World Health Organization (WHO) 2013 guidelines recommend a ritonavir-boosted protease inhibitor (bPI) plus 2 nucleoside reverse transcriptase inhibitors (NRTIs) for first-line antiretroviral therapy (ART) in HIV-infected children aged <3 years, particularly if perinatally exposed to nonnucleoside reverse transcriptase inhibitor (NNRTI), and for second-line ART in HIV-infected children receiving 2 NRTIs plus a NNRTI first-line.1 In 2011, only 28% of children urgently needing ART were receiving treatment, compared with 58% of adults,2 and only 3% of children on ART were on bPI second-line therapy. As children require ART lifelong, the numbers needing second-line treatment will rise substantially in the future.
Simplifying 2 NRTI + NNRTI pediatric first-line ART with fixed dose combination solid formulations, dosed following simplified WHO weight-band tables, has significantly aided the programmatic scale-up of first-line pediatric treatment. An important factor constraining wider use of second-line ART and bPI-based first-line ART in children aged <3 years is the lack of affordable and appropriate pediatric bPI-based formulations for resource-limited settings. The only currently available combination bPI is ritonavir-boosted lopinavir (LPV/r) in liquid formulation for young children and as pediatric tablets for older children. The syrup is unpleasant tasting, has a high alcohol concentration (42%), and contains propylene glycol, which is especially undesirable in children aged <2 years.3 Moreover, LPV/r syrup requires refrigeration, making it challenging for storage, transportation, and use in resource-limited settings. The pediatric tablet is still relatively large and should not be crushed because this decreases bioavailability by 40% versus whole tablets.4 LPV/r is the only bPI licensed for young children. Pediatric ritonavir as a booster for other protease inhibitors is only available as a very unpleasant tasting liquid or large tablet (100 mg). Therefore, alternative better tasting LPV/r formulations not requiring cold-chain transportation and storage are urgently needed.
LPV/r has recently been developed by a generic pharmaceutical company (Cipla Pharmaceuticals, Mumbai, India) as a heat-stable tablet (produced by a melt extrusion process similar to the innovator product) containing the same dose (100/25 mg LPV/r; Lopimune) as the FDA approved innovator pediatric tablets, and also a novel minitab sprinkle formulation stored in capsules (40/10 mg LPV/r; Lopimune). The capsules can be opened, and the sprinkles given even to the smallest children, mixed with food or milk. In the Children with HIV in Africa—Pharmacokinetics and Adherence of Simple Antiretroviral Regimens (CHAPAS-2) study, we evaluated the pharmacokinetics and acceptability of the minitab formulation versus the innovator syrup (80/20 mg LPV/r per mL; Kaletra, Abbott, North Chicago, IL) and versus the generic pediatric tablets in HIV-infected Ugandan infants and children.
CHAPAS-2 was an open, randomized, phase I, 2-period crossover comparative bioavailability trial in HIV-infected infants/children aged 3 months to <13 years taking or about to start first-line (infants exposed to perinatal nevirapine for the prevention of mother-to-child transmission) or second-line ART with 2 NRTIs + LPV/r (older children) from 2 pediatric clinics in Kampala, Uganda (Joint Clinical Research Centre, Baylor-Uganda; ISRCTN01946535). At enrollment, the majority of children were already on LPV/r for >1 month (Table 1). Infants/children were ineligible if they were expected to increase dose before the second pharmacokinetic sampling day at week 8, were on concomitant medication known to interact with ART (eg, TB treatment), or had liver enzymes grade 2 or higher, anemia (hemoglobin <8.5 g/dL) or other illnesses (eg, severe diarrhea, vomiting, renal or liver disease) that could influence LPV/r pharmacokinetics. Enrolled children missing any antiretroviral drug dose in the 3 days before each pharmacokinetic sampling (assessed by adherence questionnaire) were excluded from pharmacokinetic analyses, as were children with pharmacokinetic evidence of poor compliance (both C0h below the lower limit of quantification and C12h >3.0 mg/L). All caregivers, and children where appropriate, gave written informed consent and assent, respectively. The study was approved by the Ethics Committee from each participating site, the Uganda National Council of Science and Technology and by University College London, United Kingdom.
At enrollment, all infants aged 3 to <12 months were included in a nonrandomized 2-period crossover design (cohort A), and all children aged 1–4 years (cohort B) or 4 to <13 years weighing <25 kg (cohort C) in a 1:1 randomized 2-arm, 2-period crossover design. The 2 first designs compared the novel LPV/r minitabs with innovator syrup, and the third with the new generic pediatric tablets (see Figure S1, Supplemental Digital Content, http://links.lww.com/QAI/A508). Children were only included in cohort C if they were able to swallow pediatric tablets. Children in cohort B and cohort C were randomized using computer-generated randomization lists produced by the trial statistician at the Medical Research Council's Clinical Trials Unit. Randomization was performed by phoning the Clinical Trials Unit. LPV/r was dosed twice daily according to WHO 2010 pediatric weight bands (see Table S1, Supplemental Digital Content, http://links.lww.com/QAI/A508) as syrup, minitab-contained capsules (which have to be opened), and whole pediatric tablets.
Four weeks after enrollment, when steady state was achieved on allocated treatment (syrup cohort A; minitabs/syrup cohort B and minitabs/tablets cohort C), an intensive 12-hour pharmacokinetic session was performed. Samples were taken immediately before directly observed intake of the morning LPV/r dose (t = 0 hours) and 1, 2, 4, 6, 8, and 12 hours later. Children, if not breastfed, fasted >3 hours before the pharmacokinetic session, and breakfast (mainly porridge) was given with the morning dose. After the week 4 sampling session, all infants and children switched LPV/r formulation. Cohort A switched from syrup to minitabs, cohort B switched from minitabs to syrup or vice versa, and cohort C switched from tablets to minitabs or vice versa (see Figure S1, Supplemental Digital Content, http://links.lww.com/QAI/A508). At week 8, a second intensive pharmacokinetic sampling session was performed. Data on acceptability of each LPV/r formulation were collected from standardized questionnaires at baseline (if already on LPV/r at enrollment) and weeks 4, 8, and 12 after enrollment. At week 8, children and caregivers chose which formulation they wished to continue.
Lopinavir (LPV) and ritonavir concentrations in plasma samples were determined using ultra performance liquid chromatography with ultraviolet detection5 at the Department of Pharmacy, Radboud University Medical Centre, Nijmegen, the Netherlands. The analytical assay LPV and ritonavir ranges were 0.109–31.2 mg/L and 0.044–29.4 mg/L, respectively. Intraday and interday precision ranged from 0.6% to 4.2% [coefficient of variation (CV)], and 0.3%–1.8%, respectively. The assay accuracy range was 98.2%–105.6%.
For cohort B and cohort C, the sample size of 24 children provided >80% power for the width of the 90% confidence interval (CI) for the geometric mean ratio (GMR) between formulations to lie within 0.80–1.25 (bioequivalence6) if no difference was observed (GMR = 1), based on an estimated SD of change in log10 area under the concentration–time curve 0–12 hours after dose (AUC0–12h) between minitab sprinkle and tablet/syrup of 0.26. For primary analysis, steady-state LPV pharmacokinetic parameters [AUC0–12h, maximum concentration (Cmax) and concentration at 12 hours after dose (C12h)] were determined using Phoenix version 6.3 (Pharsight Corporation, Sunnyvale, CA) and compared between formulations within child using GMRs and 90% CIs. The proportions with subtherapeutic LPV concentrations (defined as <1.0 mg/L7) were compared between formulations using Fisher exact test. AUC0–12h were compared across weight bands using analysis of variance on log-transformed data [equivalent to the geometric mean (GM)].
In total, 79 children were recruited from August 2011 until September 2012. One was ineligible (not receiving LPV/r, on efavirenz-based first-line) and 1 withdrew consent shortly after enrollment (because the caregiver was unable to attend the sample session days), leaving 77 children for analyses (19 infants in cohort A, 26 children in cohort B, and 32 children in cohort C). Median (interquartile range) age was 0.5 (0.4–0.7), 2.0 (1.8–2.8), and 6.2 (5.8–7.8) years in cohort A, B, and C, respectively (Table 1). Infants/children were moderately wasted and stunted.
Primary pharmacokinetic analyses were based on 13 infants in cohort A (6 excluded), 21 children in cohort B (5 excluded), and 25 children in cohort C (7 excluded), each with 2 pharmacokinetic profiles. Children excluded from primary pharmacokinetic analyses had been wrongly dosed (3 cohort C), unable to swallow pediatric tablets (1 cohort C), noncompliant at one of the 2 sampling sessions (both C0h <lower limit of quantification and C12h >3.0 mg/L; 4 cohort A, 4 cohort B, and 1 cohort C), changed weight bands between weeks 4 and 8 (1 cohort B, 2 cohort C), developed tuberculosis (1 cohort A), or had transferred to another clinic so only 1 PK curve was undertaken (1 cohort A).
In cohort A (infants aged 3 to <12 months), LPV concentrations and pharmacokinetic parameters were slightly higher with minitabs compared with innovator syrup (Fig. 1A; Tables 2 and 3). The GMR (minitabs:syrup) for AUC0–12h, Cmax, and C12h all lay within the 0.80–1.25 range, but the lower and upper 90% CI limits fell outside it (Tables 2 and 3). This was a consequence of high but similar interindividual variability with both minitabs (CV%: 46%, 38%, and 68%, AUC0–12h, Cmax, and C12h, respectively) and innovator syrup (51%, 40%, and 73%, respectively). Two (15%) and zero (0%) of 13 children had a subtherapeutic concentration on syrup and minitabs, respectively (Fig. 2; Fisher exact P = 0.48).
Similarly to infants (cohort A), in cohort B (children aged 1 to <4 years), LPV concentrations and pharmacokinetic parameters were slightly higher with minitabs compared with innovator syrup (Fig. 1B; Tables 2 and 3). The GMR (minitab:syrup) for AUC0–12h, Cmax lay within the 0.80–1.25 range, with C12h GMR just above this, with the upper 90% CI limits all above 1.25 (Tables 2 and 3). There was no impact of randomization order on any estimated GMR (P > 0.4). The AUC0–12h, Cmax, and C12h CV% were 33%, 27%, and 50%, respectively, in minitabs, compared with 33%, 27%, and 55% with syrup. No (0%) child had subtherapeutic concentrations 12 hours after the intake of either formulation (Fig. 2).
In cohort C (children aged 4 to <13 years), LPV concentrations of the new generic pediatric tablets were higher compared with the novel minitabs (Fig. 1C). The GMR (minitabs:tablets) (90% CI) for AUC0–12h, Cmax, and C12h were 0.72 (0.60–0.86), 0.74 (0.64–0.85), and 0.59 (0.43–0.81), respectively, all lying outside the bioequivalence range of 0.80–1.25. There was no impact of randomization order on any estimated GMR (P > 0.15). The interindividual variability (CV%) of the pharmacokinetic parameters was moderately high for minitabs (49%, 42%, and 76%, for AUC0–12h, Cmax, and C12h, respectively), compared with tablets (28%, 19%, and 66%, respectively). One (4%) and 4 (16%) of the 25 included children had subtherapeutic LPV trough concentrations after receiving tablets and minitabs, respectively (Fig. 2; Fisher exact P = 0.35). Compared with historical data, LPV pharmacokinetic parameters were higher with tablets, but similar with minitabs (Tables 2 and 3).
Ritonavir pharmacokinetic data and comparisons were consistent with LPV in all 3 cohorts. There were no differences across weight bands in LPV or RTV pharmacokinetic parameters in the different cohorts, or in cohorts A and B combined (P > 0.23; analysis of variance).
Minitabs had to be administered with food. In cohort A, 83% of infants were breastfed (minitabs were dissolved in a small volume of expressed breast milk and given to the infant in a spoon or put directly on the infant tongue before breastfeeding); in older children the most common food caregivers gave minitabs with was porridge (62% cohort B, 34% cohort C).
Among younger children in cohorts A and B, more problems were reported with taking syrups than minitabs (Fig. 3). In contrast, the older children, all but one of whom were already established on tablets for >6 months, reported more problems with taking minitabs. The problem reported most often with all formulations was taste; taste was worse with minitabs than tablets (50% versus 0%), but similar with minitabs and syrup (53% minitabs versus 67% syrups, cohort A, and 38% minitabs versus 38% syrup, cohort B) (Fig. 3). Difficulty swallowing was reported as a problem in 20% minitabs versus 60% syrups (cohort A), 27% minitabs versus 19% syrup (cohort B), and 13% minitabs versus 0% tablets (cohort C). Storage, transportation, and conspicuousness were less problematic for minitabs compared with syrups. However, several caregivers were concerned about the number of capsules needing to be opened to give minitabs to older children.
About 94%, 79%, and 76% of the caregivers in cohort A, cohort B, and cohort C, respectively, reported that it was easy or very easy to switch from the original formulation to minitabs.
At enrollment, only 37%, 12%, and 41% of caregivers in cohort A, cohort B, and cohort C thought they would prefer minitab sprinkles. At week 12, 72%, 64%, and 19%, respectively, reported a preference for minitabs. At week 8, 14/18 (78%) caregivers in cohort A and 19/26 (73%) in cohort B chose to continue minitabs rather than syrups; however, in cohort C, only 7/32 (22%) caregivers/children chose minitabs, 5 of the 7 children being aged <6 years.
Caregivers provided multiple comments about the formulations at each visit. It was clear that minitabs were preferred to syrup in cohorts A and B largely because they were considered to be easier to administer, store (including not requiring refrigeration), and transport. For those preferring syrup, key issues were first, that minitabs were also bitter; second, that caregivers felt more confident that they could administer the whole dose with syrup than with minitabs even if the child struggled; and third, the requirement to give minitab sprinkles with food was of concern to some caregivers because some felt they needed to sweeten food with sugar or give with honey (which is expensive) to mask the taste. One caregiver in cohort A reported that the child subsequently refused the food, which had been previously administered with minitabs.
Only 1 grade 4 adverse event (AE) was reported (malaria with severe anemia, Hb <6.5 g/dL) at week 9 in a child on innovator syrup (second pharmacokinetic sampling deferred to week 12). This AE was also reported as a serious AE (hospitalization) and was considered unrelated/unlikely related to the study medication. No other grade 3 or 4 AEs were reported among children taking any of the formulations.
This comparative bioavailability study found slightly higher, but still broadly equivalent, LPV/r exposure from the novel minitab sprinkle formulation taken with food in infants/children aged 3 months to 4 years compared with the currently used innovator syrup and historical data.3 Exposure from minitab sprinkles in older children (aged 4 to <13 years) was also generally consistent with our younger cohort. However, LPV/r exposure from the generic tablets was significantly higher than from minitabs. Variability in LPV/r pharmacokinetic parameters was moderate to high with all formulations, but no LPV/r exposure differences were found between weight bands for twice-daily dosing recommendations across the 3 formulations. Caregivers found minitabs more acceptable than syrups for their infants/children, particularly for transportation and storage reasons. However, for older children already able to swallow tablets, these were more acceptable than minitabs, where taste remained a concern.
LPV/r exposure in children in this study was slightly higher than adults and historical data from children on the same formulation,3,8 probably because our study doses were higher (553–723 mg·m−2·d−1) following WHO 2010 weight-band dosing recommendations rather than 460/115 mg·m−2·d−1 from the manufacturers leaflet.3 Exposure with minitabs was higher than from syrup, despite equivalent mg dosing, thus, bioequivalence could not be confirmed. However, LPV/r was well tolerated, although it should be noted that at enrollment, most children had already received LPV/r for >1 month. Children both naive to and previously receiving LPV/r were eligible to allow for potentially stronger preferences in those already used to a previous formulation. Our findings are compatible with interim results from IMPAACT 1083,9 which estimated AUC0–12h, Cmax, and C12h of 119.8 h·mg/L, 13.4 mg/L, and 5.5 mg/L, respectively, in children aged 0–13 years on innovator pediatric LPV/r tablets or liquid formulations dosed following WHO weight bands with no severe adverse events definitely related to LPV/r.
Figure 1 suggests that there could be relatively large circadian variations in pharmacokinetics12 in the 1 to <4 years cohort in particular, and also with tablets in older children. However, there was considerable variability in LPV C0h and C12h values, so this may also be due to chance. An alternative explanation would be late intake of the LPV evening dose before the PK day; although this dose was observed when the child was admitted the night before, the precise timing was unfortunately not recorded so we cannot assess this further.
Apart from relatively small numbers (59 children with 2 pharmacokinetic profiles), 1 limitation is that our study included only Ugandan children. Genetic polymorphisms have minimal clinically relevant influence on LPV exposure in children,10 so this is unlikely to have materially affected results. Another limitation is the lack of concurrent virological data, although changes would not be expected since exposure from the new formulations was higher than previously reported. Variability in pharmacokinetic parameters between groups was high, but subtherapeutic trough concentrations (<1.0 mg/L) were mostly rare or absent. In only 2 groups (children aged 4 to <13 years on minitabs; infants aged 3 to <12 months on syrups) were trough concentrations subtherapeutic in 15%–16%, similar to the 13% in IMPAACT 1083, which also reported favorable 24-week HIV-1 RNA responses.9 Finally, because of challenges in sampling relatively young children over 24 hours, we were unable to directly estimate the impact of unequal AM/PM dosing. As above, unfortunately, time of intake of the preceding evening dose was also not recorded.
Acceptability data also suggest that caregivers found the minitab sprinkles to have important advantages over syrups. Most young children (aged 1 to <4 years) in cohort B and around half the infants (aged 3 to <12 months) in cohort A had already been taking LPV/r syrup for >1 month (median: 1.1 years). Despite being stable on syrup for some time, most chose to continue minitabs at study exit, mainly because of improved ease of transportation and/or storage than syrup, although taste was still a concern. However, for older children (all able to swallow tablets) minitabs were less acceptable, particularly regarding taste. A few caregivers in the younger cohorts also preferred syrup to minitabs, citing poor taste, requirement to take with food (particularly if food refusal occurred by association or if additional expensive food items such as honey were deemed necessary). Finally, because of mixing the minitab sprinkle with a small amount of food, some caregivers were unsure that the whole dose was taken when administering minitabs during the study.
The unmet need for child-friendly formulations to facilitate scale-up of bPI-based first- and second-line ART for younger and older children, respectively, in resource-limited settings is clear. Overall, the novel, heat-stable, minitab sprinkle and generic pediatric tablet formulation of LPV/r dosed following WHO weight bands had an acceptable pharmacokinetic profile and was generally more acceptable and preferred to syrup in the short term. For countries choosing to move to LPV/r first-line for children <3 years following WHO 2013 recommendations, the minitab sprinkle formulation (submitted for regulatory approval) offers an alternative to syrup, which has well-known limitations of acceptability, administration, and transportation and storage logistics. However, follow-up in this study is very short, and further follow-up is planned to describe longer-term acceptability and efficacy through to 1 year. In addition, further pharmacokinetic and acceptability studies are planned for an improved finer “granule” sprinkle formulation with better taste masking, which is under development.
The authors thank the families and children participating in the CHAPAS-2 trial. They also thank Dr. Mohammed Lamorde who assisted with pharmacokinetic training and the study teams of Joint Clinical Research Centre, Kampala, Uganda; Baylor-Uganda Paediatric Infectious Disease Clinic, Mulago Hospital, Kampala, Uganda; Medical Research Council Clinical Trials Unit, London, United Kingdom; and Radboud University Nijmegen Medical Centre, Nijmegen, the Netherlands.