Baseline Characteristics and Patient Disposition
Of the 178 patients screened, 101 (41 children and 60 adolescents) were eligible for inclusion in the study and received etravirine treatment (Fig. 1). Baseline characteristics are shown in Table 1. More than half of the enrolled patients were female (63%). Approximately half (49%) were White, 30% Black/African-American and 20% Asian. The median age was 12.0 years, ranging from 6 to 17 years. Patients most frequently originated from Thailand (20 patients; 19.8%), Argentina (15 patients; 14.9%), the USA (15 patients; 14.9%) and South Africa (10 patients; 9.9%). The median baseline log10 VL was 3.6 log10 copies/mL in children and 4.0 log10 copies/mL in adolescents. At baseline, 62.4% of patients had a VL < 20 000 copies/mL, while 6.8% had a VL > 50 000 copies/mL. The predominant mode of HIV transmission was mother-to-child transmission (99%), with one adolescent contracting HIV-1 infection by heterosexual contact. Efavirenz or nevirapine was previously used in 67% of patients, and 56.2% of patients enrolled with no phenotypic activity to these first-generation NNRTIs.
In total, 76 patients (75.2%) completed 48 weeks of treatment, with 25 (24.8%) prematurely discontinuing (Fig. 1). Mean treatment duration for the week 48 analysis was 49.9 weeks. AEs (mainly rash) and protocol noncompliance were the most frequent reasons for premature treatment discontinuation [n = 2 (4.9%) and n = 3 (7.3%) for children; n = 6 (10%) and n = 5 (8.3%) for adolescents, respectively] (Fig. 1).
Darunavir/ritonavir (51.5%) and lopinavir/ritonavir (38.6%) were the most frequently used boosted PIs in the OBR. The majority of patients used two (59.4%) or one (27.7%) N[t]RTI, most frequently lamivudine (51.2%) and zidovudine (39%) in children, and TDF (51.7%) and lamivudine (35%) in adolescents. Enfuvirtide and raltegravir were used by two (2%) and 11 (10.9%) patients, respectively.
Safety and Tolerability
Overall, 82.9% of children and 91.7% of adolescents experienced at least one AE (Table 2), and 22% and 40%, respectively, reported at least one AE possibly related to treatment. AEs considered very likely to be related to etravirine treatment occurred in 2.4% of children and 3.3% of adolescents. Mostly, AEs were of grade ≤ 2 severity (Table 2); 12% of patients experienced grade 3 AEs. Only two grade 4 AEs occurred (both were thrombocytopaenia and not related to etravirine).
Five patients (all adolescents) reported at least one serious AE. Of these, a serious AE in only one patient was related to etravirine, which was an inadvertent etravirine overdose of 250 mg bid instead of 200 mg bid for 50 days, with no safety consequences. One patient permanently discontinued the study because of drug resistance as a result of poor adherence which, along with decreased weight, was reported as serious AEs considered not related to etravirine. The remaining three patients experienced serious AEs considered not related to etravirine that resulted in periods of hospitalization, but did not necessitate permanent treatment discontinuation. No patients died during the study.
The most commonly reported AEs (in ≥ 10% of all patients) were upper respiratory tract infection, any rash, diarrhoea, cough and vomiting (Table 2). There was a higher incidence of any rash in female patients (17 of 64; 26.6%) than in male patients (six of 37; 16.2%). The median time to onset of first rash was 10 days, and the median duration of rash was 7 days. The most frequent grade ≥ 2 AEs possibly related to etravirine included any rash (13%) and diarrhoea (3%) (Table 2).
Permanent discontinuation of etravirine because of AEs occurred in eight patients, most commonly because of rash. The four patients who discontinued because of rash were all female. No Stevens–Johnson syndrome or other grade 4 rashes were reported. In a post hoc, multivariate logistic regression model, darunavir/ritonavir use in the OBR (vs. lopinavir/ritonavir or another boosted HIV PI), White race, being a North American patient and higher exposure (AUC0–12h) to etravirine were all found to be associated with increased incidence of rash. Another post hoc, multivariate analysis on the pooled PIANO and DUET data showed that female sex, but not age or participation in the PIANO study (a partial surrogate for age), was the only risk factor for developing rash.
Virological response over time to week 48 (< 50 copies/mL; ITT-TLOVR) for children, adolescents and the overall population is presented in Figure 2a. Based on NC=F imputed data, 52.5% of all patients had a plasma VL < 50 copies/mL at week 24 (58.5% in children; 48.3% in adolescents). At week 48, the overall response slightly increased from week 24 to 56.4% of all patients (68.3% in children; 48.3% in adolescents) (Table 3).
Efficacy over time to week 48. (a) Virological response over time to week 48 [< 50 copies/mL; intent-to-treat, time to loss of virological response (ITT-TLOVR)] and (b) change from baseline in CD4 cell count [noncompleter equals failure (NC = F)] over time for the overall population and children and adolescent subgroups. CI, confidence interval.
Factors found to be prognostic of virological response in the final multivariate model were level of adherence by pill count > 95% (P = 0.0452), male sex (P = 0.0156), lower etravirine weighted genotypic score at baseline (P = 0.0007) and high etravirine C 0h (P = 0.0005).
Mean (95% CI) changes from baseline in CD4 cell count to week 48 in children, adolescents and all patients are shown in Figure 2b. An increase in absolute CD4 cell count was observed at all time-points. The mean change in CD4 cell count from baseline at week 48 was +156 cells/μL overall, +178 cells/μL in children and +141 cells/μL in adolescents (P < 0.0001 for intragroup comparisons vs. baseline).
Overall, 40.6% (41 of 101) of patients (26.8% of children and 50% of adolescents) experienced VF, the majority of whom were nonresponders (Table 3). Twenty-three of 41 (56.1%) patients with VF had previously used efavirenz, nevirapine or both, and the median number of NNRTI resistance-associated mutations (RAMs) in patients with VF was somewhat higher than in patients without VF [2.0 (range 0–5.0) vs. 1.0 (range 0–4.0), respectively). There were 30 patients with VF who had available genotypic data at baseline and endpoint. Emerging NNRTI RAMs occurred in 18 (three children and 15 adolescents) of the 30 patients with genotypic data. The most commonly observed emerging NNRTI RAMs (in at least three cases) were Y181C (n = 8; one child and seven adolescents), L100I (n = 3; all adolescents), E138A (n = 3; two children and one adolescent) and V90I (n = 3; one child and two adolescents). These four NNRTI RAMs have been previously defined as etravirine RAMs.[25,26,31] Both genotypic and phenotypic development of resistance to etravirine was observed in nine of the 23 (39.1%) patients with VF for whom genotypic and phenotypic data were available. In three of 23 patients with VF, etravirine resistance was demonstrated by phenotype only. No development of phenotypic resistance to etravirine was observed in 11 of 23 (47.8%) patients with VF.
According to the PENTA adherence questionnaire, 69.2% of children and 61.4% of adolescents were adherent to etravirine up to week 48. Self-reported adherence was markedly higher compared with measurements of adherence by pill count (> 95% adherence: 45.9% in children and 34.6% in adolescents), consistent with reports in other studies.[32–34]
Mean etravirine AUC0–12h and C 0h were 5684 ng h/mL and 377 ng/mL in children and 4895 ng h/mL and 325 ng/mL in adolescents, respectively. Sex, age, adherence to etravirine and TDF use did not significantly influence the pharmacokinetics of etravirine (AUC0–12h or C 0h). Statistically significant effects were observed for race (only for AUC0–12h), weight and the use of a boosted PI in the OBR. Lower exposures to etravirine were observed in Asians (mean AUC0–12h 4316 ng h/mL) relative to both Black patients (mean 4548 ng h/mL) and White patients (6012 ng h/mL). Higher exposures to etravirine were observed in patients with lower body weights and patients using a background PI other than darunavir/ritonavir or lopinavir/ritonavir (i.e. atazanavir/ritonavir or saquinavir/ritonavir).
HIV Medicine. 2014;15(9):513-524. © 2014 Blackwell Publishing