Are pelvic side-wall recurrences of cervical cancer biologically different from central relapses?

Publisher Website
Google Scholar
Abstract
Background. By using the Combined Operative and Radiotherapeutic Treatment (CORT) procedure, pelvic side-wall recurrences of gynecologic malignancies arising in a previously irradiated pelvis may be locally controlled. Local control of central relapses may be achieved by exenteration alone. If, in cervical cancer, both relapse patterns are biologically different (as hypothesized by some investigators), distinct disease courses after local treatment may be expected. Methods. Since June, 1989, 32 pelvic recurrences of cervical cancer were treated for local control in this institution. The median size of the recurrent tumors was 5 cm (range, 2–9 cm); 84% of the patients had been extensively irradiated in the pelvis. Therapy of 14 centrally located recurrences was exenteration alone. In 18 patients with relapses fixed to the pelvic wall and histologically confirmed intralesional resection planes, the CORT procedure was applied. Results. After a median observation of 24 months (range, 5–48 months) 7 patients with central recurrences and 11 patients with pelvic wall recurrences had progressive disease in the pelvis and/or distantly. The site of recurrent tumor progression was similar in both groups. Stratified kaplan-Meier and univariate Cox regression analysis identified recurrent tumor size, age, and recurrence-free intervals, but not relapse location as prognostic factors. Only size of the recurrent tumor significantly influenced survival in the multivariate Cox analysis. Conclusions. These results suggest that central and pelvic wall recurrences of cervical cancer do not exhibit pronounced biologic differences. Patients with large (> 5 cm) recurrences have a poor prognosis in spite of extended radical treatment irrespective of tumor location. Efforts should be made to detect isolated pelvic relapses at a smaller tumor size to enhance the chance for long term survival after local control by exenteration and CORT. Cancer 1994; 74: 648-45