Early or late recurrence? Little difference after all

October 1, 2004

Houston--Early or late recurrence after induction BCG therapy with or without maintenance BCG for high-risk bladder cancer confers a significant increased risk of death, a finding that contradicts the widely held belief that late recurrence has less adverse impact on survival, according to data from a large clinical database.

Patients whose cancer recurred more than 12 months after induction BCG had a mortality rate similar to that of patients who recurred within 12 months. The analysis also revealed a higher mortality for recurrent bladder cancer than previously reported in the literature.

The Southwest Oncology Group has reported that maintenance BCG reduces long-term recurrence and improves disease-related survival (J Urol 2000; 163:1124-9). Moreover, a meta-analysis of 24 randomized clinical trials showed that the risk of progression is reduced only in patients who receive maintenance BCG (J Urol 2002; 168:1964-70).

Risky delay The availability of new drugs and increased clinical experience with maintenance and combination immunotherapy could lead to a tendency to delay definitive local therapy, Dr. Lerner said. As a consequence, patients could be exposed to an increased risk of progression to muscle-invasive and possibly metastatic disease. Among patients who undergo cystectomy, long-term survival is significantly decreased with muscle-invasive versus pT1 or less cancer (J Urol 1998; 160:1285-90).

To test the hypothesis that early recurrence after induction BCG increases the risk of progression, Dr. Lerner and associates retrospectively reviewed outcomes in 501 patients who had resected Ta or T1 tumors, with or without carcinoma in situ. All patients received induction BCG and were randomized to maintenance BCG therapy or to observation.

Induction BCG consisted of six weekly intravesical instillations and concomitant percutaneous BCG. Patients randomized to maintenance therapy were scheduled to receive three weekly instillations plus percutaneous BCG at 3 and 6 months after induction and then every 6 months for as long as 3 years from the start of induction.

Among patients with CIS, complete response was defined as no evidence of cancer on biopsy and negative cytology. For patients with Ta and T1 cancer, criteria for complete response were negative cystoscopy and cytology.

Timing isn't everything During follow-up, 251 of 501 evaluable patients had recurrences, and 229 died. An adjusted proportional hazards model of survival showed that recurrence at any time was the most potent predictor of mortality. Patients who recurred early had a hazard ratio of 2.37 (p=.0001) compared with those who did not progress or relapse, whereas late recurrence was associated with a hazard ratio of 2.57 (p<.0001).

The analysis yielded no evidence that BCG affected the relationship between timing of relapse and survival. Progression to T2 or more advanced cancer occurred in a similar proportion of patients with early (30%) or late (27.5%) recurrence.

Relatively few patients who recurred underwent cystectomy (56 of 251). Timing of recurrence had little influence on the likelihood of cystectomy, as 27% of patients with early recurrence had the surgery versus 18% of those who recurred late (p=.075).

Patients on maintenance BCG underwent cystectomy earlier-a median of 11 months compared with 24 months for patients not on maintenance therapy. A similar proportion of patients in the two treatment arms with T2 or more advanced disease had cystectomy (45% vs. 49%).