radiother Oncol 2018;126:89–95.
20. van Dijk L V., Thor M, Steenbakkers RJHM, Apte A, Zhai TT, Borra R, et al. Parotid gland fat related Magnetic Resonance image biomarkers improve prediction of late radiation-induced xerostomia. Radiother Oncol 2018;128:459–66.
21. Mehanna H, West CML, Nutting C, Paleri V. Head and neck cancer - Part 2: Treatment and prognostic factors. BMJ 2010;341:721–5.
22. Silva P, Homer JJ, Slevin NJ, Musgrove BT, Sloan P, Price P, et al. Clinical and biological factors affecting response to radiotherapy in patients with head and neck cancer: A review. Clin Otolaryngol 2007;32:337–45.
23. Leoncini E, Vukovic V, Cadoni G, Pastorino R, Arzani D, Bosetti C, et al. Clinical features and prognostic factors in patients with head and neck cancer: Results from a multicentric study. Cancer Epidemiol 2015;39:367–74.
24. Xie K, Chen J, Zou J, Chen L, Gong L. Tumor volumes predict prognosis in head and neck cancer: a meta-analysis. Transl Cancer Res 2017;6:687–97.
25. Zhai TT, Langendijk JA, van Dijk L V., Halmos GB, Witjes MJH, Oosting SF, et al. The prognostic value of CT-based image-biomarkers for head and neck cancer patients treated with definitive (chemo-)radiation. Oral Oncol 2019;95:178–86.
26. Zhai TT, van Dijk L V., Huang BT, Lin ZX, Ribeiro CO, Brouwer CL, et al. Improving the prediction of overall survival for head and neck cancer patients using image biomarkers in combination with clinical parameters. Radiother Oncol 2017;124:256–62.
27. Zwanenburg A, Leger S, Vallières M, Löck S. Image biomarker standardisation initiative 2016; arxiv. org/abs/1612.07003
28. Avilés-Jurado FX, León X. Prognostic factors in head and neck squamous cell carcinoma: Comparison of CHAID decision trees technology and cox analysis. Head Neck
17
General introduction
1
2013;35:877–83.
29. Folkert MR, Oh JH, Setton J, Apte AP, Thorstad WL, Schoder H, et al. Predictive Modeling of Outcomes Following Definitive Chemoradiation Therapy for Oropharyngeal Cancer Based on FDG-PET Image Characteristics. Int J Radiat Oncol 2013;87:S3.
30. van Den Broek GB, Rasch CRN, Pameijer FA, Peter E, Van Den Brekel MWM, Tan IB, et al. Pretreatment probability model for predicting outcome after intraarterial chemoradiation for advanced head and neck carcinoma. Cancer 2004;101(8):1809-1817.
31. de Jong RJB, Hermans J, Molenaar J, Briaire JJ, Cessie S Le. Prediction of survival in patients with head and neck cancer. Head Neck 2001;2:718-724.
32. Folkert MR, Setton J, Apte AP, Grkovski M, Young RJ, Schöder H, et al. Predictive modeling of outcomes following definitive chemoradiotherapy for oropharyngeal cancer based on FDG-PET image characteristics. Phys Med Biol 2017;62(13):5327-5343.
33. Mermod M, Tolstonog G, Simon C, Monnier Y. Extracapsular spread in head and neck squamous cell carcinoma: A systematic review and meta-analysis. Oral Oncol 2016;62:60–71.
34. Chai RL, Rath TJ, Johnson JT, Ferris RL, Kubicek GJ, Duvvuri U, et al. Accuracy of computed tomography in the prediction of extracapsular spread of lymph node metastases in squamous cell carcinoma of the head and neck. JAMA Otolaryngol - Head Neck Surg 2013;139:1187–94.
35. Dayyani F, Etzel CJ, Liu M, Ho CH, Lippman SM, Tsao AS. Meta-analysis of the impact of human papillomavirus (HPV) on cancer risk and overall survival in head and neck squamous cell carcinomas (HNSCC). Head Neck Oncol 2010;2:15.
36. Bentzen SM, Atasoy BM, Daley FM, Dische S, Richman PI, Saunders MI, et al. Epidermal growth factor receptor expression in pretreatment biopsies from head and neck squamous cell carcinoma as a predictive factor for a benefit from accelerated radiation therapy in a randomized controlled trial. J Clin Oncol 2005;23:5560-5567. 37. West CML. Intrinsic radiosensitivity as a predictor of patient response to radiotherapy.
Br J Radiol 1995;68:827-837.
38. Aerts HJWL. The potential of radiomic-based phenotyping in precisionmedicine a review. JAMA Oncol 2016;2:1636–42.
39. Caudell JJ, Torres-Roca JF, Gillies RJ, Enderling H, Kim S, Rishi A, et al. The future of personalised radiotherapy for head and neck cancer. Lancet Oncol 2017;18:e266– 73.
40. Grove O, Berglund AE, Schabath MB, Aerts HJWL, Dekker A, Wang H, et al. Quantitative computed tomographic descriptors associate tumor shape complexity and intratumor heterogeneity with prognosis in lung adenocarcinoma. PLoS One
2015;10: e0118261.
41. Cui Y, Song J, Pollom E, Alagappan M, Shirato H, Chang DT, et al. Quantitative Analysis of 18F-Fluorodeoxyglucose Positron Emission Tomography Identifies Novel Prognostic Imaging Biomarkers in Locally Advanced Pancreatic Cancer Patients Treated With Stereotactic Body Radiation Therapy. Int J Radiat Oncol Biol Phys2016;96:102-9 42. Haralick RM, Dinstein I, Shanmugam K. Textural Features for Image Classification. IEEE
Trans Syst Man Cybern 1973;3:610–21.
43. Tang X. Texture information in run-length matrices. IEEE Trans Image Process 1998;7:1602-1609.
44. Thibault G, Fertil B, Navarro C, Pereira S. Texture Indexes and Gray Level Size Zone Matrix Application to Cell Nuclei Classification. Patern Recognit Inf Process 2009:140-145. 45. Barton MB, Jacob S, Shafiq J, Wong K, Thompson SR, Hanna TP, et al. Estimating the
demand for radiotherapy from the evidence: A review of changes from 2003 to 2012. Radiother Oncol 2014;112:140-144.
46. Naghavi AO, Echevarria MI, Strom TJ, Abuodeh YA, Venkat PS, Ahmed KA, et al. Patient choice for high-volume center radiation impacts head and neck cancer outcome. Cancer Med 2018;7:4964–79.
47. Vergeer MR, Doornaert PAH, Rietveld DHF, Leemans CR, Slotman BJ, Langendijk JA. Intensity-Modulated Radiotherapy Reduces Radiation-Induced Morbidity and Improves Health-Related Quality of Life: Results of a Nonrandomized Prospective Study Using a Standardized Follow-Up Program. Int J Radiat Oncol Biol Phys 2009;74:1-8.
48. Collins GS, Reitsma JB, Altman DG, Moons KGM. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): The TRIPOD Statement. Eur Radiol 2015;67:1142–51.
