Predictors of Voriconazole Trough Concentrations in Patients with Child-Pugh Class C Cirrhosis: A Prospective Study

Abstract

This prospective observational study aimed to clinically describe voriconazole administrations and trough concentrations in patients with Child-Pugh class C and to investigate the variability of trough concentration. A total of 144 voriconazole trough concentrations from 43 Child-Pugh class C patients were analyzed. The majority of patients (62.8%) received adjustments. The repeated measured trough concentration was higher than the first and final ones generally (median, 4.33 vs. 2.99, 3.90 mg/L). Eight patients with ideal initial concentrations later got supratherapeutic with no adjusted daily dose, implying accumulation. There was a significant difference in concentrations among the six groups by daily dose (p = 0.006). The bivariate correlation analysis showed that sex, CYP2C19 genotyping, daily dose, prothrombin time activity, international normalized ratio, platelet, and Model for end-stage liver disease score were significant factors for concentration. Subsequently, the first four factors mentioned above entered into a stepwise multiple linear regression model (variance inflation factor <5), implying that CYP2C19 testing makes sense for precision medicine of Child-Pugh class C cirrhosis patients. The equation fits well and explains the 34.8% variety of concentrations (R² = 0.348). In conclusion, it needs more cautious administration clinically due to no recommendation for Child-Pugh class C patients in the medication label. The adjustment of the administration regimen should be mainly based on the results of repeated therapeutic drug monitoring.

Keywords: CYP2C19; Child–Pugh C cirrhosis; administration; trough concentrations; voriconazole.

Figure 1

The flow charts show (a) changes of voriconazole trough concentrations in 43 patients during treatment clinically; (b) statistical analysis of 144 measured voriconazole trough concentrations.

Figure 2

Distinction of voriconazole trough concentration of 43 patients (a) as the therapeutic progress by days (n = 144); (b) in different CYP2C19 phenotype groups (n = 103, p = 0.040). The Kruskal–Wallis test was used to compare C_mins of the four groups. Data are expressed as the median ± interquartile range. C_min, voriconazole trough concentration; CYP2C19, cytochrome P450 2C19; C₀, the first trough concentration after voriconazole therapy; C_ss, the final trough concentration; C₁, the repeated measured trough concentration except C_ss. *1, *2, *3, *17 reprented the single nucleotide sequence of genotype of CYP2C19.

Figure 3

For 12 patients started beyond the therapeutic target (a) Distinction of voriconazole trough concentration in different CYP2C19 phenotype groups (n = 50, p = 0.021). The Kruskal–Wallis test was used to compare C_mins. (b) Distinction of voriconazole trough concentration in different groups of C₀, C₁, C_ss (n = 50), and daily dose (n = 27). Data are expressed as the median ± interquartile range. (c) The mean concentration of such 12 patients changed by days. (d) Every dosage adjustment (n = 247) and the mode dose (n = 45) by days (It showed the highest one if there were more than one mode). *1, *2, *3, *17 reprented the single nucleotide sequence of genotype of CYP2C19.

Figure 4

The voriconazole trough concentrations by days of 8 patients who had targeted C₀ but supratherapeutic C₁ during the constant unchanged therapy dosage of voriconazole (n = 38).

Figure 5

Distribution of voriconazole trough levels over daily dosages. The numbers of measurements for each daily dose are reported.