Intravesical administration of Bacillus Calmette-Guerin (BCG), a live attenuated strain of Mycobacterium bovis, has become a mainstay of adjunctive therapy for superficial bladder cancer. While generally well tolerated, both local and systemic infectious complications can arise. When disseminated BCG infection occurs, antituberculous therapy with or without glucocorticoids should be administered.
The infectious complications of BCG immunotherapy will be reviewed here. The clinical use of this agent for the treatment of superficial bladder cancer is discussed separately. Disseminated infection has also been rarely reported in patients receiving BCG vaccination for the prevention of tuberculosis . (See "Treatment of non-muscle invasive bladder cancer" and "BCG vaccination".)
The mechanism by which BCG leads to the development of infectious complications is not fully understood. Its mechanism of action as an immunotherapeutic agent in cancer is not fully known but recent evidence suggests that elaboration of a particular helper T cell cytokine profile known as the "Th1 response" is an integral part of its mechanism . (See "Treatment of non-muscle invasive bladder cancer".)
Considerable debate exists in the literature about whether infectious complications due to BCG represent a hypersensitivity reaction or ongoing active infection. The hypersensitivity hypothesis gained early credence based upon the presence of granulomas and the absence of recoverable organisms. In a number of case reports, acid-fast bacilli have not been demonstrated and organisms have not grown despite a high clinical suspicion of BCG infection [3,4]. A response to glucocorticoids, administered along with antituberculous drugs, has also supported the notion of a hypersensitivity response.
In contrast, other case reports have demonstrated viable organisms in a variety of tissues, including lung , liver [6,7], pancreas , psoas abscess contents , mycotic aneurysm , bone marrow , vitreous fluid [12,13], and even the brain . The fastidious growth nature of BCG in culture and a doubling time of 24 to 48 hours contribute to the difficulty in its isolation. M. bovis has also been demonstrated by PCR in some cases , although in other reports these studies have been negative . Further compounding this story are several cases of delayed infection, months to even years after the original BCG administration [16,17].