Cytokine gene expression in a mouse model: the first instillations with viable bacillus Calmette-Guerin determine the succeeding Th1 response

Abstract

Purpose: Bacillus Calmette-Guerin (BCG) therapy for superficial bladder cancer is immune dependent and activation of a Th1 immune response is probably required for clinical efficacy. Given the empirical approach to improving BCG therapy we investigated in a mouse model the consequences of modifications in BCG therapy with regard to Th1 and Th2 cytokine responses in the bladder. These studies may provide a rationale for possible modifications of the established clinical treatment protocol.

Materials and methods: The dynamics of Th1 (interferon-gamma, interleukin [IL]-2, IL-12p40 and tumor necrosis factor-alpha) and Th2 (IL-10 and IL-4) cytokine responses during and after 6 once weekly intravesical BCG instillations in mice was determined by a semiquantitative reverse transcriptase-polymerase chain reaction based method.

Results: During 6 weekly BCG instillations a dose and time dependent induction of the various Th1 as well as Th2 cytokines was observed. The response pattern was comparable to urinary cytokine induction patterns in patients. Electrocauterization prior to BCG instillations led to lower and more variable levels of cytokine polymerase chain reaction products compared with noncauterization. Lowering the dose of BCG seemed to affect the Th1 cytokine response most, whereas the Th2 response was less influenced by dilution of the BCG preparation. Six instillations with nonviable BCG induced Th2 but failed to induce Th1 cytokines, which may explain the necessity of BCG viability for antitumor activity. However, when mice were first treated 3 times with viable BCG and subsequently received 3 instillations with killed BCG, the Th1 and Th2 cytokine pattern was comparable to the standard 6-week regimen with viable BCG.

Conclusions: The model seems an appropriate one in which to investigate changes in Th1 and Th2 cytokine gene expression levels in bladders resulting from modifications in intravesical BCG treatment. It was possible to induce a local Th1 cytokine response with nonviable BCG provided that local sensitization to BCG antigens had occurred during preceding instillations with a viable BCG preparation. Whether such an approach could decrease BCG therapy toxicity, while maintaining antitumor efficacy, remains to be further investigated in patients.