Anorectal fistula cancer is defined by the WHO classification as a tumor arising from the anal sinus or fistula [5]. It is thought that anorectal fistula cancer develops from the anal glands within the anal fistula and the epithelial components of the fistula ducts, becoming malignant when stimulated by inflammation over a long period of time [6, 7]. In general, early diagnosis and treatment of anorectal fistula cancer is difficult due to the clinical symptoms caused by anal fistulas. A case has been reported in which a preoperative pathological diagnosis was not made despite multiple biopsies and a diagnosis of malignancy was finally obtained from the resected specimen [8]. In our case, although the patient had been followed up regularly for more than 5 years with anal stenosis due to anorectal fistula, routine endoscopy and biopsy failed to diagnose the anorectal fistula cancer, eventually revealed by a close examination for abdominal pain and diarrhea. Several researchers have reported case series and established the following five diagnostic criteria for anal fistula malignancy: repeated inflammation over a long period of time (> 10 years), pain and induration in the region of the anorectal fistula, mucinous discharge, no primary cancer in other parts of the anorectal region, and anorectal fistula orifice in the anal canal or anal crypt [9, 10].

Although anorectal fistula cancer is a rare disease in terms of colorectal cancer as a whole, anorectal fistula cancer is not uncommon in CD because an anal fistula is the most common anal lesion associated with CD in Japan [11]. It has been reported that 15% of colorectal cancers associated with CD are anorectal fistula cancers [12]. Early disease onset, long-standing disease (> 10 years), severe chronic colitis, chronic fistula, and stenosis are important risk factors for carcinogenesis of a fistula tract [13]. Colorectal cancer associated with CD is more common in the right side of the colon in Western countries and in the left side of the colon in Japan [11]. It is particularly common in the rectum and anus [14, 15], reportedly accounting for 55% of cases [12]. In Japan, the mean age at cancer diagnosis is 58.3 years for anorectal fistula cancer overall and 38.9 years for cancers associated with CD [1, 11], indicating that anorectal fistula cancer associated with CD tends to develop at a relatively young age. The duration from the onset of anorectal fistula to the diagnosis of cancer appears to be 18.8 years for all anorectal fistula cancers and 17 years for cases associated with CD [1, 11], which is not significantly different. In this patient, though it had been more than 20 years from the onset of CD to the diagnosis of cancer, it was only 5 years from the onset of anorectal fistula to the diagnosis of cancer, a relatively short period of time.

Surgical resection is the standard treatment for anorectal fistula cancer and abdominoperineal resection is commonly performed [16]. One-third of resected cases were reported to be positive for resection margins, indicating the difficulty of radical resection in anorectal fistula cancer [1]. As it is not easy to determine the range of tumor infiltration intraoperatively, it is necessary to carefully examine the region to be resected using various preoperative imaging techniques before planning the surgery and to perform a wide local resection to avoid positive margins. Mucinous carcinoma, the main histological type of hemorrhoidal carcinoma, is known to have low sensitivity to chemotherapy [17, 18], but there have been several reported cases in which curative resection was performed after preoperative CRT for anorectal fistula carcinoma. Positive resection margins were reported in 14% of patients who were treated with preoperative CRT, and preoperative CRT may be useful to ensure negative resection margins [19]. In our case, even though the tumor had slightly shrunk on MRI with stable disease as defined by the revised RECIST guidelines (version 1.1) [3], curative resection was performed with extensive resection including the penis. Given that the floating carcinomas within the mucinous nodules were markedly reduced compared to the biopsy findings, preoperative CRT may have contributed to the negative resection margins. No large-scale trials have examined the efficacy of preoperative CRT for anorectal fistula cancer, and there are many factors to be considered, such as irradiation modalities, irradiation range, and chemotherapy regimen. In our case, we introduced the regimen that we generally apply to locally advanced rectal cancer. Considering the local control effect of preoperative CRT for rectal cancer, preoperative CRT for anorectal fistula cancer is expected to improve the negative margins, local control, and prognosis.