INTRODUCTION
Secondary extramammary Paget disease (s-EMPD) represents anal canal and rectal, bladder, and gynecological cancers, which extend horizontally within the epidermis of the anal and vulvar skin.1 It is necessary to distinguish this condition from primary extramammary Paget disease (p-EMPD), which occurs primarily in genital and perianal areas and occasionally extends to the urethra, vagina, or anus. p-EMPD and s-EMPD have different causes and require different treatment and management. However, clinical features of p-EMPD and s-EMPD remain to be fully elucidated. This study aimed to investigate the clinical and histopathological features of these two conditions in the anorectal region and to identify useful features for differentiation.
METHODS
We retrospectively analyzed perianal skin lesions in 16 patients who were suspected as EMPD in Shinshu University Hospital between 2009 and 2022. This study was approved by the ethics committee of the Shinshu University School of Medicine (#5676) and was conducted in accordance with institutional review board guidelines. Six patients were diagnosed with p-EMPD, and 10 were diagnosed with s-EMPD derived from anal canal adenocarcinoma. We performed CK-7, CK-20, GCDFP-15, and CDX-2 immunostaining. Clinical and histopathological features were collected through electronic medical records. The differences between the two conditions were compared at the following perspectives: protuberance, continuity to the mucosa, circumscription, symmetry of perianal skin lesions, boundary clarity (dermal side), and association of depigmentation. Protuberant lesion was defined as those with nodules or foci, not limited to a percentage of the lesion. A skin rash was defined as circumscribed type if it was present in more than 90% of the 360° directions centered on the anus. We have not included depigmentation in our determination of whether or not the circumscribed. Symmetry was defined as mirror-image symmetry with respect to a straight line containing the anus. Statistical significance for clinical findings between the two groups was evaluated using χ2 test. For assessment of symmetry around the anus, the coefficient of variation was calculated from the lengths in the four vertical and horizontal directions from the anus. Student t test was performed to evaluate the significance between the two groups. Statistical significance was set at p < 0.05.
RESULTS
There were more males in both cohorts, and the s-EMPD cohort had a higher proportion of males. The p-EMPD cohort had a longer disease duration than the s-EMPD cohort (average duration of 27.3 and 9.7 months, respectively). Skin symptoms, such as itchiness and pain, were observed in two of six (33%) of the patients with p-EMPD and one of 10 (10%) of the patients with s-EMPD. Defecation disturbance, such as constipation and bleeding during defecation, were observed in five of 10 (50%) of the patients with s-EMPD and zero of six (0%) of the patients with p-EMPD (Table 1).
Six and 10 of the perianal lesions associated with p-EMPD and s-EMPD, respectively, were analyzed by their characteristics, including circumference, symmetry, radiality, surface characteristics, prominence, and border distinctiveness. A previous study reported circumferential and symmetrical perianal distribution, a focal or papillomatous elevation of the margins of the rash, and a clear cutaneous border of the tumor as characteristics of s-EMPD.2 The present study, however, found no significant difference in the circumference of the rash between the two groups (p = 0.55; Table 1). On the other hand, nine of 10 (90%) patients with s-EMPD had symmetric perianal skin lesions, whereas all patients with p-EMPD had asymmetric skin lesions (p = 0.0004) (Table 1; Figure 1). When perianal symmetry between the two groups was compared, s-EMPD had a smaller coefficient of variation of length in four directions from the anus (average = 0.35) than p-EMPD (average coefficient of variation = 0.62; p = 0.048) (Figure 2); thus, these findings suggest that s-EMPD was more symmetric around the anus. The percentage of patients with raised lesions, such as foci or nodules, was nine of 10 (90%) for s-EMPD and one of six (16%) for p-EMPD (p = 0.003; Table 1). Well-defined tumor borders on the lateral margin were identified in five of 10 (50%) s-EMPD cases, and zero of six (0%) of p-EMPD borders tended to be clearer in s-EMPD; however, the difference was not significant (p = 0.078; Table 1). Depigmented lesions were observed in three of six (50%) of p-EMPD cases and four of 10 (40%) of s-EMPD cases, with no significant difference. Moreover, inclusion of depigmentation did not affect assessment of circumscribed or symmetrical perianal distribution.
Rectal examination/proctoscopy was performed in six of 10 of the patients with s-EMPD, and all six patients had no obvious tumor findings. Lower gastrointestinal endoscopy was performed in eight of 10 patients. Of these eight patients, only one patient (12.5%) had a primary advanced cancer on lower gastrointestinal endoscopy. The remaining seven of eight patients (87.5%) had primary early-stage cancer, but endoscopy showed no tumor findings (Table 1).
In histopathological findings, a previous study reported that fibroepithelioma with Pinkus-like changes and peripheral nuclear palisading is characteristic of s-EMPD and that the coexistence of subepidermal mucin deposition with fibroepithelioma is specific for s-EMPD.3In this study, similar epithelial changes were observed in five of 10 (50%) patients with s-EMPD; however, they were not observed in p-EMPD (Table 1). Fibroepithelioma with Pinkus-like changes and peripheral nuclear palisading were the histological findings of the protuberant lesions in five of five (100%) patients with s-EMPD.
4 DISCUSSION
Visual inspection of the anus, proctoscopic/rectal examination, or lower gastrointestinal endoscopy is often insufficient for diagnosis of s-EMPD associated with anal canal cancer, especially in the case of anal adenocarcinoma in situ.4 Therefore, observation of perianal skin lesions and histopathological features may be critical to find latent anal canal adenocarcinomas. To the best of our knowledge, there are few studies describing the histopathological and clinical features.3 This study focused on the clinical features of the perianal lesions of s-EMPD in detail and proposes differential points to aid in distinguishing it from p-EMPD. In particular, protuberant and well-defined lesions as well as perianal symmetrical spread are important in the diagnosis of s-EMPD. The protuberant lesions of s-EMPD may be considered to reflect pathological change of fibtoepithelioma with Pinkus-like changes. In the present study, fibroepithelioma with Pinkus-like changes without peripheral nuclear palisading was observed in two of six (33%) patients with p-EMPD (data not shown), which is consistent with the previous study.3 In addition, subepidermal mucus retention was observed in five of six (83.3%) p-EMPD cases and three of 10 (30%) s-EMPD cases; thus, our findings suggest that it was not characteristic of s-EMPD.
In conclusion, we propose that s-EMPD should be considered when anal skin lesions are raised and well defined, as well as in cases of symmetrical spread from the anus.
CONFLICT OF INTEREST STATEMENT
None declared.
Authors: Megumi Kosano, Yukiko Kiniwa, Asuka Mikoshiba, Mai Iwaya, Ryuhei Okuyama
Department of Dermatology, Shinshu University School of Medicine, Matsumoto, Japan
The Journal of Dermatology, John Wiley & Sons Australia, Ltd, on behalf of Japanese Dermatological Association.