Risk factors and clinical outcomes of pediatric liver transplant recipients with post-transplant lymphoproliferative disease in a multi-ethnic Asian cohort.

BACKGROUND: We aimed to evaluate clinical characteristics, risk factors, and disease outcomes for liver transplant recipients (LTR) with post-transplant lymphoproliferative disease (PTLD) at our center.

METHODS: Retrospective review of data of all pediatric LTR (1991-2015) was conducted.

RESULTS: The overall incidence of PTLD was 16.4% (18/110), the majority (13/18) were early lesions, while 3/18 were polymorphic/monomorphic PTLD. The risk factors significant on univariate analysis were as follows: mean age (years) at transplant (1.66 vs 4.76, P = .006); age <2 years at transplant (odds ratio [OR] 3.53 [95% confidence interval [CI]: 1.16-10.73], P = .026); cytomegalovirus (CMV) primary infection (OR 11.39 [95% CI: 3.44-37.7], P < .001); recipient CMV seronegativity (OR 7.50 [95% CI: 2.02-27.78], P = .003); presence of CMV end-organ disease (OR 4.00 [95% CI: 1.22-13.16], P = .022); Chinese ethnicity; and higher mean duration of intravenous ganciclovir prophylaxis. In multivariate analysis, CMV primary infection (OR 5.22 [95% CI: 1.25-21.87], P = .024), CMV seronegativity (OR 5.91 [95% CI: 1.13-30.90, P = .035]), and having acute cellular rejections (ACR) prior to PTLD (OR 5.53 [95% CI: 1.43-21.48, P = .013]) were significant risk factors for PTLD, with the latter two factors having a synergistic effect in increasing PTLD risk in a stratified analysis. The final multivariate model in predicting the risk of PTLD, utilizing CMV primary infection, recipient CMV seronegativity, and ACR before PTLD as predictive variables, was statistically significant (likelihood ratio chi square statistic = 25.18, P < .0001 with df = 3).

CONCLUSIONS: We report a unique clinicopathologic and risk factor profile in our cohort-early lesion PTLD accounts for the majority and the incidence of monomorphic PTLD remains low. In addition, we show a synergism between CMV naivety and ACR on PTLD risk, a higher prevalence of gastrointestinal manifestations, and a lack of significant association with Epstein-Barr virus seronegativity.