Treatments for Locally Advanced Oropharyngeal Cancer: A Systematic Review of Clinical Effectiveness and Cost-Effectiveness

Transcription

1 CADTH RAPID RESPONSE REPORT: SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer: A Systematic Review of Clinical Effectiveness and Cost-Effectiveness Service Line: Rapid Response Service Version: 1.0 Publication Date: August 2017 Report Length: 76 Pages

2 Authors: Chuong Ho, Kelsey Seal, Charlene Argáez, Cheryl Ho Cite As: Treatments for locally advanced oropharyngeal cancer: a systematic review of clinical effectiveness and cost-effectiveness. Ottawa: CADTH; 2017 Aug. (CADTH rapid response report: systematic review). ISSN: (online) Disclaimer: The information in this document is intended to help Canadian health care decision-makers, health care professionals, health systems leaders, and policy-makers make well-informed decisions and thereby improve the quality of health care services. While patients and others may access this document, the document is made available for informational purposes only and no representations or warranties are made with respect to its fitness for any particular purpose. The information in this document should not be used as a substitute for professional medical advice or as a substitute for the application of clinical judgment in respect of the care of a particular patient or other professional judgment in any decision-making process. The Canadian Agency for Drugs and Technologies in Health (CADTH) does not endorse any information, drugs, therapies, treatments, products, processes, or services. While care has been taken to ensure that the information prepared by CADTH in this document is accurate, complete, and up-to-date as at the applicable date the material was first published by CADTH, CADTH does not make any guarantees to that effect. CADTH does not guarantee and is not responsible for the quality, currency, propriety, accuracy, or reasonableness of any statements, information, or conclusions contained in any third-party materials used in preparing this document. The views and opinions of third parties published in this document do not necessarily state or reflect those of CADTH. CADTH is not responsible for any errors, omissions, injury, loss, or damage arising from or relating to the use (or misuse) of any information, statements, or conclusions contained in or implied by the contents of this document or any of the source materials. This document may contain links to third-party websites. CADTH does not have control over the content of such sites. Use of third-party sites is governed by the third-party website owners own terms and conditions set out for such sites. CADTH does not make any guarantee with respect to any information contained on such third-party sites and CADTH is not responsible for any injury, loss, or damage suffered as a result of using such third-party sites. CADTH has no responsibility for the collection, use, and disclosure of personal information by third-party sites. Subject to the aforementioned limitations, the views expressed herein are those of CADTH and do not necessarily represent the views of Canada s federal, provincial, or territorial governmentsor any third party supplier of information. This document is prepared and intended for use in the context of the Canadian health care system. The use of this document outside of Canada is done so at the user s own risk. This disclaimer and any questions or matters of any nature arising from or relating to the content or use (or misuse) of this document will be governed by and interpreted in accordance with the laws of the Province of Ontario and the laws of Canada applicable therein, and all proceedings shall be subject to the exclusive jurisdiction of the courts of the Province of Ontario, Canada. The copyright and other intellectual property rights in this document are owned by CADTH and its licensors. These rights are protected by the Canadian Copyright Act and other national and international laws and agreements. Users are permitted to make copies of this document for non-commercial purposes only, provided it is not modified when reproduced and appropriate credit is given to CADTH and its licensors. About CADTH: CADTH is an independent, not-for-profit organization responsible for providing Canada s health care decision-makers with objective evidence to help make informed decisions about the optimal use of drugs, medical devices, diagnostics, and procedures in our health care system. Funding: CADTH receives funding from Canada s federal, provincial, and territorial governments, with the exception of Quebec. SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 2

3 Reviewers External Reviewers This document was externally reviewed by content experts and the following individuals granted permission to be cited. Stephanie Snow, BA, BSc, MD, FRCPC Medical Oncologist / Assistant Professor QEII Health Sciences Centre / Dalhousie University Halifax, NS, Canada John de Almeida, MD, MSc, FRCSC Head and Neck Surgeon / Assistant Professor University Health Network / University of Toronto Toronto, ON, Canada Authorship Chuong Ho led the project protocol development; selected studies; extracted, tabulated, and analyzed data; wrote the clinical section of the report; and revised the report based on reviewers comments. Kelsey Seal contributed to article selection, study quality assessment, data extraction, tabulation of data for the both the clinical and economic reviews; and to the analysis of data for the clinical review. Charlene Argáez designed and executed the literature search strategies; wrote the search methods section; and managed report referencing. Cheryl Ho, MD, FRCPC, participated in the conceptualization of clinical approaches to the treatment of locally advanced oropharyngeal cancer and in the preparation and review of the protocol and draft report. All authors approved the final draft report. Acknowledgments The authors would like to acknowledge Laura Weeks, PhD, for providing methodological support and providing critical reviews of the study protocol and final report and Barbara Greenwood Dufour for reviewing the study protocol and final report for clarity and contribution to knowledge mobilization activities. Conflicts of Interest Dr. Cheryl Ho has received honoraria for travel or speaking engagements from Lilly, Boehringer Ingelheim, Pfizer and Bristol Myers Squibb. She has also received payment as an advisor or consultant from Merck, Astra Zeneca, Bristol Myers Squibb, Roche and Lilly, and research grants from Boehringer Ingelheim, Sanofi Genzyme, Astra Zeneca and Eisai. No other conflicts of interest were declared.. SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 3

4 Table of Contents Abbreviations... 6 Executive Summary... 7 Context and Policy Issues Introduction Issues Objectives Research Questions Methods Literature Search Strategy Selection Criteria and Methods Exclusion Criteria Data Extraction Critical Appraisal of Individual Studies Data Analysis and Synthesis Methods Summary of Evidence Quantity of Research Available Summary of Patient and Study Characteristics Summary of Critical Appraisal Summary of Findings Discussion Summary of Evidence and Discussion Strengths and Limitations of Review Conclusions and Implications for Decision- or Policy-Making References Appendix 1: Literature Search Strategy Appendix 2: Flow Chart of Included Studies Appendix 3: List of Included Studies Appendix 4: List of Excluded Studies Appendix 5: Study Characteristics Appendix 6: Patient Characteristics SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 4

5 Appendix 7: Critical Appraisal of Included Studies Appendix 8: Validity of Outcome Measures Appendix 9: Clinical Data Appendix 10: Economic Data SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 5

6 Abbreviations CRT chemoradiotherapy DFS DSS EORTC-QLQ EORTC-QLQ-C30 EORTC-QLQ-H&N35 Gy HNSCC HPV MCID IMRT OS p16 PORT QALY RCT T TLM TORS UW-QoL disease-free survival disease-specific survival European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire core module European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire head and neck module gray (unit) head and neck squamous cell carcinoma human papillomavirus minimum clinically important difference intensity-modulated radiotherapy overall survival tumour suppressor protein p16 post-operative radiotherapy quality-adjusted life-year randomized controlled trial primary tumour transoral laser microsurgery transoral robotic surgery University of Washington Quality of Life Questionnaire SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 6

7 Executive Summary The Issue Oropharyngeal cancer, a subset of head and neck cancer, is associated with human papillomavirus (HPV) infection (often detected with protein p16 expression status), and alcohol or tobacco use. It is the 13th most commonly diagnosed cancer and the 15th most common cause of cancer death for adults in Canada. Treatment strategies for locally advanced oropharyngeal cancer (T3 to 4a, lymph node [N] stages N0 to N1 or any primary tumour [T], N2 to N3) include primary surgery with or without adjuvant radiotherapy or chemoradiotherapy (CRT), or primary concurrent CRT with or without salvage surgery. Surgery is usually performed as open surgery or using recent advances in minimally invasive techniques, namely, transoral robotic surgery (TORS) and transoral laser microsurgery (TLM). Systemic therapy options usually consist of chemotherapy using platinum-based drugs (e.g., single-agent cisplatin) or antibody therapy using cetuximab in the concurrent CRT setting, and cisplatin in the adjuvant setting. Radiotherapy includes both conventional and intensity-modulated radiotherapy (IMRT), which delivers a precise radiation dose to the tumour while sparing surrounding structures. Altered fractionation schedules have been explored, including accelerated fractionation that reduces the total treatment time, and hyper-fractionation that involves daily administration of two reduced-dose fractions. The superiority of initial surgery versus initial concurrent CRT for the treatment of oropharyngeal cancer in terms of oncologic and functional outcomes, toxicities, complications, and quality-of-life outcomes is not apparent. Likewise, the costeffectiveness of the two regimens is unclear. Objectives This systematic review aims to compare the clinical effectiveness of primary surgical therapy (with or without adjuvant radiation or CRT) versus primary CRT (with or without salvage surgery) for adults with locally advanced oropharyngeal cancer. Costeffectiveness of the two treatment strategies will also be examined. Methods This review is based on a protocol developed a priori. Published and unpublished literature describing studies that actively compare both treatment strategies and cost studies were identified through systematic searches of multiple databases and resources. The outcomes investigated were oncologic and functional outcomes, toxicities and complications of treatments, quality of life, and cost-effectiveness. Two reviewers independently screened the titles and abstracts of all citations retrieved from the literature search and ordered the full text of articles based on the selection criteria. The reviewers independently reviewed the full text of the selected articles and compared the independently chosen included and excluded studies. Data were extracted independently by reviewers and any disagreements were resolved through discussion until consensus was reached. The quality of clinical studies and cost evaluations was assessed using the Downs and Black checklist and Drummond checklist. Due to heterogeneity in both study and patient characteristics, metaanalyses were not performed. A narrative summary of the included study findings was constructed instead. SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 7

8 Results Clinical Review No randomized controlled trials (RCTs) were identified through the literature search. This systematic review of 14 head-to-head observational studies revealed three main findings. First, oncologic outcomes between primary surgery and primary concurrent CRT after five years showed inconsistent results across studies and a lack of consistent statistical significance between the outcome differences. Functional outcomes after one year seemed to favour concurrent CRT over open surgery in one study, but another study seemed to favour surgery over concurrent CRT when TLM or TORS was the surgical approach. There were no statistically significant differences between the two treatment strategies for the most common complications, such as acute dermatitis, mucositis, chronic swallowing difficulty, dry mouth, and trismus after four years of follow-up, but statistically more CRT patients than surgery patients experienced hematological toxicities and pharyngitis during treatment. Reported quality-of-life scores for physical and social functions were favourable in the CRT group compared with surgery after a median follow-up time of 56 months. Data from studies with shorter follow-up times suggested similar quality of life between the groups after two years, and TLM or TORS led to better swallowing outcomes after one year. Second, subgroup analyses showed that the type of surgery (i.e., open versus transoral [TORS or TLM]) did not change the oncologic outcomes when compared with concurrent CRT. Third, the quality of current evidence is weak due to the heterogeneous nature of the included observational studies where selection bias is probable, preventing a solid comparison between the two treatment options. Larger controlled trials are needed in future to strengthen the evidence. Economic Review No Canadian cost-effectiveness studies were identified. The literature search identified one US cost-effectiveness study that compared TORS with primary CRT plus platinum-based therapy based on a base case of a non-smoking, HPV-positive 65- year-old man with locally advanced oropharyngeal squamous cell carcinoma. The study found that the recurrence rate was the determining factor for the costeffectiveness of the treatment options, with concurrent CRT being the dominant strategy. Primary CRT dominated TORS in almost every situation, except when the recurrence risk after CRT was high, or the risk after TORS was very low. SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 8

9 Conclusions This review provides evidence that comparison of oncologic outcomes, after five years of follow-up, between primary surgery and primary concurrent CRT for locally advanced oropharyngeal cancer is inconclusive, due to inconsistent results across studies and lack of consistent statistical significance. The type of surgical approach (i.e., open or transoral [TORS or TLM]) did not seem to change the oncologic outcomes compared with concurrent CRT. Limited evidence on functional outcomes after one year seemed to favour surgery over concurrent CRT when TLM or TORS was the surgical approach. The most common complications were similar after more than four years of follow-up, but a larger number of CRT patients may experience hematological toxicities and pharyngitis during treatment compared with surgery. Quality-of-life outcomes favoured CRT compared with surgery in general, but TLM or TORS led to better swallowing outcomes after one year. Based on a US economic analysis that may not be generalizable to the Canadian context, compared with TORS, primary CRT plus platinum-based therapy may be a cost-effective therapy under most circumstances for patients with T1 to T2, N2a to N2b, p16-positive oropharyngeal cancer. Findings may not be applicable to patients with different characteristics. Based on a sensitivity analysis in that evaluation, recurrence rates following treatment are a determining factor for cost-effectiveness. No cost-effectiveness analyses involving TLM or immunologic systemic therapy (e.g., cetuximab) were identified. The cost-effectiveness of CRT versus surgical treatment for Canadian patients with locally advanced oropharyngeal cancer remains unknown. The results from the clinical review must be interpreted with caution, given the heterogeneity of the included studies and the absence of RCTs, which prevented a pooled estimate of the clinical outcomes. Because of the non-randomized nature of the studies, there is an increased risk of selection bias, and it is possible that patients were selected to receive surgery or systemic therapy based on their characteristics or the complexity of their individual case, which may confound results. The outcomes analyzed are reflective of the relatively short-term follow-up times that have been reported to date. Until long-term data become available, no further conclusions can be drawn beyond those previously outlined. Large RCTs with long-term follow-up times are necessary to provide robust evidence. The generalizability of the results of the cost-effectiveness study to a Canadian context may not be strong, since the costs of the intervention were based on one American jurisdiction, and analyses were not based on a sufficiently large volume of data on which to estimate clinical efficacy and harms of treatment strategies. SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 9

10 Context and Policy Issues Introduction Head and neck squamous cell carcinoma (HNSCC) encompasses multiple sites of origin, including the oral cavity, oropharynx, and larynx, with diverse drivers of carcinogenesis and clinical outcomes. Oropharyngeal cancer, the subset of HNSCC that originates in the oropharynx, is associated with human papillomavirus (HPV) infection (often detected with protein p16 expression status), alcohol or tobacco use, and is the 13th most common diagnosed cancer and the 15th most common cause of cancer death for adults in Canada. 1 It is estimated that in 2015, 4,400 Canadians (2,900 men and 1,450 women) were diagnosed with HNSCC, and 1,200 Canadians would die of the disease. 2-5 In 2012, 3,760 Canadians were diagnosed with an HPV-associated cancer and, among them, oropharyngeal cancer was the most common, with 1,335 cases. 6 Oropharyngeal cancer is defined as locally advanced when the cancer is in clinical stage III, IVa, or IVb (i.e., tumour stage T3 to T4a plus lymph nodes stage N0 to N1; or any tumour stage plus lymph nodes stage N2 to N3). 7,8 Both primary surgery and systemic therapies play important roles in oropharyngeal cancer treatment. Non-surgical treatments and transoral surgical approaches have, in most centres, become mainstays of treatment, alongside open-surgical treatment. 9 A 2003 trial by Adelstein et al. 10 established concurrent cisplatin and conventional fractionation radiotherapy as the standard of care over radiotherapy alone, with a 13% improvement in three-year survival. Of the population of HNSCC patients included in this study, 59% had oropharyngeal cancer. More recently, another radiosensitizing drug, cetuximab, has been made available in Canada as an alternative to cisplatin as part of systemic therapy for HNSCC. 11 Existing guidelines, including the 2015 Alberta Health Clinical Practice Guideline on oropharyngeal treatment 8 and the 2017 National Comprehensive Cancer Network Guidelines for head and neck cancer, 12 recommended either surgery followed by radiotherapy or chemotherapy, or concurrent chemoradiotherapy (CRT) for locally advanced oropharyngeal cancer. The systemic therapy options recommended by the guideline are single-agent cisplatin or cetuximab in the concurrent CRT setting, and cisplatin in the adjuvant setting. 8 Single-agent curative intent radiotherapy can be used for patients with a contraindication to concurrent systemic therapy who are also not surgical candidates, and for HPVpositive oropharyngeal cancers that carry a better prognosis and may not require systemic radiation sensitizers to achieve favourable outcomes. Intensity-modulated radiotherapy that delivers precise radiation doses to the tumour while sparing surrounding structures is the accepted standard of care (the recommended radiation dose is 66 Gy to 70 Gy when part of concurrent CRT setting, or 60 Gy to 66 Gy when following surgery in the adjuvant setting). 8 Altered fractionation schedules have been explored, including accelerated fractionation that reduces the total treatment time, and hyper-fractionation that involves daily administration of two reduced-dose fractions. Relatively recent advances in minimally invasive techniques transoral robotic surgery (TORS) and transoral laser microsurgery (TLM) are being incorporated as favourable options for early-stage oropharyngeal cancer A 2015 costeffectiveness analysis 16 showed that from the societal perspective, for early T-stage oropharyngeal carcinoma, TORS is cost-effective for treatment. For oropharyngeal cancer that is at a more advanced stage, treatment with these minimally invasive techniques is still associated with functional impairment, and advanced-stage cancer SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 10

11 cases are not prototypical for TORS or TLM treatment. For residual or recurrent oropharynx cancer, the more traditional open-surgical approach for performing a mandibulotomy, lateral pharyngotomy, or neck dissection is used. In these cases, microvascular reconstruction of the subsequent defect is almost universally required. Issues The superiority of primary surgery versus primary CRT in the treatment of locally advanced oropharyngeal cancer in terms of oncologic and functional outcomes, toxicities, complications, and quality of life are not apparent. The cost-effectiveness of the two regimens is likewise unclear. Objectives This systematic review aims to compare the clinical effectiveness of primary surgical therapy (with or without adjuvant radiation and chemotherapy) versus primary CRT (with or without salvage surgery) for adults with locally advanced oropharyngeal cancer. The cost-effectiveness of the two treatment strategies will also be examined. Research Questions 1. What is the comparative clinical effectiveness of primary surgery (with or without adjuvant radiotherapy and chemotherapy) versus primary CRT (with or without salvage surgery) for the treatment of adults with a diagnosis of locally advanced oropharyngeal cancer? 2. What is the cost-effectiveness of primary surgery (with or without adjuvant radiotherapy and chemotherapy) for the treatment of adults with a diagnosis of locally advanced oropharyngeal cancer? 3. What is the cost-effectiveness of primary CRT (with or without salvage surgery) for the treatment of adults with a diagnosis of locally advanced oropharyngeal cancer? Methods This review is based on a protocol developed a priori. 17 Literature Search Strategy A limited literature search was conducted on key resources including PubMed, Ovid, the Cochrane Library, University of York Centre for Reviews and Dissemination databases, and Canadian and major international health technology agencies. A focused Internet search was also conducted. Filters were applied to limit the retrieval to health technology assessments, systematic reviews, meta-analyses, randomized controlled trials (RCTs), non-randomized studies, and economic studies. Where possible, retrieval was limited to the human population. The search was also limited to English-language documents published between January 1, 2001 and September 8, Bi-weekly search updates were run until April 30, See Appendix 1 for the detailed search strategy. SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 11

12 Selection Criteria and Methods Two reviewers independently screened the titles and abstracts of all citations retrieved from the literature search and, based on the selection criteria (Table 1), ordered the full text of any articles that appeared to meet those criteria. The reviewers independently reviewed the full text of the selected articles, applied the selection criteria, and compared the independently chosen included and excluded studies. Disagreements were resolved through discussion until consensus was reached. Duplicate publications of the same study were excluded unless they provide additional outcome information of interest. The study selection process is presented in a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow chart in Appendix 2. Table 1: Selection Criteria Population Intervention Comparator Outcomes Study design Adult patients (aged 18 years) with biopsy-proven primary locally advanced oropharyngeal carcinoma (American Joint Committee on Cancer stage III, IVa, or IVb) Concurrent primary chemoradiotherapy (with or without salvage surgery) a Primary surgery (with or without adjuvant radiotherapy and chemotherapy) a Oncologic outcomes Overall survival, recurrence-free survival, local-regional control (freedom from local progression) Functional outcomes Rate of percutaneous feeding-tube dependence, rate of tracheostomy dependence Quality-of-life outcomes Quality of life, anxiety, recreation, pain, saliva, dry mouth, swallowing, taste, and speech measured with a standardized scale Toxicities Nephrotoxicity, neurotoxicity, adverse infusion reactions, mortality rates due to chemotherapy-induced neutropenia, rates of skin and mucosal toxicity, osteoradionecrosis, soft tissue necrosis, cerebellar necrosis, xerostomia Complications Fistula formation, post-operative hemorrhage, hematoma formation, surgical-site infections, pneumonia, peri-operative mortality Cost-effectiveness outcomes Cost of primary surgery ± adjuvant radiotherapy ± adjuvant chemotherapy Cost of primary chemoradiotherapy ± salvage surgery All specific related costs Quality-adjusted life-years Incremental cost-effectiveness ratio Randomized controlled trials, non-randomized studies with a comparator group, cost studies a Any chemoradiotherapy regimen or surgical approach reported in publications in the last 15 years was considered for inclusion. Exclusion Criteria Studies were excluded when they did not meet the selection criteria or presented preliminary results in abstract form. Duplicate publications, narrative reviews, case studies, and editorials were excluded. Studies with mixed populations were excluded if results for patients with locally advanced oropharyngeal cancer were not reported separately. SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 12

13 Data Extraction A data-extraction form for the review was designed a priori to document and tabulate relevant study characteristics. Data were extracted independently by two reviewers, and any disagreements were resolved through discussion until consensus was reached. While it was planned to contact study authors in case of missing or unclear data, this was not necessary. A calibration exercise (data from a sample of studies were extracted independently by each reviewer and compared before all studies were extracted) was conducted to ensure consistency in data extraction. Critical Appraisal of Individual Studies The quality of clinical studies and cost evaluations was assessed using the Downs and Black 18 and Drummond 19 checklists, respectively. Quality assessments were completed by the lead researcher and verified by the second researcher. Any disagreements were resolved through discussion until consensus was reached. Studies were not excluded based on methodological assessments, but assessments were used to explain any potential differences across study results. Data Analysis and Synthesis Methods Tables were created to summarize quantitative findings for each outcome listed in Table 1. Data were synthesized separately for each question, by outcome. Head-to-head trials between the two treatment strategies were included; RCTs were not found from the literature search. Due to the high level of heterogeneity in the study and patient characteristics, meta-analyses were not performed. A narrative summary of the included study findings was instead constructed. Planned subgroup analyses based on use of additional treatments (i.e., surgery alone or surgery with adjuvant chemotherapy or radiotherapy; primary CRT with or without salvage surgery), surgical approach (open, TORS, or TLM), systemic drug therapy (cisplatin or cetuximab), and patient characteristics such as HPV status, smoking habit, and alcohol use were performed when sufficient data were available. For the economic review, costs and cost-effectiveness outcomes for primary surgery (with or without adjuvant radiotherapy or CRT), and primary CRT (with or without salvage surgery) from the available evidence were summarized, and reported narratively and in tables. Summary of Evidence Quantity of Research Available A total of 772 citations were identified in the literature search. Following the screening of titles and abstracts, 719 citations were excluded and 53 potentially relevant reports from the electronic search were retrieved for full-text review. No potentially relevant publications were retrieved from the search of the grey literature (i.e., reports and government information that are not published commercially and that are inaccessible via bibliographic databases). Of these 53 potentially relevant articles, 38 publications were excluded for various reasons, while 15 publications (14 clinical studies and one economic evaluation) met the inclusion criteria and were included in this report. A PRISMA flowchart for the study selection can be found in Appendix 2. SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 13

14 A list of included studies is provided in Appendix 3. A list of excluded studies, with reasons for exclusion, is provided in Appendix 4. Summary of Patient and Study Characteristics Details of the individual study characteristics are provided in Appendix 5. Study Design Fourteen non-randomized studies were identified regarding the comparative clinical effectiveness of primary surgery versus primary CRT for locally advanced oropharyngeal cancer. Six studies were prospective, seven studies were retrospective, and one study 33 was cross-sectional. Specifically, four studies were prospective cohort studies, one study 25 was a prospective case series compared with a historical cohort, one study 24 was a prospective observational pre- and post-study compared with a historical cohort, seven studies were retrospective cohort studies. Two studies 20,22 had an overlapping patient population, but reported on different outcomes. One economic evaluation 34 of the cost-effectiveness of primary surgery versus primary CRT was identified. The study 34 had a cost-effectiveness analysis and a Markov model was designed. Country of Origin Of the 14 studies, three studies 21,25,33 were conducted in Italy, two studies 20,22 were conducted in Canada, three studies 23,26,27 were conducted in the United States, one study 24 was conducted in the United Kingdom, one study 28 was conducted in Australia, one study 29 was conducted in Japan, two studies 30,32 were conducted in South Korea, and one study 31 was conducted in Taiwan. The economic evaluation 34 was conducted in the United States. The analysis was performed from the payer perspective with a 10-year time horizon. Patient Population Additional details of the patient characteristics in the included studies is provided in Appendix 6. All of the studies included patients with locally advanced oropharyngeal cancer. Two of the studies 21,23 included only patients with oropharyngeal squamous cell carcinoma, but they provided separate data for locally advanced oropharyngeal squamous cell carcinoma. Six studies reported patients HPV status, 20,23,24,26-28 and five studies reported smoking habits. 20,23,27,29,30 One study reported separate data on patients HPV status and smoking habits. 20 While eight studies had patients in treatment groups with different demographics such as age, sex, or risk profile, six studies 20,25,27,29,31,33 reported there was no statistically significant differences in patient demographics or comorbidities between the intervention groups. One study reported a different radiation dose used between the treatment groups. 32 The authors of the economic evaluation 34 designed a Markov model to simulate the clinical history of a non-smoking 65-year-old with T1 to T2, N2a to N2b, HPV-positive oropharyngeal squamous cell carcinoma. SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 14

15 Interventions All 15 studies used concurrent CRT as their intervention. All studies used cisplatin or other platinum-based drugs for chemotherapy. In addition to platinum-based drugs, three studies used 5-fluorouracil, 25,32,33 one study used cetuximab and tirapazamine, 21 one study used 5-fluorouracil, doxorubicin, and taxanes, 22 and one study used mitomycin C. 24 For radiotherapy, either a conventional approach or intensitymodulated radiotherapy (IMRT) was used in all studies, but details or dose were not specified in eight studies ,26,29-32 Comparator Among the 14 included clinical studies, the surgical approach was open surgery in four studies, 22,28,30,31 open surgery or transoral surgery in six studies, 20,25,26,29,32,33 TLM in one study, 24 TORS in one study, 23 TLM or TORS in one study, 27 and the specific approach was not reported in one study. 21 Eight studies had post-operative radiotherapy (PORT) as adjuvant therapy, 21,25,28-33 and six studies had PORT or CRT as adjuvant therapy. 20,22-24,26,27 The economic evaluation study used TORS as the surgical approach, and PORT or CRT as adjuvant therapy. 34 All studies used cisplatin or other platinum-based drugs for chemotherapy. For radiotherapy, a conventional approach or IMRT was used in all studies, but details or dose were not specified in eight studies ,26,29-32 Outcomes Thirteen studies 20,22-33 reported on oncologic, functional, toxicity, and complication outcomes, in which seven studies reported oncologic outcomes after five years followup, 20,22,23,28-31 and the remaining six studies reported outcomes after less than five years of follow-up ,32,33 Four studies 21,27,32,33 reported quality-of-life outcomes. The cost-effectiveness evaluation 34 reported costs, and cost per quality-adjusted life-year (QALY). The analysis was performed from the payer perspective with a 10-year time horizon. Cost and QALYs were discounted at an annual rate of 3%. The Markov model was based on a base case of a non-smoking 65-year-old man with T1 to T2, N2a to N2b, HPV-positive oropharyngeal squamous cell carcinoma. Summary of Critical Appraisal All included clinical studies were non-rcts and thus had a potential bias in patient selection. Due to the nature of the interventions, blinding was not possible in the included studies. Hypotheses were described; the method of selection from the source population was described; main outcomes, interventions, patient characteristics, and main findings were clearly described; and estimates of random variability and actual probability values were provided in most studies. 20,22-27,29-33 Patients and treatment options were representative of the condition under study. Among the trials reporting oncologic outcomes, only one study had clear definitions of clinical end points. 31 The lack of clinical end point definitions may lead to variability in the observed results both within and between studies. There was imbalance in patient demographics between the treatment groups in eight studies, despite attempts to correct for baseline imbalance, with six studies 20,25,27,29,31,33 reporting there were no statistically significant differences between the groups. Two studies 24,25 used a historical control group, so it is possible that changes in diagnostic methods or patient characteristics (e.g., HPV rates) between time periods may have influenced the results. It was unclear in all studies whether there was sufficient power to detect a clinically important effect, as SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 15

16 studies did not report power calculations to determine an appropriate sample size Loss to follow-up was not clearly described in eight studies ,27-31 The cost-effectiveness study 34 based cost calculations on the 2015 Medicare payment schedule for the Chicago region. Medicare payments vary across the country, influencing the relative cost-effectiveness of the two treatment strategies. The utility data were extracted from one study that favoured surgery, thus presenting a potential bias toward surgery. The Markov model was based on a base case of a non-smoking 65-year-old man with T1 to T2, N2a to N2b, HPV-positive oropharyngeal squamous cell carcinoma. The 10-year time horizon seems appropriate. All relevant costs were considered. Appropriate sensitivity analyses were done. The generalizability of the results may not be strong, since the costs of the intervention varies across jurisdictions, and analyses were not based on a sufficiently large volume of data on which to estimate clinical efficacy and harms of treatment strategies. While the costeffectiveness analysis examined a slightly different population (T1 to T2) than the patients described in most of the clinical studies (T3 to T4), the base case for the economic model is considered locally advanced due to nodal involvement. 8 Details of the quality appraisal of the individual included studies are provided in Appendix 7. The validity of outcome measures for the quality-of-life instruments used in the identified studies are described in Appendix 8. Evidence for validity was found for each tool, though minimum clinically important differences (MCID) were not described for the European Organisation for Research and Treatment of Cancer head and neck module (EORTC-QLQ-H&N35), which was used in two studies. 21,32 Summary of Findings Clinical Review The literature search identified 14 observational studies comparing clinical outcomes between primary surgery and primary concurrent CRT in patients with locally advanced oropharyngeal cancer. A narrative synthesis was completed, with a focus on the outcomes reported after the longest follow-up time (five years for oncologic outcomes, and one year for functional outcomes). Details of the clinical findings in the included studies are provided in Appendix 9. Clinical findings comparing surgery (all types) and concurrent CRT is listed in Table 2, and subgroup analyses between open surgery and concurrent CRT, and between TLM or TORS and concurrent CRT is listed in tables 3 and 4, respectively. Appendix 9, Table 14, and Table 15 provide the individual study findings according to outcome. Oncologic and Functional Outcomes As elaborated below, in general, oncologic outcomes between the two treatment strategies after five years were similar. There were inconsistencies in findings among studies and lack of consistent statistically significant differences between outcomes. When patients were stratified by risk factors such as HPV status and smoking habits, data from one study suggested that a primary surgery plus adjuvant CRT approach offered a higher disease-specific survival (DSS) rate than concurrent CRT, with or without salvage surgery, in patients with p16-negative cancers and in smokers. Data from one study found similar overall survival (OS) rates between the two treatment strategies for high-risk patients, but a significantly higher survival rate with CRT for patients with low or intermediate risk. Data on oncologic outcomes with follow-up times SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 16

17 of less than five years mostly showed non-statistically significant differences between the two treatment strategies. Limited evidence on functional outcomes after one year of follow-up seems to favour concurrent CRT over open surgery, but appears to favour surgery over concurrent CRT when TLM or TORS is the surgical approach. Oncologic Outcomes After Five Years of Follow-up Three prospective cohort studies 20,22,23 and four retrospective chart reviews compared five-year oncologic outcomes between surgery and concurrent CRT in patients with locally advanced oropharyngeal cancer. Examination of study details in terms of patient demographics or treatment did not show a clear pattern to explain the differences in outcomes between studies. For DSS, surgery led to a greater survival rate in three studies 20,22,23 and a lower rate in one study; 31 statistical significance was found in two studies, 20,22 and differences were not significant in two studies. 23,31 When patients were stratified by p16 (HPV status) and smoking status, data from one study 20 on 200 patients found that surgery (transoral or open) plus adjuvant CRT was associated with statistically significantly higher DSS for p16-negative patients (72.0% versus 37.1%) and for smokers (76.9% versus 47.5%) compared with CRT alone. When patients were further stratified into subsets, smokers with p16-positive cancers (n = 81) and p16-negative smokers (n = 65) had a statistically significantly higher DSS rate with surgery than with CRT (85.4% versus 66.2% and 60.3% versus 27.4%, respectively). Differences in DSS survival between the two strategies for p16-positive non-smokers (n = 35), and p16- negative non-smokers (n = 19) were not statistically significant. For OS, findings were mixed among studies, with surgery leading to a greater survival rate than CRT in three studies, 20,22,23 and a lower rate in four studies; statistical significance was found in three studies 20,22,28 and not found in the rest. There was no clear pattern regarding the characteristics of studies that demonstrated statistically significant differences and those that did not. Data from one study 20 showed that p16- negative smokers had a statistically significantly higher OS rate with surgery than with CRT (43.1% versus 20.8%). Differences in OS between the two strategies for the other subsets such as p16-positive non-smokers, p16-positive smokers and p16-negative non-smokers were not statistically significant. Data from one study that stratified patients into high-, intermediate-, and low-risk groups 28 found similar survival rates between the two treatment strategies for high-risk patients, but a statistically significantly higher rate with CRT for intermediate- and low-risk patients (hazard ratio: 4.46; 95% confidence interval [CI], 1.68 to 11.86). For disease-free survival (DFS), surgery led to a higher survival rate in one study, 31 and a lower rate in one study; 30 the difference was not statistically significant in either study. For locoregional control (LRC), surgery led to a higher control rate in one study, 31 and a lower rate in one study; 23 the difference was not statistically significant in either study. For locoregional recurrence rate, surgery led to more patients with locoregional recurrence than CRT in one study, 30 and fewer patients in one study. 29 Statistical significance was not reported in either study. For locoregional recurrence free survival, data from one study suggested surgery led to a lower survival rate than CRT 30 ; the difference was not statistically significant. For progression-free survival, SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 17

18 data from one study showed surgery led to a lower survival rate than CRT; 29 the difference was not statistically significant. For distant metastasis free survival, surgery led to a greater survival rate in one study, 31 and a lower rate in one study; 30 the difference was not statistically significant in either study. Oncologic Outcomes After Less Than Five Years of Follow-up Five studies reported oncologic outcomes after less than five years of followup. 22,25,26,32,33 DSS, OS, DFS, and progression-free survival were not found to be statistically significantly different compared with surgery (transoral or open) and concurrent CRT in four trials with follow-up times from two to four years, 25,26,32,33 while one trial found open surgery significantly reduced the likelihood of DSS and OS after two years of follow-up. 22 Functional Outcomes Two retrospective cohort studies 27,29 and one observational pre- and post-study 24 compared functional outcomes between surgery and concurrent CRT in patients with locally advanced oropharyngeal cancer with up to one year of follow-up. Table 2: Surgery (All Types) Versus Chemoradiotherapy At completion of treatment, surgery led to a greater rate of gastrostomy-tube dependence than CRT in the study in which the surgery approach was mostly open surgery, 29 and a lower rate of dependence in the study in which TLM or TORS were the surgical approaches. 27 At one year, surgery led to a lower rate of gastrostomy-tube dependence (3% versus 10%), and lower esophageal stricture rate (0% versus 7%) than CRT, when TLM or TORS were used. 27 The differences in both studies in tubedependence rates were statistically significant. A pre- and post-observational study 24 found that fewer patients undergoing TLM (76%) reported swallowing problems after treatment than those undergoing CRT (92%). (In this study, the majority of patients in the TLM group were HPV-positive, but testing was not done in the CRT group.) Change in MD Anderson Dysphagia Inventory score between pre-treatment and three months post-treatment was greater for CRT than for TLM (effect size not reported); the difference was statistically significant and appeared to be clinically important. Information on other functional outcomes was not reported in the included studies. Outcome Surgery + Adjuvant RT or CRT Oncologic Outcomes (at 5 Years [Range]) CRT ± Salvage Surgery Notes OS 52.9% to 86% 20,22,23, % to Surgery led to a higher 88.9% 20,22,23,28-31 OS rate in 3 studies, 20,22,23 and a lower rate in the remaining DSS 62.9% to 91% 20,22,23, % to Surgery led to a greater 78% 20,22,23,31 DSS rate in 3 studies, 20,22,23 and a lower rate in the remaining 1 31 Statistical Significance Between Surgery and CRT SS was demonstrated in 3 studies, 20,22,28 but not in the remaining 4 studies 23,29-31 SS was demonstrated in 2 studies, 20,22 but not in the remaining 2 studies 23,31 SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 18

19 Outcome Surgery + Adjuvant RT or CRT CRT ± Salvage Surgery Notes Statistical Significance Between Surgery and CRT DFS 53.7% to 79.5% 30,31 48% to Surgery led to a greater 84.2% 30,31 DFS rate in 1 study, 31 and a lower rate in the remaining 1 30 LRR 9.2% to 25.8% 29,30 8.1% to 28.0% 29,30 Surgery led to more patients with LRR than CRT in 1 study, 30 and fewer patients in 1 study. 29 LRC 59.2% to 81% 23, % to 91.0% 23,31 Surgery led to a greater control rate in 1 study, 31 and a lower rate in the remaining 1 23 LRRFS 88.4% % 30 Surgery led to a lower rate of LRRFS than CRT PFS 51.0% % 29 Surgery led to a lower PFS rate than CRT DMFS 88.9% to 90.7% 30, % to Surgery led to a greater 92.3% 30,31 rate of DMFS in 1 study, 31 and lower rate in the remaining 1 30 Functional Outcomes Rate of gastrostomy-tube dependence At completion of PORT: 13% 27 to 25.8% 29 At 1 year: 3% 27 At completion of CRT: 11% 29 to 29% 27 At 1 year: 10% 27 At completion of treatment, surgery led to a greater rate of gastrostomy-tube dependence in 1 study, 29 and a lower rate in the remaining study. 27 No SS was demonstrated SS values not reported No SS was demonstrated No SS was demonstrated No SS was demonstrated No SS was demonstrated SS was demonstrated Esophageal stricture rate Swallowing function At 1 year, surgery led to a lower dependence rate than CRT 27 At 1 year: 0% 27 At 1 year: 7% 27 At 1 year, surgery led to a lower esophageal stricture rate than CRT At 3 months, compared with baseline Based on MDADI, 16/21 patients (76%) reported more swallowing problems after treatment 24 At 3 months, compared with baseline Based on MDADI, 24/26 patients (92%) reported more swallowing problems after treatment 24 Change in Surgery led to a reduced chance of swallowing problems than CRT after treatment SS value not reported SS was demonstrated SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 19

20 Outcome Surgery + Adjuvant RT or CRT CRT ± Salvage Surgery MDADI score between baseline and 3 months was greater for CRT than for surgery Notes Statistical Significance Between Surgery and CRT CRT = concurrent chemoradiotherapy; DFS = disease-free survival; DSS = disease-specific survival; DMFS = distant metastasis free survival; LRC = locoregional control; LRR = locoregional recurrence; LRRFS = locoregional recurrence free survival; MDADI = MD Anderson Dysphagia Inventory; OS = overall survival; PFS = progressionfree survival; PORT = post-operative radiotherapy; RT = radiotherapy; SS = statistical significance. Toxicity Outcomes and Complications A retrospective cohort study examining the role of primary surgery in 82 patients with resectable stage III or stage IV tonsillar carcinoma 31 found no statistically significant differences in radiotherapy-related toxicity between open surgery plus adjuvant radiotherapy versus concurrent CRT after a median follow-up time of 39 months, with the most common acute and late toxicity being radiation mucositis (54.5% in the surgery group; 80% in the CRT group), dry mouth (77.3% in the surgery group; 80% in the CRT group), and neck fibrosis (54.5% in the surgery group; 32% in the CRT group). In this trial, conventional radiotherapy, not IMRT, was used. A retrospective cohort study compared open surgery plus adjuvant radiotherapy with concurrent CRT in 114 patients with locally advanced tonsillar carcinoma after a median follow-up time of 58 months in the surgery group and 44 months in the CRT group. 30 More CRT patients than surgery patients experienced grade 3 (or higher) hematological toxicities (leukopenia and neutropenia) (17 CRT patients, 4 surgery patients) and grade 2 pharyngitis during treatment (46 CRT patients, 49 surgery patients); the differences were statistically significant. Acute dermatitis, mucositis, chronic swallowing difficulty, dry mouth, and trismus were the most common complications, but there was no statistically significant difference between the two treatment strategies. Neither study reporting on toxicity outcomes 30,31 reported on nephrotoxicity, neurotoxicity, ototoxicity, or peri-operative mortality. One study 30 reported that no patient in either group experienced severe acute complications requiring hospitalization or surgery. Quality of Life The validity and MCID for various quality-of-life instruments, where available, are described in Appendix 8. Reported quality-of-life scores for physical and social functions were favourable in the CRT group compared with surgery, after a median follow-up time of 56 months. Data from studies with shorter follow-up times suggested similar quality of life between the groups after two years, and TLM or TORS led to better swallowing outcomes after one year. Long-term quality of life was evaluated in a cross-sectional study of 57 patients with locally advanced cancer who underwent surgery (transoral, open, or both) plus SYSTEMATIC REVIEW Treatments for Locally Advanced Oropharyngeal Cancer 20