The current problems with BCG immunotherapy pertain to efficacy and toxicity.
In spite of its remarkable success, a significant proportion of patients does not respond initially or will relapse later. Also, toxicity albeit self-limiting in many instances, curtails its widespread use despite its proven efficacy. The original therapeutic dose and schedule were derived empirically, the latter by mimicking that for intravesical chemotherapy. The possibility exists that modification of the dose and treatment schedule may help to improve efficacy and toxicity.
1.9.1 Schedule
The original schedule of 6 once-a-week intravesical instillations of BCG was not inferior to schedules of 6 instillations followed by one dose of BCG 3-monthly or 1 dose of BCG monthly for up to 2 years (Hudson et al 1987; Badalament et al 1987). Tachibana also reported that a randomized trial using a monthly schedule of 12 or 18 BCG instillations demonstrated no advantage for additional monthly instillations (as quoted by O’Donnell MA 2001). Six instillations 6-monthly for 2 years have also showed no advantage.
(Palou et al 1997). Also, Lamm has reported that a continuous 12-week course of BCG caused a reduction in immune stimulation in patients as measured by in vitro immunoproliferative assays (Lamm 1995).
The first controlled trial to show a benefit with additional BCG was a study in patients with CIS who received 81 mg of Connaught strain BCG weekly either in a schedule of 6 instillations or 6+3 instillations (the additional 3 instillations being given after a 6-week rest interval) of using. In this study, the “6+3” schedule had a significantly higher 82%
response rate at 6 months compared with 70% in the 6-instillation arm (Lamm 1992).
Several years after the start of our clinical trial, data from a trial designed to address the issue of additional (maintenance) BCG therapy, SWOG 8507, was published. In this study, patients received a “6+3” course of instillations followed by 3 weekly instillations every 6 months for a total of 3 years. The findings of this study will be reported in the Discussion.
1.9.2 Dose
1.9.2.1 BCG dose reduction – when more isn't better!
Studies in murine transitional cell carcinoma have demonstrated a typical bell-shaped dose-response curve for BCG immunotherapy (Lamm et al 1982). However, with repeated administration, cumulative doses of BCG, which would, if given in one instillation, have suppressed the immune response, have resulted in heightened antitumour response. Clearly, optimal dose is dependent upon the schedule of administration. Blumenstein et al (unpublished data from Connaught Laboratories) found that variations in Connaught BCG between 0.83 -19.39 X 108 CFUs resulted in no detectable range in efficacy. In animal work on the protective effect of Connaught BCG to challenge from tuberculosis, Siebenmann and Barbara (1974) showed that a plateau of protective efficacy was reached at a certain dose; greater doses did not improve efficacy. A likely explanation of Blumenstein's finding is that the
conventional dose of Connaught BCG is already on the plateau and could be lowered with no reduction in efficacy.
Data from controlled dose reduction trials suggest that dose reduction lowers the toxicity of immunotherapy. Halving the 120mg standard dose of Armand Frappier strain BCG in a 6 weekly instillation schedule decreased toxicity from 33% to 12% but reduced response from 67% to 37% at a mean follow-up of 21 months in patients with Ta, T1 or CIS tumours (Morales 1992). In a 6 weekly (repeated if no response) plus monthly single maintenance instillation schedule for 2 years, halving the standard dose of 150 mg of Pasteur strain BCG reduced the incidence of cystitis from 57% to 32%
and fever from 33% to 18% (Pagano et al 1995; Bassi et al 1999 and 2000). At a
median follow-up of 59 months, in 210 patients, the non-recurrence rates for the standard and low dose arms were 58% v. 56% for Ta (n.s.), 44% v. 53% for T1 (n.s.) and 30% v. 62% for CIS disease (p=0.006). Comparing conventional-dose Connaught BCG to one-third dose in an unblinded 500-patient study, Martinez-Pineiro found no difference in efficacy but significantly less toxicity with the reduced-dose treatment (Martinez-Pineiro 2003). This occurred in spite of using a 6 weekly induction schedule followed by 6 fortnightly instillations. Patients with multifocal tumours had a lower recurrence rate and progression rate with standard dose BCG but no statistically significant differences occurred for other high risk subgroups.
Hence, reducing the dose of BCG in schedules, which use more than 6 instillations, maintains efficacy whilst reducing toxicity in these controlled but unblended trials.
Possibly, the standard doses used for the various strains may be excessive and may contribute to toxicity.
1.10 BCG and Interferon alpha
There is sound immunological data to expect an enhanced antitumour response to a combination of BCG and interferon alpha. Although an initial study suggested little additive effect in the combination (Pryor et al 1995), our own studies as well as others, the combination had additive antiproliferative activity and cytotoxicity against a panel of human bladder cancer cell lines (Zhang et al 1997) and in enhancing cytokine production (Zhang et al 1999). These advantages translated to superior T cell activation and antitumour efficacy in a subcutaneous murine bladder cancer model (Gan et al 1999).
In a phase I/IIa trial, using 6 instillations of half dose Pasteur strain BCG (Tice) with interferon alpha-2b at 10MU, 30MU, 60MU and 100MU, Stricker et al (1996) reported good responses in 9 of 12 patients and good safety. These data however, required validation in a larger phase II study.