To the Editor:
We have read with interest the article by de Lemos et al  regarding the implementation of a preventive protocol for latent tuberculosis infection (LTBI) in kidney transplant recipients residing in Brazil. In that study, all recipients and living donors were intended to be screened preoperatively or at the time of transplantation for epidemiological, clinical and radiological evidence of LTBI. The authors reported a lower incidence of posttransplant active tuberculosis in the group of patients appropriately treated for LTBI. Such an approach is consistent with the American Society of Transplantation Infectious Disease guidelines and other consensus documents [2, 3] and is broadly suggested for low- and high-incidence areas. However, we suggest that it may not be necessary to screen all pretransplant patients in a low-incidence setting and that targeted screening based on epidemiological or clinical risk factors may have a better predictive value.
For example, from 2006 to 2007, we performed a study to compare the performance of the QuantiFERON-TB assay with that of the standard tuberculin skin test (TST) for diagnosing LTBI in 153 liver transplant (LT) candidates in a single Canadian center. Approximately one-quarter of them had a positive result in either one or both tests . In the long-term follow-up, 82 patients (53.6%) had undergone transplantation as of October 2013, and 42 of them (51.2%) were considered at high-risk for posttransplant tuberculosis (on the basis of a positive screening test for LTBI, origin from a country with moderate-to-high incidence, household or occupational exposure, or abnormal chest X-ray findings). The decision on whether or not to initiate treatment was ultimately left at the discretion of the attending physician. Only two patients (2.4% of the overall cohort and 4.8% of the high-risk group) received treatment for LTBI. Both regimens were administered after transplantation and consisted of rifabutin in one patient (300 mg daily) and rifampin in one patient (600 mg daily). The latter patient developed severe hepatotoxicity after 2 weeks, which resolved with rifampicin discontinuation and replacement with levofloxacin. Overall, the therapy was maintained for 4.2 and 10.1 months, respectively. Sixteen further patients (19.5%) received at some point levofloxacin or moxifloxacin for other causes, although the cumulative exposure to the drug was lower than 3 months in all cases. Interestingly, despite the lack of prophylaxis in the majority of patients, none of them developed posttransplant active tuberculosis after a median follow-up of 5.5 years (interquartile range: 3.8–6.2), accounting for a risk-exposure period of 145 400 transplant-days.
Despite the low adherence to the current recommendations for treating LTBI in LT candidates and recipients [2, 3], no apparent consequences were observed in the present cohort. Canada has one of the lowest incidences of tuberculosis in the world, estimated for 2008 in 4.9 cases per 100 000 population . Although limited by its small sample size, our experience suggests that the actual risk of active tuberculosis in LT recipients is very low in our setting. The positive predictive value of the TST—the most widely used diagnostic approach for LTBI—largely depends on the incidence of tuberculosis in the screened population. Therefore, we suggest a targeted strategy where only patients with a higher pretest probability of LTBI are screened (i.e. those who are born or have resided for a significant time in a high-incidence country, other epidemiological exposure, and/or with abnormal chest X-ray findings).
Our study also exemplifies the practical obstacles for the application of the current recommendations for therapy of LTBI in the specific population of LT candidates and recipients due to the elevated risk of drug-induced hepatotoxicity perceived by the attending hepatologists and transplant physicians. In that sense, the choice of rifamycin-based regimens in the two treated cases was likely motivated by the higher likelihood of hepatotoxicity attributable to the standard 9-month isoniazid regimen.
While we agree with the conclusions raised by de Lemos et al  that a screening strategy will detect LTBI in transplant recipients residing in high-incidence areas, the effectiveness of such a universal screening needs to be further studied on a larger scale in various epidemiological settings, particularly in view of the cost of interferon-γ release assays. Screening high-risk patients based on clinical and radiologic factors and treating for a positive test may be more effective.
M. Fernández-Ruiz, A. Humar, D. Kumar*
Multi-Organ Transplant Program, Toronto General Hospital, University Health Network, University of Toronto, Toronto, ON, Canada
*Corresponding author: Deepali Kumar, email@example.com