Risk Groups of Developing Active Tuberculosis in Liver Transplant Recipients in a Tuberculosis Endemic Area: Risk Stratification by Chest Image and Interferon Gamma Release Assay

https://doi.org/10.1016/j.ijid.2021.10.043Get rights and content
Under a Creative Commons license
open access

Highlights

  • We analyzed factors that may predict active TB in liver transplantation recipients.

  • IGRA positivity itself was not associated with TB development in a TB-endemic area.

  • IGRA positive predicted active TB when chest images suggested a previous TB infection.

  • Both IGRA and chest CT should be considered to identify LTBI treatment candidates.

Abstract

Objectives

We implemented a stratified risk analysis to predict the development of active tuberculosis (TB) in liver transplantation (LT) recipients based on IGRA and chest images in a TB-endemic area.

Methods

In this retrospective cohort study, LT recipients who tested for IGRA between May 2008 and December 2017 were included. Chest images compatible with old TB lesions were considered as positive images. LT recipients were divided into six groups: LT recipients with history of treated TB, image (+)/IGRA (+), image (+)/IGRA (-), image (-)/IGRA (+), image (-)/IGRA (-) and LTBI treated. The Cox regression model was used to analyze risk groups.

Results

Among the 717 eligible LT recipients included in this study, 21 patients developed active TB. Incidence rates of TB were 2,261, 724, and 119 cases/100,000 person-years in the 1st, 2nd, and ≥ 3rd year after transplantation, respectively. History of treated TB (HR 18.92; 95% CI 4.10–87.25) and image (+)/IGRA (+) (HR 10.86; 95% CI 2.75–42.89) were independent risk factors for developing active TB. IGRA (+) with a negative image was not a risk factor.

Conclusions

Our findings suggested that both IGRA and chest images should be considered to identify risk groups for LTBI treatment.

Keywords

Tuberculosis
Latent tuberculosis infection
Liver transplantation
Interferon-gamma release assay
Chest image

Cited by (0)

#

Si-Ho Kim and Suhyun Oh contributed equally to this study.