Mycobacterium tuberculosis and prosthetic joint infection

We read with great interest the recent Clinical Picture by Arvind von Keudell and colleagues

about a case of tuberculosis infection in a prosthetic joint, which had probably resulted from the reactivation of latent tuberculosis. We agree that tuberculosis infection in prosthetic joints is rare but should be considered in differential diagnoses for culture-negative prosthetic joint infections in patients with previous episodes of tuberculosis or epidemiological risk factors. Moreover, we believe that tuberculosis prosthetic joint infections appear to be under-reported.

We treated five cases at our centre between 1993 and 2013, four of which were associated with a hip prosthesis and one with a knee prosthesis (appendix). The medical history of tuberculosis was identified in only three of these cases, including one case of urinary tract tuberculosis, one case of pulmonary tuberculosis, and one case of tuberculosis in childhood. Diagnosis of tuberculosis prosthetic joint infection was reached through positive cultures of deep samples taken during surgical biopsy in three cases and during percutaneous biopsy in two cases. The mean time from symptom onset to diagnosis was 34 months. All Mycobacterium tuberculosis isolates were susceptible to rifampicin, isoniazid, pyrazinamide, and ethambutol. Three of our cases were associated with one or more other microorganisms (Staphylococcus aureus, Enterococcus faecalis, Staphylococcus hominis, Corynebacterium pseudodiphtericum, and Aeromonas spp). We treated one of our cases with only antituberculosis therapy without surgery, and treated four with surgery followed by antituberculosis therapy, including debridement, antibiotics, and implant retention in one case, and two-stage prosthesis exchange in the three other cases. The mean length of antituberculosis therapy was 11 months. Remission was observed in four cases with an average follow-up of 24 months. One patient died of systemic tuberculosis after 2 months of treatment.

Despite the high volume of arthroplasty, tuberculosis prosthetic joint infection is rare. Early prosthetic joint infection might be explained by pre-existing tuberculous coxitis. In other cases, tuberculosis prosthetic joint infection is likely to be a haematogenous infection following systemic tuberculosis. Diagnosis of tuberculosis prosthetic joint infection is mainly reached through positive cultures of deep samples. Nevertheless, percutaneous biopsy has been shown to be an alternative diagnostic method, as used in two of our cases. On the basis of our data, we believe that tuberculosis prosthetic joint infection is certainly under-reported and should be considered in culture-negative prosthetic joint infections, and not only in patients with a past medical history of osteoarticular tuberculosis. Previous or concomitant infection with another pathogen should not exclude diagnosis. Careful histological analysis, systematic mycobacterial cultures, and the contribution of specific PCR detection could help physicians to diagnose tuberculosis prosthetic joint infection.

We declare no competing interests.