Human papillomavirus associated oropharyngeal cancer now the most common mucosal head and neck cancer in Queensland

The profile and outcomes of head and neck cancer throughout Australia has changed over the past decade. The aim of this study was to perform a population‐based analysis of incidence, demographics, stage, treatments and outcomes of patients diagnosed with oropharyngeal squamous cell carcinoma (OPSCC), with a particular focus on HPV‐associated disease.


Introduction
Head and neck cancer arising from the upper aerodigestive tract are commonly squamous cell carcinomas (SCC) related to tobacco and/or alcohol abuse. 1 In many parts of the world, including North America and Europe, there has been a rise in the incidence of oropharyngeal SCC (OPSCC) compared with those from other head and neck sites.4][5] Where a non-surgical approach is employed in the curative management of head and neck cancer the standard of care has been concurrent chemoradiotherapy with high-dose cisplatin. 4he presence of p16 immunostaining of greater than 70% of tumour cells in OPSCC is considered to be a reliable surrogate for HPV-associated disease. 6ueensland Cancer Control Analysis Team (QCCAT) compiles and analyses information about cancer incidence, mortality, treatment, and survival housed in the Queensland Oncology Repository (QOR).The QOR contains approximately 50 million records between 1982 and 2019.
Between 2011 and 2015 the oropharynx became the most common newly diagnosed head and neck cancer in Queensland, surpassing oral cavity cancer. 7This change in incidence has resource and treatment implications, as the majority are more likely to be treated with chemoradiotherapy rather than surgery and are likely to be long-term survivors. 8,9The primary aim of this study was to analyse incidence trends of OPSCC relative to other head and neck cancers, patient demographics, p16 status, treatments utilised, and survival outcomes, between 1 January 2015 and 31 December 2019, based on QOR.The analysis predominantly focuses on the OPSCC p16+ cohort.
Previously published outcomes for this disease are mainly based on selected patients enrolled in clinical trials.This data provides real-world outcomes for unselected patients presenting with newly diagnosed OPSCC in Queensland, and is likely to reflect what is happening throughout Australia.This study received Metro South Health Human Research & Ethics Committee approval (HREC/2022/QMS/81561).

Methods
The primary aim is to analyse OPCSCC and, in particular, OPCSCC p16+, using the Queensland Oncology Repository (QOR) to benchmark incidence trends relative to other cancers.
This retrospective population-based study used linked QOR data, which collates and matches data from the Queensland Cancer Register together with public and private hospital admissions, surgery, radiation therapy, multidisciplinary team records (primarily in the public sector), and mortality data.

Staging classification and p16 status
Staging data was primarily obtained from Multidisciplinary Team meeting systems, chemotherapy systems, or radiation therapy systems across Queensland during patient episodes of care.The Cancer Alliance Queensland reviewed all staging information from these data sources to confirm accuracy.
Due to the emerging survival data based on p16/HPV status, the staging classification for this disease changed from the 7th Edition American Joint Commission on Cancer (AJCC) classification to the 8th edition in 2017.As such, the entire cohort was retrospectively re-staged according to the new staging classification for the purposes of this analysis. 10,11here p16 status was unknown, patients were staged according to the p16/HPV-negative classification (as per the TNM staging classification rules). 12Cancer stage and p16 status were recorded in multidisciplinary QOOL meeting software.This data was then linked to the individual patient record in QOR.

Treatment modalities utilised
As this is a population-based database, accordingly the cohort is composed of patients receiving a range of treatment modalities and/or in differing sequences along individual institutional policies.The use of surgery, chemotherapy, radiotherapy and concurrent chemoradiotherapy treatments will be described, providing an insight into the resources utilised for this disease in a population-based setting.
Head and neck cancer surgeries were identified from hospital admissions using the Australian Classification of Health Interventions (ACHI) 11th edition. 13Radiation therapy and intravenous systemic therapy records were linked to the earliest records created from treatment systems.Surgery and treatment records were then linked to the head and neck cancer diagnosis record if the start date was 30 days before diagnosis and up to 365 days after histological diagnosis to capture adjuvant therapy.

Comorbidities
The individual list of comorbidities considered are those contained within the Charlson Comorbidity Index (1987).No weighting was applied to the individual comorbidities, but a simple count of comorbidities present was used. 14he capturing of comorbidity data is limited to conditions coded in any hospital admission between 12 months before and 12 months after the date of cancer diagnosis.For any given cancer diagnosis, comorbidity is restricted to conditions other than the primary cancer.Smoking data is not captured at a population level.

Remoteness of residence and socioeconomic status (SES)
In this report, relative remoteness of residence is classified into three groups: Major city, Inner regional, and Other which combines Outer regional, Remote, and Very remote categories.This is based on the Australian Standard Geographical Classification (ASGC) Remoteness Structure. 15ocioeconomic status (SES) was assigned according to the Australian Bureau of Statistics Socio-Economic Indexes for Areas (SEIFA), a census-based measure of social and economic well-being and was categorised as affluent (deciles 1-2), middle (deciles 3-8), and disadvantaged (deciles 9-10). 16

Statistical analysis Time period and incidence
QCCAT performed an analysis on incidence counts and incidence rate trends, demographics, treatment modalities utilised, and outcomes of patients with newly diagnosed OPSCC between 1 January 2015 and 31 December 2019 in Queensland.Incidence relates to newly diagnosed Queensland residents only.Age-standardised rates (ASR) were calculated using the 2001 Australian population as a reference.A 3-year rolling average of the most recent 3 years of incidence is used.

Overall survival
The 2-and 5-year overall survival (all causes of mortality) was calculated from the date of histological diagnosis to the data cut-off date, 31 December 2020.Patients still alive at this date were censored.These rates were calculated in STATA 17 using the Kaplan-Meier method. 17Survival was plotted based on p16 status and 8th Edition AJCC Staging Classification.Only survival curves where the p16 status was known are shown.

Multivariable analysis
A Cox proportional hazards model was used to assess the impact of the following covariates on survival among patients with p16+ disease: stage, sex, age groups, SES, comorbidities, and residence at diagnosis.
The test of equality of survivor function was used to determine whether there were significant differences in survival across stage groups.We tested for any violation of the proportional hazards assumptions and found no evidence of violation.Post-estimation testing was done using the command estat phtest in Stata17 (StataCorp LLC, College Station, TX, USA).

Incidence trends for head and neck cancer
Between 1 January 2015 and 31 December 2019 there were 135,576 new cancer diagnoses in Queensland of which 3% (4192/135,576) were of the head and neck.A total of 38% (1584/4192) arose from the oropharynx, of which the majority, 96% (1527/1584), were OPSCC (the focus of this analysis).
Oral cavity cancer was more common than OPC (N = 485, N = 253) in females, while in males OPC was more common than oral cavity cancer (N = 1331, N = 686) (Fig. 1a).

Age-standardised incidence rate
When comparing the change in age-standardised incidence rate (ASR, 3-year rolling average) from 1982 to 2019 based on the available data from the QOR, cancer of the oropharynx rose from 2.6 to 6.3 per 100,000 persons, an increase of 144%, predominantly due to OPSCC.Age-standardised incidence rates (3-year rolling average) for selected head and neck cancers from 1982 to 2019 in Queensland are shown in Figure 1b.Among males, there was an increase from 3.8 to 11.0 per 100,000, and in females, there was an increase from 1.3 to 1.8 per 100,000.During the same time period, there was a decrease in the ASR for oral cavity cancer from 5.6 to 4.7 per 100,000 persons.This decrease was driven by a decrease in incidence among males (8.1 to 6.1 per 100,000), while in females incidence remained steady at 3.0 per 100,000.

Demographics for OPSCC diagnosis based on p16 status
Patient demographics and tumour stage (AJCC 8th Edition) for OPSCC diagnoses by p16 status are summarised in Table 1.

Treatments
Ninety three percent (1418/1527) of OPSCC patients had a record of a discussion at a Queensland multidisciplinary clinic in QOOL.

Variables impacting on survival for p16+ disease
An analysis of covariables and impact on survival for OPSCC p16+ disease is shown in Table 2. On multivariable analysis, increasing stage relative to Stage I, older age (≥65 years), increasing comorbidities (≥2) and disadvantaged SES had an adverse impact on survival.Sex or residence was not shown to have an impact on survival.

Discussion
This report provides a statewide population-based analysis confirming the emergence of HPV-associated OPSCC as the most common newly diagnosed head and neck cancer in Queensland, having surpassed oral cavity cancer.
This data provides real-world outcomes for unselected patients presenting with OPSCC, within the Australian health system context.It also provides an insight into resource implications given the rising incidence of this disease and the treatments utilised.The rise in OPC has primarily been driven by the increase in OPSCC p16+ and is considered a distinct clinical entity.Human papillomavirus subtypes 16 and 18, the same ones responsible for cervical cancer, predominantly cause HPV-associated OPSCC.
The majority of our patients with OPSCC p16+ were male, had fewer comorbidities and were younger than p16À patients.
This increase throughout the Western world has been associated with an increased rate of oral sex, its commencement at a younger age, and an increase in sexual partners. 3he latency from infection, from when people become sexually active, to the development of cancer is generally between 20 and 30 years. 3 HPV vaccination for girls was first introduced in the Australian National Immunisation Program in 2007, and for boys in 2013.It is therefore anticipated that the incidence will continue to rise for at least another 20 years before the impact of the national vaccination program is seen.According to the current Australian Institute of Health and Welfare (AIHW) figures there has been a rise in the number of actual new cases of OPC from 1982 (n = 181) to 2017 (n = 767), with an estimated projected number of new cases of 933 in 2021.This corresponds to an increase in the ASR (per 100,000) from 1.3 in 1982 to 3.1 in 2021. 18 randomised phase III trial of chemoradiotherapy demonstrated superior overall survival for HPV-associated OPSCC compared with non-HPV-associated OPSCC. 4In a single institution series from the Princess Alexandra Hospital, Brisbane, of 418 patients, Daniels et al. 5 reported 5-year locoregional relapse failure free survival and cancer specific survival following chemoradiotherapy of 91% (95% CI; 88-94) and 90% (95% CI; 87-93), respectively.
Chemoradiotherapy has therefore been predominantly used for this disease.This analysis confirms the high utilisation rates of chemoradiotherapy in Queensland, compared with surgery for this disease.In this series, 91% (1063/1170) of OPSCC p16+ received radiotherapy or chemoradiotherapy while only 31% (362/1170) underwent surgery.With a decline in oral cavity cancer incidence, where surgery is the preferred treatment of choice, and the increase in OPSCC, where chemoradiotherapy is the preferred treatment, there are important resource implications for the Australian health system with respect to potentially requiring greater chemotherapy and radiotherapy services.
Unlike clinical trials, this cohort is made up of patients with a wide range of co-morbidities, disease extent, stage, and fitness for treatment.As such the intent of treatment will have varied from curative, palliative, or salvage.It is therefore beyond the scope of this report to perform a detailed analysis to compare the impact of treatment modality received and outcome because of the potential bias of one treatment over another based on intent.Nonetheless the overall utilisation rates for the various modalities provide important insights into the treatment resources used, a main goal of the analysis.There was good concordance between stage and survival, with the increasing stage (8th Edition AJCC) resulting in lower survival.The majority of OPSCC p16+ presented with stage I disease.The 5-year overall survival for the entire OPSCC p16+ cohort of 79%, compares favourably with traditional non-HPV associated cancers, which normally have 5-year survival rates ranging between 30% and 50%. 19he association between tobacco abuse, the development of head and neck cancer and the adverse impact on survival outcomes has long been established. 20The repository used for this analysis does not capture smoking history at a statewide level and so an analysis of smoking impact could not be performed.However, the Princess Alexandra Hospital in Brisbane, Queensland, analysed 404 patients with OPSCC p16+, of which a proportion of these patients also forming part of the cohort for this analysis, and found that at diagnosis 30% (121/404) were current smokers, 32% (129/404) had never smoked and 38% (154/404) were former smokers.Current smokers had an inferior survival compared with never and former smokers with a HR 2.37 (95% CI; 1.26-4.45,P < 0.01) and 2.58 (95% CI; 1.40-4.73,P < 0.01), respectively. 21n multivariable analysis we found that for OPSCC p16+, increasing stage, older age (≥65 years) and increasing comorbidities (≥2) had an adverse impact on survival.We also found that disadvantaged SES had a worse survival, and the reasons for this are not clear but warrant further evaluation.We did not find sex or remoteness of residence had an adverse impact on survival.Chemoradiotherapy achieves high cure rates for HPV-associated OPSCC; however, these treatments are toxic and result in both acute and long-term toxicity, such as neuropathy, hearing loss, xerostomia and swallowing difficulties.3][24] Another promising option is the use of minimally invasive surgery, such as Transoral robotic surgery (TORS) or Transoral laser microsurgery (TLM) to treat the primary tumour, with or without neck surgery and then risk-adjusted post-operative radiotherapy.In a recent Phase II trial reported by Ferris et al., excellent oncologic and favourable functional outcomes were achieved in a group deemed as having intermediate risk OPSCC p16+, with primary TORS and post-operative radiotherapy. 25n conclusion, our findings confirm that OPSCC is now the most common mucosal head and neck cancer diagnosed in Queensland, having surpassed oral cavity cancer.The majority are HPV-associated (p16+), presenting with early-stage disease.Our findings confirm the favourable prognosis of this disease at a population-wide level, provide insight into the growing need for resources required to manage this disease and set the benchmark to which other jurisdictions can compare outcomes.

Table 1 .
Patient demographics and tumour stage at diagnosis of oropharyngeal squamous cell carcinoma by p16 status, 2015-2019, Queensland, Australia to oropharyngeal squamous cell carcinoma; p16+ refers to p16 positive, p16À refers to p16 negative.‡Age-standardised rate (Australian per 100,000 persons).§8th Edition American Joint Commission on Cancer (AJCC) classification/Union for International Cancer Control (UICC) TNM staging.© 2024 The Authors.Journal of Medical Imaging and Radiation Oncology published by John Wiley & Sons Australia, Ltd on behalf of Royal Australian and New Zealand College of Radiologists.

Table 2 .
Factors associated with mortality for patients with p16+ disease of oropharyngeal squamous cell carcinoma, 2015-2019, Queensland, Australia © 2024 The Authors.Journal of Medical Imaging and Radiation Oncology published by John Wiley & Sons Australia, Ltd on behalf of Royal Australian and New Zealand College of Radiologists.