Anal carcinoma:: Prognostic value of endorectal ultrasound (ERUS).: Results of a prospective multicenter study

Background and Study Aims: The classification of anal carcinoma is based on the clinical examination and the estimation of the tumor height (Union Internationale Contre le Cancer (UICC) 1987 Classification). This classification has a direct therapeutic application since tumors which are designated T1 and T2 are generally treated by radiotherapy whereas T3, T4 or N+ lesions are treated by concomitant radiation and chemotherapy. The aim of this prospective multicenter study was to evaluate endorectal ultrasound (ERUS) and to define an ERUS-based classification. Patients nad Methods: Between January 1994 and May 1997, 146 patients (42 men and 104 women; mean age, 63) from eight different centers were studied prospectively. The ERUS classification incorporates disease of the anal canal and the perirectal lymph nodes, thus: usT1 describes involvement of the mucosa and submucosa with sparing of the internal sphincter; usT2, involvement of the internal sphincter with sparing of the external sphincter; usT3, involvement of the external sphincter; usT4, involvement of a pelvic organ; NO describes no suspicious perirectal lymph nodes, and N+, perirectal lymph nodes fulfilling endosonographic criteria for malignancy (e.g. round, hypoechoic). Tumors classified as UICC T1-T2 (<4 cm) NO were treated by radiotherapy alone, whereas lesions with a UICC classification of T2 (> 4 cm), T3-T4, N0-N1-2-3 received combined radiochemotherapy. Results: Data concerning the treatment and follow-up were available for 115/146 patients (78.7%). We compared the prognostic importance of the two classification schemes for treatment response and the rate of local relapse (chi-squared test). A significantly greater proportion of T1-T2N0 lesions classified by ERUS had a complete response to treatment than those classified by conventional UICC staging (94.5 % vs. 80 %, respectively; P = 0.008). The ERUS T and N stage were significant predictors of relapse (P = 0.001 and P = 0.03, respectively) whereas the corresponding clinical (UICC) stages were not (P = 0.4 and P = 0.5, respectively). Using a Cox model, usT stage was the only significant predictive factor for patient survival. Conclusion: This muticenter prospective study demonstrated the superiority of ERUS-based staging over traditional clinical staging in the prediction of important outcomes such as local tumor recurrence and patient survival.