RESUMO
Background: The Roussel Uclaf Causality Assessment Method (RUCAM) is a validated tool for assessing causality in cases of suspected drug-induced liver injury (DILI). However, RUCAM cannot discriminate between concomitant hepatotoxic drugs with the same temporal sequence. Objective: To analyse the utility of the lymphocyte transformation test (LTT) for assisting updated RUCAM in 45 patients and 40 controls with a clinical diagnosis of DILI. Methods: Suspected DILI cases were detected through the Prospective Pharmacovigilance Program from Laboratory Signals in Hospital (PPLSH) or by consultations. The controls completed the drug therapy with no adverse reactions during the study period. A receiver operating characteristics (ROC) curve analysis was performed to calculate the optimal cut-off value for the stimulation index (SI), corresponding to the largest sum for the specificity and sensitivity values of LTT for true DILI cases. Results: Out of 45 patients diagnosed with DILI, 42 cases were detected by the PPLSH, two cases by consultation and one case by both methods. Most DILI cases (64.4%) arose during hospitalization. According to the biochemical parameters, 24 cases (53.3%) had the hepatocellular phenotype, 14 (31.1%) had the cholestatic phenotype, and 7 cases (15.6%) had the mixed phenotype. Considering the severity criteria, 7 (15.5%) cases were classified as moderate DILI, and 4 (8.9%) were severe DILI; there were no fatal cases. A total of 149 drugs (median/case, 3; IQR, 2-5) were suspected to be involved in the DILI cases (RUCAM score ≥3). In 8 cases, only one drug was suspected, and polypharmacy (≥5 drugs) was identified in 29% of the cases. Of all DILI cases, 46 (30.9%) of the 149 suspected drugs produced positive LTT results, and the LTT was positive in 34 (75.5%) of the 45 patients. No exposed controls produced positive LTT results. The optimal cut-off of 1.95 for the SI was obtained with a sensitivity of 77% and specificity of 100% (area under the curve, 0.91; 95% asymptotic confidence interval 0.84-0.97; p < 0.001). The sensitivity of the hepatocellular phenotype was 92%. Conclusion: Our results demonstrate that LTT is an add on strengthening causality in cases of suspected idiosyncratic DILI, especially for patients with several suspected drugs and a hepatocellular phenotype.
RESUMO
Many factors have been described to contribute to voriconazole (VCZ) interpatient variability in plasma concentrations, especially CYP2C19 genetic variability. In 2014, Hicks et al. presented data describing the correlation between VCZ plasma concentrations and CYP2C19 diplotypes in immunocompromised pediatric patients and utilized pharmacokinetic modeling to extrapolate a more suitable VCZ dose for each CYP2C19 diplotype. In 2017, in our hospital, a clinical protocol was developed for individualization of VCZ in immunocompromised patients based on preemptive genotyping of CYP2C19 and dosing proposed by Hicks et al., Clinical Pharmacogenetics Implementation Consortium (CPIC) clinical guidelines, and routine therapeutic drug monitoring (TDM). We made a retrospective review of a cohort of 28 immunocompromised pediatric patients receiving VCZ according to our protocol. CYP2C19 gene molecular analysis was preemptively performed using PharmArray®. Plasma trough concentrations were measured by immunoassay analysis until target concentrations (1-5.5 µg/ml) were reached. Sixteen patients (57.14%) achieved VCZ trough target concentrations in the first measure after the initial dose based on PGx. This figure is similar to estimations made by Hicks et al. in their simulation (60%). Subdividing by phenotype, our genotyping and TDM-combined strategy allow us to achieve target concentrations during treatment/prophylaxis in 90% of the CYP2C19 Normal Metabolizers (NM)/Intermediate Metabolizers (IM) and 100% of the Rapid Metabolizers (RM) and Ultrarapid Metabolizers (UM) of our cohort. We recommended modifications of the initial dose in 29% (n = 8) of the patients. In RM ≥12 years old, an increase of the initial dose resulted in 50% of these patients achieving target concentrations in the first measure after initial dose adjustment based only on PGx information. Our experience highlights the need to improve VCZ dose predictions in children and the potential of preemptive genotyping and TDM to this aim. We are conducting a multicenter, randomized clinical trial in patients with risk of aspergillosis in order to evaluate the effectiveness and efficiency of VCZ individualization: VORIGENIPHARM (EudraCT: 2019-000376-41).