Reactive arthritis induced by intravesical BCG therapy for bladder cancer: our clinical experience and systematic review of the literature.

OBJECTIVE

Intravesical instillation of BCG (ivBCG) is an effective and safe immunotherapy of bladder carcinoma but it may have, as side effect, a reactive arthritis (ReA). The authors describe 5 cases observed during their own clinical experience along with the updated review of the literature on this topic.

METHODS

Seventy-three papers were present in the world literature, each reporting almost 1 case for a total of 112 patients. However, the review focused on 61 papers, selected on the basis of reporting suitable for a correct clinical evaluation; thus, a total of 89 patients, including the cases observed in our clinic, were carefully analyzed.

RESULTS

Among the 89 patients identified 73 were males and 16 females. Europe is the geographical area with the higher number of reports, namely 80.6% of the papers including 74.2% of the patients. The Mediterranean area accounts for 62.9% of the papers and 59.6% of the cases. The symptoms of ReA appeared after a mean number of instillations of 5.8. Polyarthritis was present in 55.1%, oligoarthritis in 37.0% and monoarthritis in 7.9%. Polyarthritis was symmetric in 51.0% and asymmetric in 49.0% of the cases; oligoarthritis was symmetric in 33.3% and asymmetric in 66.7% of the cases. Overall, an asymmetric distribution of arthritis was present in 59.6%. Knee and ankle were the joints most frequently involved. The antigen HLA B27 was positive in 42.6%. The synovial fluid analysis was defined as flogistic-aseptic in 71.9% of the patients. Arthritis was recovered within 6months in 93.2% of the cases and in 70.5% of the patients within the first two months. NSAIDs and corticosteroids, alone or in conjunction with other drugs, are used in 65.1% and in 40.4% of the cases, respectively. The clinical features of ivBCG ReA are compared with ReA from other triggering agents, from which it differs for some clinical aspects and overlaps for others.

CONCLUSIONS

Compared with a previous report, this review allows to modify some figures of this topic as a reduced prevalence of polyarthritis (from 70% to 55.1%) and of spinal and sacroiliac involvement; polyarthritis remains the more frequent clinical pattern of ivBCG ReA that, however, is characterized by rather asymmetrical distribution and involvement of the large joints of lower limbs. A definite linkage to HLA B27 is present, although without prognostic value. Moreover, arthritis is aseptic, has a latency time from antigen exposure, and is associated with extra-articular features as commonly observed in ReA from other triggering agents. Arthritis is usually benign and rarely develops into a chronic form. NSAIDs and/or corticosteroids are largely effective. Noteworthy, the overall clinical picture of arthritis triggered by ivBCG emerging from this updated review is comparable to that of ReA from other bacterial agents.