Trends in epidemiology and primary treatment of anal squamous cell carcinoma in the Netherlands (1990–2021)

A rapid increase in the incidence of anal squamous cell carcinoma (SCC) was reported in several countries over the past decades. This study assessed trends in epidemiology and primary treatment over a 32‐year period (1990–2021) using the Netherlands Cancer Registry. The study population included 4273 patients, 44.2% male and 55.8% female (median age 63 years). The age‐standardised incidence rate (European Standardised Rate, ESR) increased from 0.5 to 1.6 per 100,000, which entailed an average annual percentage change (AAPC) of 5.0% (95% confidence interval [CI]: 4.5%–5.8%). While incidence among females increased continuously over the total period (AAPC 4.9%; 95%CI: 4.4%–5.6%), to 1.8 per 100,000 ESR in 2021, incidence among males increased until 2016 (annual percentage change [APC] of 6.3%; 95%CI: 5.6%–10.7%), after which it seemed to stabilise (APC −2.1%; 95%CI: −16.8%–4.5%). Significant trends were also observed in distribution of age, tumour stage and primary treatment modalities. Five‐year relative survival (RS) was estimated using the Pohar–Perme estimator, and this improved from 56.1% in 1990–1997 (95%CI: 49.3%–62.4%) to 67.9% in 2014–2021 (95%CI: 64.7%–70.9%), but remained poor for stage IV disease. Evaluation through a multivariable Poisson regression model demonstrated diagnosis in the most recent period to be independently associated with better RS, in addition to female sex, younger age, early disease stage and any treatment. In conclusion, the rising incidence of anal SCC seems to decline in males, but not in females, and advances in diagnostics and therapeutic management have likely contributed to improved prognosis.

among Dutch women and among both men and women over age 65.For men younger than 54, incidence increased until 2016, then stabilized, survival rates improved across all disease stages, excluding stage IV, possibly owing to diagnostic and therapeutic advances.

| INTRODUCTION
Squamous cell carcinoma (SCC) constitutes the most common histological type of anal cancer. 1 It arises from precursor high-grade squamous intraepithelial lesions (HSIL), and may develop throughout the anal canal, particularly along the transitional zone between the squamous and columnar epithelium.It has been hypothesised that epithelial cells in this zone have a higher turnover which translates into an elevated risk of acquiring malignancy-inducing genetic alterations. 2tients with anal SCC most often present with a palpable mass, pain, rectal bleeding, and pruritus, and may initially be misdiagnosed with a benign anorectal condition such as haemorrhoids. 3though classified as a rare cancer, 4 a rapidly rising incidence of anal SCC has been reported over the past decades, particularly in several high-income countries in the Americas, Australia, and Northern and Western Europe, [5][6][7] including the Netherlands. 8In almost all populations studied, incidence was higher in females than in males.Furthermore, rates were highest and showed the steepest increase in patients aged 50 years and older.0][11] Decreasing incidence specifically among younger males has also been reported. 12e observed incidence patterns of anal SCC have been linked to a rising prevalence of known risk factors aside from higher age, the most prominent being infection with high-risk human papillomavirus (hr-HPV). 13More than 85% of anal SCC cases are associated with hr-HPV. 14,15In women, the anatomical proximity of the anal and genital areas favours HPV cross-site transmission, hence increasing the risk of developing anal SCC. 16In addition, this risk is increased in particular populations, for instance among men who have sex with men, immunocompromised individuals, and patients with other HPV-related cancer or precancerous lesions. 17nagement of anal SCC has changed considerably during the past decades.9][20] Standard of care for patients with local or locoregional disease consists of firstline chemoradiotherapy (CRT): radiotherapy (RT) preferably combined with 5-fluorouracil (5FU) or capecitabine and other cytotoxic agents, usually mitomycin C (MMC).The majority of patients who undergo this regimen will show a clinically complete response and do not require extensive surgery such as an abdominal-perineal resection (APR), including an end-colostomy.Nevertheless, this procedure still has a role as salvage treatment or in patients not amenable to CRT (e.g., previous irradiation).Local excision is only considered suitable as definitive treatment for a subgroup of early-stage cT1 lesions for which radical resection can be obtained without major morbidity, located in the anal canal and perianal margin.
The aim of this study was to assess the trends in both epidemiology and primary treatment of anal SCC in the Netherlands between 1990 and 2021.Besides confirmation of the abovementioned developments in the Dutch population, the analyses centred on whether the incidence of anal SCC continued to rise in the most recent years, and whether trends have translated to changes in patients' prognosis over time.

| Data source
Patient records were retrieved from the Netherlands Cancer Registry (NCR), hosted by the Netherlands Comprehensive Cancer Organization (IKNL).This population-based registry contains data on all newly diagnosed cases of cancer in the Netherlands, including information on patient and disease characteristics and on therapeutic interventions as part of primary treatment, registered according to international coding systems.These include the International Classification of Diseases for Oncology (ICD-O) for categorising the primary site (topography) and histology (morphology), and the TNM system for staging of malignant diseases.Annual linkage with the Municipal Personal Records Database (Gemeentelijke Basisadministratie, GBA) establishes follow-up information on vital status for every patient in the database, which was updated until 31 January 2023 for this study.
Patients under 18 years at diagnosis were excluded, as well as foreign residents and patients diagnosed abroad.Also excluded were cases with incomplete registries, and secondary primary tumours according to the reporting rule for multiple primary cancers as maintained by the International Agency for Research on Cancer (IARC) and the International Association of Cancer Registries (IACR). 21

| Statistical analysis
Baseline characteristics were summarised by frequencies for categorical variables, and by medians including interquartile ranges for continuous variables.Age at diagnosis was studied both continuously and in groups (18-54/55-64/65-74 and ≥75 years).As the NCR only contained data on therapies provided as part of the initial treatment plan for the period under study, no information was available on second-or higher line treatment, including salvage APR after failed CRT.While main treatment modality, for example, surgery, was registered for the whole study period, the type of procedure (local excision or APR) was only specified after 2005.
Incidence rates (per 100,000 person-years) were calculated by dividing the number of patients of each incidence year by the corresponding annual Dutch population (provided by Statistics Netherlands/Centraal Bureau voor de Statistiek, CBS), stratified by 5-year age groups.The rates were age-adjusted through standardisation to the European standard population (European Standardised Rate, ESR), and trends were evaluated using joinpoint regression analyses. 22,23Significant pattern changes, or joinpoints, were identified through a Monte Carlo permutation method and evaluation of the Bayesian Information Criterion (BIC).From the analyses, average annual percent changes (AAPC) were calculated for the total study period, and annual percent changes (APC) for time segments between joinpoints.Trends in baseline characteristics and primary treatment were examined by 4-year incidence periods, and assessed by calculating a chi-squared test statistic for Pearson's correlation coefficient.
Relative survival (RS) within 5 years following diagnosis of anal SCC was used to approximate disease-specific survival in the study population.RS was calculated by matching the observed probability of overall survival among patients to the expected probability of survival within the corresponding Dutch population, according to age, sex and calendar year (retrieved from CBS) using the Pohar-Perme estimator. 24RS rates were age-standardised through the weights defined by the International Cancer Survival Standard (ICSS). 25Trends in RS over time were assessed by 8-year periods.
Excess mortality (as mortality analogue of RS) was further modelled using multivariable Poisson regression analyses.Independent predictors were evaluated following their inclusion on the basis of a p-value of <.1 in univariable analyses and definitive selection based on the BIC and likelihood ratio tests comparing nested models.Their effects were presented as excess hazard ratio's (EHR's) against reference categories for categorical variables.In all statistical analyses, two-sided p-values <.05 were considered significant and confidence intervals (CI) were calculated at a confidence level of 95%.Analyses

| Trends in characteristics and treatment
The

| Trends in survival
Prognosis improved for patients with anal SCC: the  and after the start of the pandemic event, contrary to colorectal cancer. 27,28Similar trends in males were recently reported across the United States of America, which were linked to improved care for HIV patients and smoking patterns in some states. 29The current study findings are assumedly also linked to the changing dynamics surrounding HIV infection and care.They anyhow call for close monitoring in the coming years to determine whether the increasing incidence of anal SCC has indeed been halted, at least for some subgroups.
Tumour staging improved over time, which is demonstrated by the reduction of patients with unknown disease stage.This may be attributed to more frequent and improved imaging, for instance through the combined use of positron emission tomography (PET) imaging and computed tomography (CT) or magnetic resonance imaging (MRI).The increased proportion of stage III disease may then be due to better detection of tumours invading adjacent organs or lymph nodes being affected.
Over two-thirds of patients is nowadays treated initially with CRT, while the use of RT only has decreased.1][32] Patients fare better in terms of being less likely to require a colostomy and showing lower recurrence and higher survival rates. 33Following careful patient selection, APR is mainly reserved as salvage procedure after failed CRT, and is only performed in a minority of patients as primary treatment, mostly when (chemo)radiation is not possible (e.g., after previous irradiation).
5][36] The period of diagnosis remained independently associated with better survival even after adjustment for primary treatment modalities as well as patients' sex, age and tumour stage, which likely indicates improvements in salvage or therapeutic management of recurrent disease.Also in line with previous findings is the lower excess mortality found among female patients compared to males. 7,9Unlike other analyses, the impact of tumour differentiation was largely diminished in the multivariable analyses. 36 for many other malignancies, further advances are anticipated with a role for novel immunotherapeutic agents, 37,38 Specifically those targeting the programmed cell death receptor 1 (PD-1) have shown promising results in patients with metastatic and surgically unresectable recurrent anal SCC. 39,40Anti-PD-1 therapies are currently being studied in patients with locally advanced disease (ClinicalTrials.govidentifiers NCT03233711 and NCT02628067).
Given its rarity, systematic screening for anal SCC is not recommended, and even screening focused on high-risk groups has been subject for debate. 41,42Therefore, vaccination against HPV, particularly the high-risk variants 16 and 18, is considered a mainstay to reduce incidence rates of anal SCC and its precursor lesions on the long term. 43Indeed, the implementation of a national vaccination program in Denmark may have contributed to the decrease in incidence of HSIL that was recently observed. 9In the Netherlands, a multivariable analyses, omitting them would introduce other sources of bias since they were not allocated at random.Lastly, due to the lack of data on cancer recurrence or progression of disease, the study could not assess progression-free survival or the benefits of salvage treatment.
The study's main strength lies in the use of a population-based registry with a high completeness that included information on treatment since its onset.This enabled accurate evaluation of incidence and survival estimates for anal SCC in unselected patients over a considerable time period, and allowed for an overall assessment of their initial management.
In conclusion, the rising incidence of anal SCC seems to decline in males, but not in females, and advances in diagnostics and therapeutic management have likely contributed to improved prognosis.
were performed with Stata version 17.0 (StataCorp, College Station, Texas) and the Joinpoint regression program, version 5.0.2 (National Cancer Institute, USA).

2 F I G U R E 3
Joinpoint regressions on age-specific incidence rates of anal SCC: (A) males; (B) females.AAPC, average annual percentage change; APC, annual percentage change; CI, confidence interval.This population-based study showed that the decade-long rise in anal SCC incidence seems to stabilise among males in the Netherlands, but not in females.This course primarily occurred in males aged 18-54 years, as incidence rates for other age groups continued to rise.While incidence also slowed down among the youngest age group, it seemed to speed up among those aged 65-74 years.Although caution is warranted given the COVID-19 pandemic-which reached the Netherlands in February 2020-,26 the abovementioned trends started well before its onset.Furthermore, no significant changes in stage distribution were observed before1994-1997 1998-2001 2002-2005 2006-2009 2010-2013 2014-2017 2018-2021   Distribution according to 4-year time periods for patients with anal SCC for: (A) tumour stage; (B) primary treatment modality.APR, abdominal-perineal resection; CRT, chemoradiotherapy; RT, radiotherapy.
Baseline characteristics of patients diagnosed with anal SCC in the Netherlands in 1990-2021.
Univariable and multivariable analysis on factors associated with excess mortality among patients with anal SCC.Surgery comprised all procedures performed as part of primary treatment, including local excision, APR, and not specified surgery until 2005.Abbreviations: CI, confidence interval; CRT, chemoradiotherapy; EHR, excess hazard ratio; RT, radiotherapy.
reaching the age of 10 years are offered vaccination, and catch-up vaccination is offered to anyone up to18-year (2022-2023)and to19-27 year-olds (2023).However, whether this program will indeed reduce the incidence of anal SCC remains to be determined.Some limitations of this study should be acknowledged.As it concerns an observational study, potential confounding by indication needs to be considered when interpreting the reported associations between treatments and outcome.Although these were adjusted in multivariable models, residual confounding cannot be precluded since data on patients' health status and more detailed information on tumour characteristics were unavailable for analyses.No information was recorded on HPV status, previous HSIL or malignancy, HIV or other immune-compromising conditions.While treatments should be considered intermediate variables, potentially adding bias to the T A B L E 2