Perianal Bowen's disease: a clinicopathologic study of 47 patients

Perianal Bowen's disease is an uncommon squamous-cell carcinoma in situ usually treated by surgical excision. There are controversies concerning surgical margin extent, because the disease is likely to recur in nonexcised skin areas of the anal and perianal skin. The aims of this study were 1) to determine the recurrence rate after different surgical treatments and 2) to determine if molecular markers might have a prognostic role in perianal Bowen's disease.

Retrospective chart review from 1972 to 1993 of 47 patients with perianal Bowen's disease was undertaken. Follow-up was obtained by office visits and/or phone questionnaire. Immunohistochemical analysis for p53 protein and Ki-67 nuclear antigen was conducted on fixed tissue specimens.

Twenty-six patients were treated by wide local excision with microscopic clearance of resection margins, 15 by local excision with only macroscopic clearance of resection margins, 5 by CO2 laser vaporization, and 1 by abdominoperineal resection because of fecal incontinence. Median follow-up for the entire population was 104 (range, 16-273) months. The incidence of local recurrence was 23.1 percent (6/26) after wide local excision, 53.3 percent (8/15) after local excision, and 80 percent (4/5) after CO2-laser vaporization. Recurrence rate estimated by Kaplan-Meier analysis is statistically different (P = 0.002) between radically treated patients (wide local excision/abdominoperineal resection; n = 27) and patients undergoing conservative treatment (local excision/laser vaporization; n = 20). Among patients with recurrence, the median time until recurrence was 38.5 (range, 3-89) months and 41.5 (range, 4-111) months after conservative and radical treatment, respectively. Nine of 20 (45 percent) patients in the conservative group and none of the 27 patients in the radical group had multiple episodes of recurrence (P < 0.001). In addition, 3 of 20 and 0 of 27 patients in the respective groups developed an invasive cancer (P = 0.034). Positive staining for p53 protein was observed in 12 (33.3 percent) of the 36 tissue specimens available for immunohistochemical analysis. Recurrence occurred in 9 of 24 (37.5 percent) patients negative for p53 and in 6 of 12 (50 percent) patients with positive p53 expression (P = not significant). Ki-67 antigen-graded expression from 1+ to 4+ did not reveal any correlation with incidence of recurrence. Recurrence rate did not differ by p53 and Ki-67 results, either in the overall group of 36 patients or stratified by surgical treatment groups.

Wide local excision for perianal Bowen's disease leads to a significantly lower recurrence rate than local excision or laser therapy. Follow-up longer than five years is recommended because of the risk of late recurrence. p53 protein and Ki-67 antigen immunohistologic expression may not have a prognostic role in perianal Bowen's disease.