Differences between a probable and proven BCG infection following intravesical instillations: 16 years experience in a tertiary care hospital

Intravesical instillation of Bacillus Calmette-Guérin (BCG) is an efficient immunotherapy for superficial bladder cancer. BCG infection represents a major yet uncommon adverse event that occurs in 5-10% of the patients treated with BCG instillations, though the pathogenesis of this entity is not clear.

We report two cases of patients presented at our institution with BCG infection after instillation: one with microbiological BCG isolate and another without, and review all the medical records of patients instilled with BCG in our institution from 1996 until 2012, comparing patients with probable and proven BCG infection.

During the study period, a total of 786 patients received BCG intravesical instillations. Of them, 31 (4%) patients had to suspend treatment because of adverse events and, specifically, 11 (1.3%) patients had to interrupt treatment because of suspected BCG infection. The incidence of BCG infection during our study period was 0.87 episodes per 1,000 instilled patients/year and 140 cases per 10,000 instilled patients. Of the 11 patients with suspected BCG infection, 7 (64%) had a probable BCG infection, while 4 (36%) patients had a proven BCG infection. All patients with a proven infection had a previous underlying condition, compared to a high proportion of patients with probable infection (57%) that did not present with underlying diseases. Common findings between both groups of patients were abnormal imaging studies and laboratory tests. Regarding treatment, 8 (73%) of the 11 patients with BCG infection received at least two first line drugs active against M. bovis (isoniazid, rifampicin or ethambutol), four patients (36%) received steroids as part of the treatment and curation was obtained in 10 (91%) patients, while 1 patient with a proven infection had a death related to BCG infection.

We can conclude that BCG infection after intravesical instillations has a low incidence in our institution. Patients with previous underlying conditions seem to have more proven infections. A high proportion of patients do not yield positive microbiological tests; in those cases the diagnosis relies in clinical, radiological and laboratory findings. Treatment for BCG infection should include at least two active drugs against M. bovis and coadjuvant steroid treatment for systemic BCG infections.