In the current study, we report real-world data from a large monocentric cohort of 188 patients with a median follow-up of 83 months, all treated with (chemo)radiotherapy for anal carcinoma. Special focus was placed on sex-specific differences. Additionally, the prognostic significance of the recently updated AJCC staging system (now in its 9th edition) was tested.

Over 70% of our patients were female. The 5‑year OS rate in our study was significantly better for women at 82.9% compared to men at 59.9%. Similarly, male sex was identified as a significant risk factor for both CFS and DFS. Regarding FFR, a trend toward poorer FFR outcomes for male patients was observed, with a p-value of 0.09. Given the small number of events, we assume that a larger cohort might have demonstrated a statistically significant difference. Overall, these findings are in line with previously published data [13,14,15,16,17,18]. Arora et al. reported a median OS of 148 months for white women, 146 months for black women, 111 months for white men, and 82 months for black men [14]. Similarly, Koerber et al. demonstrated a 2-year OS of 63.5% for male and 89.5% for female patients in a cohort also treated in Germany [18]. Significant sex differences in terms of incidence and oncological endpoints have also been reported for a variety of other solid tumors. For example, Cook et al. showed high male-to-female incidence rate ratios (IRR) for numerous oncological diseases, with an IRR of 0.81 for anal carcinoma, indicating a higher incidence in women [2]. In another study, they also demonstrated elevated male-to-female mortality rate ratios (MMR) for a variety of tumors, with the most pronounced differences observed in lip cancer (MMR 5.5), hypopharyngeal cancer (MMR 4.5), esophageal cancer (MMR 4.1), and bladder cancer (MMR 3.4) [19]. Nakamura et al. demonstrated in a large meta-analysis of over 86,000 patients that female patients with non-small cell lung cancer had significantly better OS compared to male patients, regardless of tumor stage, histology, or smoking status [21]. Additionally, significant sex differences in survival favoring women were observed for oropharyngeal carcinomas, malignant melanomas, and colorectal carcinomas, among others [20, 23, 24]. The differences described in the literature highlight the need for sex- and gender-sensitive medicine. As early as 2015, the National Institutes of Health (NIH) called for the consideration of sex in the design and analysis of all studies [25]. Gender medicine encompasses all facets of sex-specific differences, considering not only biological distinctions but also varying gender identities and roles. The observed differences in incidence and oncological outcomes between men and women have been partially attributed to gender-specific variations in the use of healthcare services, such as screening programs. In our cohort, the proportion of female patients in AJCC stage I was significantly higher at 19.4% compared to 7.4% for male patients, suggesting that women may seek medical attention earlier.

Additionally, there is an evident difference in sex chromosomes, even in non-sex-related cancers, as well as in the varying levels of sex hormones [22]. Thus, the interplay between sex chromosomes and hormones impacts both the local drivers of carcinogenesis, including cancer-initiating cells and the components of the tumor microenvironment, as well as systemic factors such as cellular metabolism and the immune system [26]. Furthermore, significant sex-specific differences in the pharmacokinetics of many antitumor drugs have been documented. For instance, Mueller et al. reported a 26% higher elimination rate of 5‑FU in men compared to women [27]. Moreover, there are significant differences in body composition between sexes, with men having a substantially higher percentage of metabolically active fat-free body mass compared to women of the same height and weight [28]. Since calculation of the individual 5‑FU dose is based on body surface area, there are significant sex-specific differences in the circulating concentration profile. This impacts both the efficacy and toxicity of the treatment.

In addition to the sex-specific differences in oncological outcomes, our study also revealed a distinct difference in the prognostic value of the AJCC staging system within our cohort. For male patients, there was a strong trend towards better OS, DFS, and FFR, along with significantly better CFS for tumors in earlier stages. In contrast, the categorized AJCC stage (early/intermediate/advanced tumors) did not serve as a relevant prognostic factor in female patients.

As previously mentioned, the TNM and AJCC staging systems are intended, among other things, to function as prognostic tools to predict oncological outcomes. However, as stated by Janczewski et al., there was a lack of hierarchical prognostic order in the 8th version of the AJCC staging system, with 5‑year survival rates of 84.4%, 77.4%, 63.7%, 73.0%, 58.4%, 59.9%, and 22.5% for stages I, IIA, IIB, IIIA, IIIB, IIIC, and IV, respectively [11]. In the revised and recently published 9th edition, tumor stages IIB, IIIA, and IIIC have been updated, assigning greater importance to the T stage. Application of the 9th edition now shows a good hierarchical order, with 5‑year OS rates of 84.4%, 77.4%, 73.0%, 62.1%, 58.4%, 56.9%, and 22.5% for stages I, IIA, IIB, IIIA, IIIB, IIIC, and IV, respectively [11]. In the study by Janczewski et al., as well as in our cohort, women constitute the majority of patients, with 64.8% and 71.3%, respectively. Regarding ethnicity, 79% of the patients were non-Hispanic whites. The stage distribution patterns within the cohorts are similar, although we excluded stage IV patients from our analysis. A significant difference between the cohorts lies in the period during which patients were included in the analysis. Janczewski et al. included patients diagnosed between 2012 and 2017, while our cohort includes patients treated at our clinic between 2004 and 2020. Although the imaging modalities used for staging are not specified, the different inclusion periods suggest that a higher proportion of patients may have undergone MRI and/or PET/CT for staging. Given the known superior sensitivity and specificity of MRI and PET/CT compared to conventional CT alone, a stage shift may have occurred due to lymph node metastases not detected by CT alone. However, this factor alone does not explain the differing prognostic value of the AJCC stage for men and women in our cohort.

The main limitations of our study are the relatively small sample size and its retrospective character. There is evidence, for example from Grabenbauer et al., suggesting that tumors originating from the perianal skin are clearly associated with a worse prognosis [29]. Due to the retrospective nature of our analysis, a precise distinction between anal canal and anal margin carcinomas was not possible. Consequently, we cannot make any statements regarding the distribution of anal canal and anal margin carcinomas between men and women in our cohort or assess whether this had an influence on the oncological endpoints.

Notably, the limited number of male patients in the low-risk group, coupled with the absence of events in this group, affects the statistical significance of our findings. However, we anticipate that with a larger sample size or more events in this group, our analysis would reach statistical significance with respect to tumor stages. To address why the AJCC staging performed poorly in female patients in our cohort, we can only propose potential explanations at this stage. A possible reason might be the excellent 5‑year overall survival rate of nearly 83% across all stages for female patients. Such a high survival rate might render stage-based prognostic subdivision more challenging or potentially require a significantly larger cohort than ours to achieve statistical significance. As described, the retrospective analysis was conducted over an extended period. Consequently, there were variations beyond the investigated radiation techniques (3D-CRT vs. IMRT), which showed no significant difference. Notably, there were additional variations, particularly in target delineation and dose prescription, that were not specifically analyzed in the present study but which could potentially influence oncological outcomes.

Yang et al. developed a prognostic nomogram to predict 1‑, 3‑, and 5‑year survival incorporating the AJCC stage, sex, age, and whether or not radiotherapy was performed [30]. The nomogram, with a C-index of 0.684 and 0.730 in the training and validation groups, respectively, outperformed the AJCC stage (with C‑indices of 0.610 and 0.659), supporting the hypothesis that sex should be considered in prognostic assessments.