We evaluated the prognostic value of purified protein derivative skin test reactivity and a granulomatous response in intravesical bacillus Calmette-Guerin therapy. We treated 62 patients with intravesical bacillus Calmette-Guerin once a week for 6 weeks. Purified protein derivative skin tests were performed before and after therapy. Cold cup bladder biopsies were examined in a blind retrospective manner for the presence of granulomas 6 weeks after completion of therapy. A significant correlation between status free of tumor and the presence of either granulomas or positive purified protein derivative skin tests was observed for the total patient population. Of 25 patients whose purified protein derivative test was converted from negative to positive 19 (77 per cent) remained free of tumor, while only 11 of 32 (34 per cent) whose test did not convert to purified protein derivative positive remained free of tumor (p equals 0.0006, chi-square). Similarly, 28 of 37 patients (77 per cent) who had a granulomatous response remained free of tumor, while only 8 of 25 (32 per cent) without a granulomatous response remained free of tumor (p less than 0.003, chi-square). The correlation was similar for each parameter when the total patient population was subdivided into patients treated for carcinoma in situ, residual tumor or prophylaxis. Calculation of predictive values showed that neither purified protein derivative responsiveness, granuloma formation nor a combination of both provided a highly accurate predictive index of therapeutic response in individual patients. False positive or negative rates, ranging from 23 to 24 per cent and 32 to 39 per cent, respectively, were observed. These results suggest that a link between immunological responsiveness and response to therapy exists but that neither the purified protein derivative skin test nor the granulomatous response exhibits sufficient immunological specificity to serve as accurate prognostic indicators in individual patients.
ASJC Scopus subject areas