Letter to the Editor
Screening for Latent Tuberculosis Infection in Low-Incidence Areas
Version of Record online: 29 MAY 2014
© Copyright 2014 The American Society of Transplantation and the American Society of Transplant Surgeons
American Journal of Transplantation
Volume 14, Issue 7, page 1709, July 2014
How to Cite
Santoro-Lopes, G. (2014), Screening for Latent Tuberculosis Infection in Low-Incidence Areas. American Journal of Transplantation, 14: 1709. doi: 10.1111/ajt.12792
- Issue online: 20 JUN 2014
- Version of Record online: 29 MAY 2014
To the Editor:
We thank Fernández-Ruiz et al  for their interest in our article . We agree with the authors that the effectiveness and safety of the current recommendations to prevent active tuberculosis (TB) in solid organ transplant recipients should be further studied in different settings. The assumption that in low-incidence areas a more accurate risk stratification would be obtained if the screening for latent tuberculosis infection (LTBI) is targeted to patients with other risk factors is very plausible. In fact, this approach has been recommended in the TBNET consensus statement , but its performance remains to be tested in a large-scale study. In the illustrative example presented in their letter to the editor, it is likely that the risk for posttransplant TB was strongly overestimated. It is surprising that 51% of liver transplant recipients in a low-incidence area were classified as at high-risk for TB while in the study of de Lemos et al , carried out in an endemic area, only 26% of the patients who were appropriately screened were defined as such. It is not clear whether universal screening for LTBI was the main source of risk overestimation in that setting. Although the distribution of risk factors among the 82 liver transplant recipients were not presented, the analysis of the published data from the original cohort of 153 transplant candidates  shows that only 10 (6.5%) patients would be defined as at high-risk for TB-based solely on a positive result of an interferon-gamma release assay (IGRA). Apparently, a more important source of risk overestimation might have been the fact that 33% of the transplant candidates were considered to be at high-risk, because they came from a country with TB prevalence >10 cases/100 000 inhabitants , regardless of the results of LTBI screening tests. This criterion is probably associated with a low positive predictive value. According to it, for example, LTBI treatment would be indicated for all transplant recipients in Brazil. In the study of de Lemos et al , the use of a simple set of predictive rules (which included the result of a tuberculin skin test), in a group of transplant recipients that lived in a Brazilian city with high incidence of TB, allowed targeting LTBI treatment to a subset of subjects who were at a significantly higher risk of developing posttransplant TB. Therefore, we think that this might be a better preventive approach for transplant patients coming from endemic areas. Finally, because of the study protocol , another potential source of risk overestimation to be considered is that transplant candidates were screened with both tuberculin skin test and IGRA and considered to have LTBI if one of these tests were positive. This double test approach is not routinely recommended , especially in low-incidence areas where there is a concern that discordant results in the screening tests may further reduce the positive predictive value of the diagnosis of LTBI.
Infectious Diseases Clinic, Hospital Universitário Clementino Fraga Filho, Rio de Janeiro, Brazil
Department of Preventive Medicine, School of Medicine, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil
*Corresponding author: Guilherme Santoro-Lopes, email@example.com
The author of this manuscript has no conflicts of interest to disclose as described by the American Journal of Transplantation.