A prediction model may enable better targeting of early interventions for carbapenem-resistant Enterobacteriaceae (CRE) infection after liver transplantation (LT) among carriers, according to study results published in Clinical Infectious Diseases.
In this retrospective, multinational cohort study, researchers aimed to develop a prediction model for CRE infection within 30 and 60 days after LT when the overall CRE infection threshold probability exceeded 10%. Data from 840 patients were gathered from 15 hospitals (majority having endemic CRE with Klebsiella pneumoniae carbapenemase as the main mechanism of CRE), with follow-up occurring 180 days after LT.
In the 840 patients (65.4% men; median age 55 years), the primary indication for LT was viral hepatitis (44.8%) followed by alcoholic hepatitis (24.6%). A total of 250 patients had CRE infection post-LT. There were 203 patients who were carriers of CRE before LT, and 637 patients were colonized with CRE after LT.
Although pre-LT CRE carriers had more severe liver disease and a complicated clinical course compared with post-LT CRE carriers before transplantation, the CRE infection rate was similar between the groups (33% and 28.7%, respectively), and there was no difference in all-cause mortality at 180 days (56.7% vs 58.5%, P =.46). However, the time to infection following LT was earlier for the pre-LT CRE carriers (median, 9 vs 23 days).
To assess risk factors for CRE infections, the researchers conducted a multivariable analysis that included the following 6 variables: surgical reintervention, prolonged mechanical ventilation, acute kidney injury, CRE colonization 60 days before transplant, CRE colonization within 60 days after transplant, and multisite colonization within 60 days after transplant. The predicted risk of CRE infection within 30 and 60 days after LT was 15% (interquartile range [IQR], 11%-24%) and 21% (IQR, 15%-33%), respectively.
“The model showed acceptable 60-day discrimination and prediction accuracy for CRE infection when assessed against the derivation AUC [area under the curve] 74.6 (95% CI, 70.9-78.4), Brier index 16.3 (95% CI, 6.7-25.9), and bootstrapped validation dataset AUC 73.9 (95% CI, 67.7-79.1), Brier index 16.6 (95% CI, 14.6-19.1),” the researchers noted.
Findings from this study highlight that a default strategy for treating all CRE carriers may be inferior to a targeted strategy, given a probability of CRE infection higher than 10%. “Further studies are needed to further validate the model and establish which strategy would be more effective, in terms of adverse events and collateral resistance impact and, ultimately patient survival after transplantation,” concluded the researchers.
Giannella M, Freire M, Rinaldi M, et al. Development of a risk prediction model for carbapenem-resistant Enterobacteriaceae infection after liver transplantation: a multinational cohort study. Clin Infect Dis. Published online February 10, 2021. doi:10.1093/cid/ciab109