Time to treatment disruption in children with HIV-1 randomized to initial antiretroviral therapy with protease inhibitors versus non-nucleoside reverse transcriptase inhibitors.

BACKGROUND: Choice of initial antiretroviral therapy regimen may help children with HIV maintain optimal, continuous therapy. We assessed treatment-naïve children for differences in time to treatment disruption across randomly-assigned protease inhibitor versus non-nucleoside reverse transcriptase inhibitor-based initial antiretroviral therapy.
METHODS: We performed a secondary analysis of a multicenter phase 2/3, randomized, open-label trial in Europe, North and South America from 2002 to 2009. Children aged 31 days to <18 years, who were living with HIV-1 and treatment-naive, were randomized to antiretroviral therapy with two nucleoside reverse transcriptase inhibitors plus a protease inhibitor or non-nucleoside reverse transcriptase inhibitor. Time to first documented treatment disruption to any component of antiretroviral therapy, derived from treatment records and adherence questionnaires, was analyzed using Kaplan-Meier estimators and Cox proportional hazards models.
RESULTS: The modified intention-to-treat analysis included 263 participants. Seventy-two percent (n = 190) of participants experienced at least one treatment disruption during study. At 4 years, treatment disruption probabilities were 70% (protease inhibitor) vs. 63% (non-nucleoside reverse transcriptase inhibitor). The unadjusted hazard ratio (HR) for treatment disruptions comparing protease inhibitor vs. non-nucleoside reverse transcriptase inhibitor-based regimens was 1.19, 95% confidence interval [CI] 0.88-1.61 (adjusted HR 1.24, 95% CI 0.91-1.68). By study end, treatment disruption probabilities converged (protease inhibitor 81%, non-nucleoside reverse transcriptase inhibitor 84%) with unadjusted HR 1.11, 95% CI 0.84-1.48 (adjusted HR 1.13, 95% CI 0.84-1.50). Reported reasons for treatment disruptions suggested that participants on protease inhibitors experienced greater tolerability problems.
CONCLUSIONS: Children had similar time to treatment disruption for initial protease inhibitor and non-nucleoside reverse transcriptase inhibitor-based antiretroviral therapy, despite greater reported tolerability problems with protease inhibitor regimens. Initial pediatric antiretroviral therapy with either a protease inhibitor or non-nucleoside reverse transcriptase inhibitor may be acceptable for maintaining optimal, continuous therapy.

RCT Entities:   Population Interventions Outcomes

Introduction

Globally, 1.8 million children are living with HIV, and 110,000 die annually due to AIDS-related illnesses [1]. For HIV-infected children, greatest survival outcomes can be achieved only with optimal, uninterrupted treatment on effective antiretroviral therapy (ART). Treatment disruptions, defined as any interruption or alteration of initial ART, may result from patient-level factors (e.g., poor adherence, drug intolerance), provider-level factors (e.g., prescription stops, changes, or errors), or systems-level factors (e.g., stock outs, interruptions in drug delivery). Unfortunately, treatment disruptions may result in treatment failure, acquisition of resistance mutations, and loss of future treatment options—which are particularly consequential in children. Compared with adults, children have greater pharmacokinetic variability and fewer available licensed drugs [2, 3]. Due to longer lifetime antiretroviral exposure, children have more potential for long-term toxicity [4, 5]. Children have greater social vulnerability related to their dependence on others for medical care and medication administration [6, 7]. If inadequately treated, children progress much faster to AIDS and death [8-10]. As children’s initial ART regimens are often their best opportunity for effective, tolerable treatment, optimizing the time on a successful initial regimen may result in greater long-term effectiveness of ART and more lifetime treatment options [11]. Analyzing longitudinal relationships between pediatric ART regimens and time to treatment disruption allows identification of initial ART regimens that pose greater challenges to maintaining optimal, continuous ART.

When deciding which regimen to prescribe to optimize clinical outcomes, clinicians must consider both drug pharmacology and potential adherence to ART regimens [12]. Boosted protease-inhibitor (PI)-based regimens appear more forgiving of treatment disruptions than do non-nucleoside reverse transcriptase inhibitor (NNRTI)-based regimens [13-17]. However, certain PI characteristics decrease adherence and tolerability, particularly in children: poor taste; gastrointestinal toxicity; and regimen complexity, such as pill burden, storage requirements, and dosing frequency [7, 17–22]. Prior pediatric studies that have assessed the ability of children to maintain continuous therapy did not do so in settings in which use of PI- vs. NNRTI-based ART regimens was randomly allocated, nor have prior studies measured treatment disruptions longitudinally. As a result, these previously conducted studies have potential for residual confounding from unmeasured covariates. Furthermore, most studies have isolated analyses of prescription patterns, adherence, and tolerability, rather than evaluating the total effect of the regimen on maintaining optimal, continuous therapy. In the PENPACT-1 study, 266 HIV-1-infected, treatment-naïve children from Europe, North America, and South America were randomized to ART with either a PI or NNRTI and followed longitudinally for at least 4 years [23]. We aimed to assess PENPACT-1 participants for differences in time to treatment disruption across randomized PI vs. NNRTI treatment arms at 4 years and end of study.

Methods

Study design and participants

PENPACT-1 (Paediatric European Network for Treatment of AIDS [PENTA] 9 / Pediatric AIDS Clinical Trials Group [PACTG] 390) was an international multicenter phase 2/3, randomized, open-label trial enrolling children living with HIV-1 from 68 clinical centers in 13 countries in Europe and North and South America between September 25, 2002, and September 7, 2005 (S1 Protocol) [23]. Eligible children aged 31 days to less than 18 years were HIV-1-infected and had not received ART or received only antiretrovirals for <56 days to reduce mother-to-child transmission (excluding single-dose nevirapine). All parents or guardians and children, as appropriate, gave written consent for the parent trial; this protocol was conducted in accordance with the Declaration of Helsinki and approved by the relevant ethics committee or institutional review board (IRB) for each participating center. The secondary analysis on time to treatment disruption was performed under a data request and was reviewed only at IRBs where the analysis was performed. The secondary analysis was deemed exempt by the Duke University IRB and approved by the University of North Carolina-Chapel Hill and Children’s Mercy Kansas City IRBs. This study is registered with the International Standard Randomised Controlled Trial Number Registry (ISRCTN73318385) at https://doi.org/10.1186/ISRCTN73318385 and ClinicalTrials.gov (NCT00039741) at https://clinicaltrials.gov/ct2/show/NCT00039741.

Children were randomized 1:1 to start ART with two nucleoside reverse transcriptase inhibitors (NRTIs) plus either a PI or NNRTI. Randomization was stratified by age (<3 years or ≥3 years); receipt of perinatal ART prophylaxis; and research network (PENTA or PACTG), which varied by region; with variable block sizes. The study was open label, and the treating clinician chose the two NRTI drugs combined with a drug from the randomly assigned PI or NNRTI class. Children underwent clinical and HIV-1 RNA viral load assessments at randomization (week 0), weeks 2, 4, 8, 12, 16, 24, and then every 12 weeks until the last child assigned to treatment reached 4 years of follow-up (August 31, 2009). Treatment starts, changes, and stoppages were recorded at these clinical visits and ad hoc throughout the study. Trained study personnel administered validated adherence questionnaires every 24 weeks after randomization, or if missed, at the following attended visit [24]. Adherence questionnaires were harmonized across networks to collect key data. Specifically, adherence questionnaires recorded the number of missed doses to all antiretrovirals over the 3 days prior to these 24-weekly visits and barriers to adherence experienced within 2 weeks prior to these visits. Four years of follow-up was defined as the week 192 visit plus a 6-week lag to capture late visits.

Outcomes

We defined time to treatment disruption as the number of weeks between randomization and the first documented treatment disruption event. We defined treatment disruption as stopping, switching, or reporting missed doses of any component of the initial ART regimen for any reason except recall of nelfinavir (June 2007) or planned treatment interruptions. Stopping was defined as any duration of treatment discontinuation, regardless of whether treatment was restarted or changed in the future, whereas switches were defined as immediate changes of therapy. Information on ART stoppages or switches was derived from participants’ treatment records, and missed doses were defined as any questionnaire-reported missed doses within 3 days prior to the study visit.

Additional analyses included adjustment for stratified randomization factors (age, receipt of perinatal ART prophylaxis, research network), assessed differences in outcome for the primary follow-up time point (4 years) vs. the entire study, and explored reasons for treatment disruptions. Reasons for treatment disruptions were analyzed using (1) the treatment record’s documented rationale for ART stop or change and (2) any questionnaire-reported barriers to adherence within 2 weeks prior to the visit when missed dose(s) were reported. Only one reason for treatment disruption was allowed on the treatment record; thus a single response, such as “caregiver request,” may not exclude additional reasons. Multiple reasons were allowed on adherence questionnaires.

We assessed the sensitivity of our results to our definition of treatment disruption. Our alternative outcome definitions included restricting treatment record-based treatment disruptions (or any questionnaire event) to drug changes or stops lasting more than 3 days or 14 days and restricting treatment record-based treatment disruptions (or any questionnaire event) to only events including the PI or NNRTI drug component.

Statistical analysis

PI vs. NNRTI treatment groups were assessed according to a modified intention-to-treat (mITT) analysis consistent with the original study [23]. The sole modification was removal of three participants: two who withdrew consent prior to ART initiation, and one with a major eligibility violation. Follow-up began at date of randomization. Participants were right-censored for initial treatment contrary to randomization, planned treatment interruption, death, withdrawal of consent, loss to follow-up, or study end.

For the primary outcome, we estimated the risk of treatment disruptions using the complement of the Kaplan-Meier estimator. We estimated the hazard ratio for treatment disruptions using Cox proportional hazards models. Proportional hazards assumptions were assessed graphically, using time-interaction terms, and with martingale residuals. In adjusted analyses, we stratified by baseline randomized stratification variables: age, exposure to perinatal ART, and research network. Analyses were conducted in SAS® version 9.4 (Cary, NC).

Results

PENPACT-1 enrolled 266 HIV-1 infected children from 68 centers in 13 countries in Europe, North America, and South America. The mITT analysis was restricted to 263 participants who initiated ART. Participants were a median age of 6.5 years at enrollment (IQR [interquartile range], 1.8–12.9), 52% male, 49% black, and 79% exposed to HIV via vertical transmission (Table 1). Fifty-one percent had moderate to severe clinical symptoms (CDC stage B or C). Median growth parameters were below average (weight-for-age Z score -0.6; height-for-age Z score -0.9). Median CD4 Z-score was -3.5, consistent with predominance of moderate to severe immunosuppression, and median viral load was 5.0 log10 copies/mL. Whereas 15% of children had ART exposure for prevention of mother-to-child transmission, 4% had at least one major resistance mutation at baseline. Although treatment groups had differences in racial distribution, baseline characteristics relating to mode of HIV-1 acquisition, clinical and immunological status, and ART resistance were generally balanced across ART regimens, consistent with the randomized design.

Table 1

Baseline characteristics of study participants according to initial ART regimen.

		Randomized Group
Variable		PI	NNRTI	Total
N		131	132	263
Age
<3 years	n (%)	34 (26%)	36 (27%)	70 (27%)
3–17 years	n (%)	97 (74%)	96 (73%)	193 (73%)
Age in years	Median (IQR)	7.1 (2.8, 13.7)	6.4 (2.7, 11.0)	6.5 (2.8, 12.9)
Sex
Male	n (%)	69 (53%)	67 (51%)	136 (52%)
Race
Black, Non-Hispanic	n (%)	60 (46%)	69 (52%)	129 (49%)
White, Non-Hispanic	n (%)	40 (31%)	29 (22%)	69 (26%)
Hispanic/Other	n (%)	31 (24%)	34 (26%)	65 (25%)
Research Networka
PENTA	n (%)	95 (73%)	93 (70%)	188 (71%)
PACTG/IMPAACT	n (%)	36 (27%)	39 (30%)	75 (29%)
Route of Infection
Vertical	n (%)	103 (79%)	106 (80%)	209 (79%)
Other/Unknown	n (%)	28 (21%)	26 (20%)	54 (21%)
CDC Clinical Stage
N	n (%)	27 (21%)	29 (22%)	56 (21%)
A	n (%)	35 (27%)	37 (28%)	72 (27%)
B	n (%)	41 (31%)	43 (33%)	84 (32%)
C	n (%)	28 (21%)	23 (17%)	51 (19%)
Weight-for-Age Z-score	Median (IQR)	-0.5 (-1.6, 0.1)	-0.7 (-1.6, 0.2)	-0.6 (-1.6, 0.1)
Height-for-Age Z-score	Median (IQR)	-0.9 (-1.5, -0.2)	-0.9 (-1.8, 0)	-0.9 (-1.7, -0.2)
CD4 Z score	Median (IQR)	-3.6 (-7.2, -1.7)	-3.4 (-6.5, -1.4)	-3.5 (-6.8, -1.6)
Viral Load log10 copies/mL	Median (IQR)	5.1 (4.5, 5.7)	5.0 (4.5, 5.6)	5.0 (4.5, 5.7)
Perinatal ART Exposure	n (%)	19 (15%)	20 (15%)	39 (15%)
≥1 Major Resistance Mutationb	n/N (%)	5/116 (4%)	5/123 (4%)	10/239 (4%)
HIV-1 subtype
B	n (%)	52 (42%)	49 (39%)	101 (41%)
C	n (%)	13 (11%)	12 (10%)	25 (10%)
F	n (%)	25 (20%)	23 (18%)	48 (19%)
A/CRF_AG/D/G	n (%)	21 (17%)	31 (25%)	52 (21%)
Unclassified	n (%)	12 (10%)	11 (9%)	23 (9%)
Switching Threshold
1,000 copies/mL	n (%)	66 (50%)	68 (52%)	134 (51%)
30,000 copies/mL	n (%)	65 (50%)	64 (48%)	129 (49%)
Duration of Follow-Up in weeks	Median (IQR)	263 (217, 313)	260 (219, 316)	261 (217, 313)

ART, antiretroviral therapy; IQR, interquartile range; N, total sample size; n, subsample size; NNRTI, non-nucleoside reverse transcriptase inhibitor; PACTG, Pediatric AIDS Clinical Trials Group; PENTA, Paediatric European Network for Treatment of AIDS; PI, protease inhibitor.

a PENTA sites were predominantly in Europe, South America, and the Bahamas. PACTG sites were based primarily in the United States.

b Not all patients had successful baseline genotypic resistance assays.

Median follow-up time was 261 weeks (IQR, 217–313). Two participants in each arm were started on a PI or NNRTI contrary to randomization; two underwent planned treatment interruption; five withdrew from study after ART initiation; 37 were lost to follow-up; and one patient died, due to HIV-related complications (Fig 1). Two hundred forty-nine participants ever completed an adherence questionnaire, totaling 2,112 questionnaires over the duration of the study for a mean of 8.5 questionnaires per participant.

Fig 1

Study profile.

ART, antiretroviral therapy; NNRTI, non-nucleoside reverse transcriptase inhibitor; PI, protease inhibitor.

Overall, 191 of 263 participants had at least one treatment disruption event during the study, with 66% (95% confidence interval [CI] 61–72%) treatment disruption probability at 4 years (primary follow-up period) and 83% (95% CI 76–91%) treatment disruption probability at study end (6.5 years). At 4 years, probabilities of treatment disruption were 70% (95% CI 62–78%) vs. 63% (95% CI 55–72%) in the PI and NNRTI arms, respectively (Fig 2). Hazards for treatment disruption, however, were similar for PI vs. NNRTI-based regimens (unadjusted hazard ratio [HR] 1.19, 95% CI 0.88–1.61), even after adjustment for stratification factors of age, receipt of perinatal ART, and research network/region (adjusted HR 1.24, 95% CI 0.91–1.68).

Fig 2

Proportion of children experiencing treatment disruption from initial ART regimen by study week.

The vertical line delineates 4 years on study. ART, antiretroviral therapy; n, subsample size; NNRTI, non-nucleoside reverse transcriptase inhibitor; PI, protease inhibitor.

After 4 years, treatment disruption probabilities converged, such that treatment disruption probabilities at study end were 81% (95% CI 72–90%) for PI vs. 84% (95% CI 73–94%) for NNRTI arms, but changes over time in the hazard ratio of treatment disruption by treatment arms were non-significant (unadjusted P for interaction = 0.33, adjusted P = 0.21). Hazards for treatment disruption over the entire study period were similar for PI vs. NNRTI-based regimens, unadjusted (HR 1.11, 95% CI 0.84–1.48) and adjusted (HR 1.13, 95% CI 0.84–1.50).

Of 191 treatment disruption events, 126 events were based on ART regimen stoppages or changes in the treatment record, and 67 events were reported missing doses on adherence questionnaires, with two participants experiencing both event types simultaneously. Of the treatment stops or changes, 25% of events were substitutions of at least one first-line ART drug (PI 32%, NNRTI 16%), 53% were stoppage or suspension of the entire first-line ART regimen (PI 48%, NNRTI 59%), and 22% were switches to a second-line ART regimen (PI 20%, NNRTI 25%). Most frequent reasons documented for ART stops or changes were adverse events (34%), viral failure (22%), caregiver request (18%), non-adherence (7%), and temporary break (6%), with the greatest difference between PIs over NNRTIs for adverse events (Table 2).

Table 2

Reasons listed for treatment disruption events.

Reason / Barrier		PI	NNRTI	Total
Treatment Recorda
Adverse event	n (%)	24 (37%)	19 (31%)	43 (34%)
Viral failure	n (%)	13 (20%)	15 (25%)	28 (22%)
Caregiver request	n (%)	11 (17%)	12 (20%)	23 (18%)
Non-adherence	n (%)	6 (9%)	3 (5%)	9 (7%)
Temporary break	n (%)	3 (5%)	5 (8%)	8 (6%)
Unknown	n (%)	5 (8%)	1 (2%)	6 (5%)
Drug supply problem	n (%)	1 (2%)	2 (3%)	3 (2%)
Intercurrent illness	n (%)	0 (0%)	2 (3%)	2 (2%)
Resistance	n (%)	1 (2%)	1 (2%)	2 (2%)
Parent forgot	n (%)	1 (2%)	0 (0%)	1 (1%)
Simplification	n (%)	0 (0%)	1 (2%)	1 (1%)
Treatment record total	n	65	61	126
Adherence Questionnaireb
Forgot/lack of support	n (%)	10 (29%)	10 (30%)	20 (30%)
Ran out of drug	n (%)	8 (24%)	9 (27%)	17 (25%)
Problems taking some of the drugs (e.g., intolerance, taste, medication volume)	n (%)	11 (32%)	5 (15%)	16 (24%)
Fear of disclosure to others	n (%)	10 (29%)	5 (15%)	15 (22%)
Patient refused/didn’t want to take drugs	n (%)	10 (29%)	4 (12%)	14 (21%)
Scheduling/lifestyle interference	n (%)	9 (26%)	3 (9%)	12 (18%)
Drug toxicity concerns	n (%)	7 (21%)	4 (12%)	11 (16%)
Supervised by someone else or multiple caregivers	n (%)	6 (18%)	5 (15%)	11 (16%)
Patient unwell	n (%)	6 (18%)	4 (12%)	10 (15%)
Other	n (%)	4 (12%)	5 (15%)	9 (13%)
Different routine/change in living situation	n (%)	3 (9%)	4 (12%)	7 (10%)
Fed up giving/taking drugs	n (%)	3 (9%)	2 (6%)	5 (7%)
Think medication is not needed or not helping	n (%)	2 (6%)	2 (6%)	4 (6%)
Caregiver unwell/depressed	n (%)	0 (0%)	0 (0%)	0 (0%)
Total listed problems on questionnaireb	n	89	62	151
Total participants with questionnaire-reported missed doses	n	34	33	67
Total Treatment Disruption Eventsc	n	97	94	191

n, subsample size or number of events; NNRTI = non-nucleoside reverse transcriptase inhibitor; PI = protease inhibitor.

a One category allowed per treatment record change or stop.

b Participants may have answered in more than one category.

c Some participants had both a treatment record and adherence questionnaire event at the same time.

Reports of missed doses on adherence questionnaires were balanced between PI and NNRTI arms, as 35% of non-adherence events in each arm were from patient or caregiver reports. The most common questionnaire-reported barriers to adherence, forgetting/lacking support (30%) or running out of medications (25%), were balanced between PI and NNRTI regimens. Other common questionnaire-reported adherence problems—including difficulties with administration, such as those attributed to intolerance, taste, patient refusal (24%); fear of disclosure to others (22%); patient refusal (21%); difficulties with scheduling or lifestyle (18%); and concerns about drug toxicity (16%)—were more frequently reported in participants in the PI arm (Table 2).

In sensitivity analyses, modifications of the outcome definition did not result in substantial hazard ratio changes. Point estimates at 4 years remained similar to the primary analysis when restricting events on the treatment record (or any event on questionnaire) to only ART regimen stops or changes lasting >3 days (unadjusted HR 1.16, 95% CI 0.85–1.57; adjusted HR 1.19, 95% CI 0.88–1.63), only ART regimen stops or changes lasting >14 days (unadjusted HR 1.27, 95% CI 0.93–1.74; adjusted HR 1.32, 95% CI 0.96–1.81), or only stops or changes including the PI or NNRTI drug (unadjusted HR 1.14, 95% CI 0.84–1.55; adjusted HR 1.18, 95% CI 0.87–1.61).

Discussion

In PENPACT-1, our estimates were not compatible with large differences in time to treatment disruption between participants randomized to PIs versus NNRTIs. Point estimates were mildly in the direction of more treatment disruptions in PI-based regimens, particularly in the primary end point of 4 years, but differences were small, possibly due to chance, and appeared to decrease by study end. Exploration of reasons for treatment disruptions suggested that PI-based regimens may be less tolerable, both due to adverse events leading to treatment stoppages or substitutions and to regimen-specific adherence barriers reported on the adherence questionnaire. However, these PI-associated difficulties did not interrupt continuous therapy to the initial PI-based regimens more than they did to NNRTI-based regimens.

Although we did not find a meaningful difference in treatment disruptions in PI vs. NNRTI-based regimens, the secondary analyses exploring reasons for treatment disruptions suggested that administration of a PI-based regimen to a child may be a struggle, even if not resulting in actual missed doses. The treatment record suggested that participants experienced more adverse events to PIs over NNRTIs, but adherence questionnaire responses formed a pattern of difficulties with PI tolerability, whether attributed to taste, medication volume or pill burden, toxicity, or simply patient refusal. This pattern would be consistent with existing literature on PI vs. NNRTI regimens. PIs have higher drug toxicity, especially gastrointestinal side effects, and intolerance, particularly regarding their noxious taste [7, 18–20, 25–27]. Even if children are able to swallow pills, certain PIs are available only as large pills [28, 29]. At the time of PENPACT-1, no PIs were available as complete-regimen combinations for children, whereas single-tablet NNRTI regimens could facilitate adherence through administration of fewer pills [2, 30–34]. More recently, a novel four-in-one fixed-dose combination of abacavir, lamivudine, and LPV/r granule-filled capsules has been under study and submitted to the FDA for approval [35]. Participants reported more barriers to adherence in PIs related to scheduling or lifestyle interference, which may relate to dosing frequency. We hypothesize that increased fear of disclosure to others, as noted in the PI arm, may relate to difficulties concealing drug administration when given more frequently. Higher dosing frequency has been associated with more frequent treatment disruptions [20, 30, 33, 34, 36–39]. Some NNRTIs, most notably efavirenz, have more suitable pharmacokinetics for once daily administration. In our study, most PI-based regimens were administered at least twice daily, whereas some commonly used NNRTI-based regimens allowed once-daily dosing.

Most children in PENPACT-1 experienced a treatment disruption event during the study. Only about one-third of participants remained continuously on their initial ART at 4 years; only one-sixth remained continuously on initial ART at study end. These results are consistent with other pediatric data on the durability of first-line treatment regimens [40]. Maintaining continuous therapy on ART is critical to sustained HIV-related outcomes, as suppressing viral load decreases the probability of HIV sub-populations acquiring antiretroviral resistance mutations and chances of forward infection [41-49]. Although optimal adherence targets vary by PI vs. NNRTI class, adherence has been modest across ART studies, especially patients failing to achieve viral suppression [12, 16, 17, 30, 50–56]. Notably, ART appears to be less successful in producing viral suppression in children, who are more prone to viral failure and resistance due to higher plasma viral loads, less robust antiviral immune responses, greater pharmacokinetic variability, and social dependency [44, 57]. Adolescents have particularly worse viral and immunological outcomes, due to poor ART adherence [48, 49, 58–60]. The large proportion of children in PENPACT-1 with disruptions of their initial ART raises concerns regarding long-term durability, especially as these patients were receiving adherence support on a clinical trial protocol at specialty pediatric HIV centers.

Based on our data, choice of an initial PI- vs. NNRTI-based regimen may not have a major impact on ART treatment disruption. Despite differences in reported regimen-related adherence barriers, participants in both treatment arms persevered in taking their regimens similarly. Moreover, the most common questionnaire-reported barriers were not regimen-specific: forgetting/lack of support and running out of drug. Novel interventions may still be able to improve the experience of drug administration. Pediatric pellets are heat-stable and generally more acceptable than syrups, but palatability and administration problems persist and may increase over time [61-63]. Pediatric granules, especially in the four-in-one combination, may improve palatability and decrease pill burden [35, 64]. Precision medicine related to taste-sensing genotypes may hold promise for prescribing according to individualized palatability [65]. In adult data, integrase strand transferase inhibitors (INSTIs) have been at least as tolerable as PIs or NNRTIs, if not more so, and INSTIs are increasingly preferred drugs in children [66-69]. Nevertheless, a primary goal of optimizing continuous therapy to ART is durable viral suppression, which was comparable across PI vs. NNRTI arms in this study’s parent trial, although similar trials had variable results [23, 70–74]. In this study population, choice of either PI- or NNRTI-based initial ART appears acceptable.

Our estimates of treatment disruption may have had measurement error. First, we had no direct measures of drug exposure, such as therapeutic drug monitoring. Treatment records captured only prescribing events and documented ART disruptions, and the adherence questionnaires relied on accurate reporting by either the child or the caregiver, if present and willing to answer. Although we relied on a questionnaire that has previously been validated [24], reporting biases and unanswered questionnaires may have affected our measures of missed doses. Our combining treatment records and adherence questionnaires into a composite outcome should have decreased measurement error from either instrument individually. Second, adherence questionnaires in this study focused on ART adherence over the 3 days prior to the most recent visit and inquired about adherence barriers encountered over the prior 2 weeks, rather than a daily measure of adherence throughout the study. The time-varying nature of treatment disruption means that patients may have experienced an initial or temporary period of treatment disruption that was subsequently corrected [75, 76], but our analysis presents only data on time to first event of treatment disruption. Third, limited participant report of individual drugs missed on the adherence questionnaire precluded definitive identification of treatment disruptions of individual drugs. Instead, we assessed treatment disruption to any component of the ART regimen. Fourth, heterogeneity of adherence questionnaires across networks, ages, and respondents regarding barriers to therapy should caution against rigorous interpretation of reasons for treatment disruptions. Finally, this study size was not sufficient to distinguish differences on the order of 7%, as was seen at 4 years.

Conclusions

In conclusion, children in PENPACT-1 had similar time to treatment disruption for initial PI-based regimens and NNRTI-based regimens. Although secondary analyses suggest that PI-based regimens may be more difficult to tolerate and may be less convenient to administer, these difficulties did not result in a large difference in children stopping, changing, or missing doses at 4 years (PI 70%, NNRTI 63%), and any suggested differences diminished by study end (PI 81%, NNRTI 84%). Initial ART with either a PI or NNRTI may be acceptable for maintaining continuous therapy on ART in children.

CONSORT checklist.

(DOC)

Click here for additional data file.

PENPACT-1 trial protocol.

(PDF)

Click here for additional data file.

30 Jun 2020

Please carefully address the reviewer's comments and suggestions.

Please submit your revised manuscript by Aug 13 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

We look forward to receiving your revised manuscript.

Kind regards,

Patricia Evelyn Fast, MD, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments:

This paper applies a new analysis to the PENPACT-1 data, to investigate time to treatment interruption for Protease-Inhibitor-based vs NNRTI-based regimens over a 4 year period. The paper is well written and convincing. Two reviewers have provided valuable suggestions to improve the presentation of results.

In addition to addressing the reviewers' comments, please comment on whether important changes in the treatment regimens used, including the availability of palatable protease inhibitors have occurred during the years since the trial ended.

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following in the Competing Interests section:

"Although the authors have no financial competing interests for this manuscript, DEY has been an investigator on studies unrelated to HIV supported by Astellas, Chimerix, Merck, Pfizer, Viracor-Eurofins, Kansas City Area Life Sciences Institute, and the Marion Merrell Dow Fund; DEY was a founder and is an unpaid advisory board member for the non-profit Maipelo Children’s Trust/Cover the Globe, which provides HIV services in Botswana. The other authors report no conflicts of interests. "

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests).  If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

3. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

4. One of the noted authors is a group 'PENPACT-1 (PENTA 9 / PACTG 390) Study Team'. In addition to naming the author group and listing the individual authors and affiliations within this group in the acknowledgments section of your manuscript, please also indicate clearly a lead author for this group along with a contact email address.

5. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Absence of evidence is not evidence of absence so said the late great Doug Altman writing along with his longtime co-author Martin Bland back in 1985 in their great Statistics Notes series in the BMJ. The caution raised there, I think, applies somewhat to the well presented and succinct piece of work laid out in this mansuscript. The present study is based on an extended followup of a large trial that had an extensive geographical footprint looking at paediatric HIV within children aged 1 month through to just under 18 years. The parent trial was a 2x2 factorial study whose aim was to look at changes in viral load after 4 years. The present study looked to asesss whether tehre was a differenec in the timing of treatment disruption both at the end of the initial followup period which was 4 years and at the end of study at 6.5 years.

I have some issue with some of the wording and presentation of the data in lines 194 through 197. Specifically I would like to see 95% Confidence Intervals for all the estimates given ie the 66% and 83% on line 195, the 70% and 63% on line 197 and the estimates of 81% and 84% on line 212. I believe the term "treatment distruption probability" on lines 195 and 196 would be better put as eg "66% 95CI(LB,UB) had a treatment disruption" and on line 196 consider changing "At 4 years..." to something like "By four years 70% (95 CI(LB,UB) and 63%... had experienced at least one treatment disruption respectively".

Table 2 seems rather unnecessary given that all the information it contains already and fully exists within the text on lines 199-201 and 215-6.

I would suggest that a) given this is secondary analysis of pre-existing data and b) that this hints a little at a post-hoc power analysis, the last sentence of the results could be excised without any loss of import for this work.

Finally, I always find the use of a risk table (showing numbers at risk and events occuring within a time interval) under KM plots to be most helpful and would respectfully request that the authors consider adding these to their Figure 2.

Reviewer #2: The authors have not made the underlying data readily available; access is restricted to protect the privacy and confidentiality of the study participants. The authors have provided an institutional email address through which researchers can access the primary data. This is in line with the Plos data policy.

The authors need to clarify in the ethics statement why exemption from an IRB outside the USA was not sought for the secondary analysis.

Line 76, consider stating that '...children progress much faster to AIDS and death' rather than refer to HIV disease.

Line 78 implies that longer time on their initial regimen means greater efficacy. Please confirm whether there is any evidence that duration of treatment improves efficacy of ART; in the context you have used, it might be more appropriate to state that longer time on the initial regimen may result in greater effectiveness of ART.

Line 116; please confirm if block randomization was carried out with blocks based on the site of enrollment.

Line 124; it would be important to clarify what data was collected on the adherence questionnaires that were administered every 24 weeks and what data was collected during the other scheduled visits, or ad hoc as stated in line 122.

As it reads now, it implies that data on treatment disruptions for each drug were collected covering 3 days prior to every 24 week visit when the questionnaire was administered and barriers to adherence experienced in the 2 weeks prior to this visit. This is a very narrow window to collect this data over 4 years.

Line 146 implies that the adherence questionnaire was administered strictly every 24 weeks. As the visit schedule meant volunteers were seen at least once every 12 weeks, please clarify if it was permissible for an adherence questionnaire to e administered in the visit immediately following a 24 week visit when it was missed.

Lines 167 and 173; please confirm whether the classification of race was based on self-reporting at all sites.

Line 220; please confirm if the 25% of events reported, involved a substitution of either the PI or NNRTI or if these cases include instances of substitution of one or more of the NRTIs without interference with the administration of the PI or NNRTI in the regimen.

Line 221; please clarify if stoppage or suspension of the entire first line regimen resulted in starting the second line regimen in all cases. Please also specify how the group that had the first line regimen stopped differs from the group that had switches to a second-line regimen.

Table 3; Please clarify if caregiver requests were distinguished from adverse events. Please state in line 24 if the caregiver requests listed in the table were all those that were for reasons other than adverse events.

Line 332; please clarify in the methods section if the adherence questionnaire was standardized across study sites and age groups. The statement on line 332 implies that the questionnaires differed across sites and age groups.

Figure 1; what was the reason for volunteers who withdrew consent after initiation of ART?

There were more cases of withdrawal of consent in the NNRTI group than the PI group; please state whether any of these withdrawals were related to adverse events or issues of tolerability of the drugs?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Greg Fegan

Reviewer #2: Yes: Vincent Muturi-Kioi

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

30 Sep 2020

Thank you for your kind, thoughtful review of our manuscript. We have revised the manuscript and included a letter responding to academic editor and reviewer comments. Our responses are also listed below.

Editor Comments

1. “In addition to addressing the reviewers' comments, please comment on whether important changes in the treatment regimens used, including the availability of palatable protease inhibitors have occurred during the years since the trial ended.”

We address this point on pages 16 and 18 of the revised manuscript.

2. “Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at [Websites removed].”

We have modified our style accordingly.

3. “Thank you for stating the following in the Competing Interests section:

‘Although the authors have no financial competing interests for this manuscript, DEY has been an investigator on studies unrelated to HIV supported by Astellas, Chimerix, Merck, Pfizer, Viracor-Eurofins, Kansas City Area Life Sciences Institute, and the Marion Merrell Dow Fund; DEY was a founder and is an unpaid advisory board member for the non-profit Maipelo Children’s Trust/Cover the Globe, which provides HIV services in Botswana. The other authors report no conflicts of interests.’

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests). If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.”

We confirm that these competing interests do not alter our adherence to PLOS ONE policies on sharing data and materials, and we have included this statement in our resubmission.

4. “Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.”

We have included our updated Competing Interests statement in our cover letter.

5. “We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.”

IMPAACT and PENTA are willing to share the data, upon review and approval of a data request. The information for such a data request are provided.

Please note that the original data were obtained by our own data request to IMPAACT, and publication of our results required approval of PACTG/IMPAACT and PENTA. As such, the data are not our own to share, but rather are of PACTG/IMPAACT and PENTA. In addition, the data we received are both potentially identifiable, as determined by the multiple IRBs listed in the manuscript. Moreover, the data are considered sensitive, as they include the HIV diagnosis of participants in the trial. Therefore, receipt of the data for our analysis was contingent upon our not sharing the data, given its sensitive nature. Our original statement on not being able to share the data is verbatim from the IMPAACT instructions we were given.

6. “One of the noted authors is a group 'PENPACT-1 (PENTA 9 / PACTG 390) Study Team'. In addition to naming the author group and listing the individual authors and affiliations within this group in the acknowledgments section of your manuscript, please also indicate clearly a lead author for this group along with a contact email address.”

The listed senior author of this manuscript is Ross E. McKinney, Jr., who was a North American co-chair of the study and one of the authors of the parent study primary manuscript. We have included his contact email address on page 2.

5. “Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.”

We have included captions for the Supporting Information files and updated in-text citations accordingly.

Reviewer comments

Reviewer 1

1. “Absence of evidence is not evidence of absence so said the late great Doug Altman writing along with his longtime co-author Martin Bland back in 1985 in their great Statistics Notes series in the BMJ. The caution raised there, I think, applies somewhat to the well presented and succinct piece of work laid out in this manuscript. The present study is based on an extended followup of a large trial that had an extensive geographical footprint looking at paediatric HIV within children aged 1 month through to just under 18 years. The parent trial was a 2x2 factorial study whose aim was to look at changes in viral load after 4 years. The present study looked to assess whether there was a difference in the timing of treatment disruption both at the end of the initial followup period which was 4 years and at the end of study at 6.5 years.”

Thank you for your comments. We agree that absence of evidence is not evidence of absence. We have re-worded our statements to reflect that our estimates are not compatible with large differences between treatment groups in time to treatment disruption. Changes are noted on page 15.

2. “I have some issue with some of the wording and presentation of the data in lines 194 through 197. Specifically I would like to see 95% Confidence Intervals for all the estimates given ie the 66% and 83% on line 195, the 70% and 63% on line 197 and the estimates of 81% and 84% on line 212. I believe the term "treatment disruption probability" on lines 195 and 196 would be better put as eg "66% 95CI(LB,UB) had a treatment disruption" and on line 196 consider changing "At 4 years..." to something like "By four years 70% (95 CI(LB,UB) and 63%... had experienced at least one treatment disruption respectively".”

We agree with including 95% confidence intervals with these estimates and have inserted them as recommended on pages 12-13.

2. “Table 2 seems rather unnecessary given that all the information it contains already and fully exists within the text on lines 199-201 and 215-6.”

We appreciate Reviewer #1’s suggestion and have eliminated this table. We have renumbered our tables accordingly.

3. “I would suggest that a) given this is secondary analysis of pre-existing data and b) that this hints a little at a post-hoc power analysis, the last sentence of the results could be excised without any loss of import for this work.”

Thank you for pointing out that this information suggests a post-hoc power analysis. This information was included as a sensitivity analysis to show potential influences of various treatments of missing questionnaire data. However, given the potential for misinterpretation, we have removed this sentence (page 15) and removed the corresponding descriptions from the Methods (page 9).

4. “Finally, I always find the use of a risk table (showing numbers at risk and events occuring within a time interval) under KM plots to be most helpful and would respectfully request that the authors consider adding these to their Figure 2.”

We appreciate this suggestion and have revised Figure 2 to include the numbers at risk, censored, and events occurring within the time intervals.

Reviewer 2

1. “The authors have not made the underlying data readily available; access is restricted to protect the privacy and confidentiality of the study participants. The authors have provided an institutional email address through which researchers can access the primary data. This is in line with the Plos data policy.”

Thank you for discussing this matter. As per academic editor recommendations, we have submitted the privacy risks and sensitive nature of the data, as well as recognizing that the data are not ours to share, but rather were obtained by our own data request. We have included the official statement of IMPAACT regarding data access and the mechanism by which these data may be requested.

2. “The authors need to clarify in the ethics statement why exemption from an IRB outside the USA was not sought for the secondary analysis.”

As these data were received via a data request from the central data management center, we only sought approval from the institutional directly overseeing the investigators performing the secondary analysis. We have included this clarification in the Methods section (page 7).

3. “Line 76, consider stating that '...children progress much faster to AIDS and death' rather than refer to HIV disease.”

Thank you for this important distinction. We have edited this sentence accordingly (page 5).

4. “Line 78 implies that longer time on their initial regimen means greater efficacy. Please confirm whether there is any evidence that duration of treatment improves efficacy of ART; in the context you have used, it might be more appropriate to state that longer time on the initial regimen may result in greater effectiveness of ART.”

Thank you for this clarification. We have re-phrased as per the reviewer’s suggestion, cited data showing that 2nd line treatment failure is common and problematic on page 5, and cited corroborating data on pediatric first-line durability on page 17.

5. “Line 116; please confirm if block randomization was carried out with blocks based on the site of enrollment.”

Randomization was both stratified and in blocks. Stratification was within categories of age, receipt of perinatal ART, and within the research network (PACTG or PENTA), not by site. Trial statisticians employed variable block sizes. We have clarified the blocked randomization design in the manuscript (page 7).

6. “Line 124; it would be important to clarify what data was collected on the adherence questionnaires that were administered every 24 weeks and what data was collected during the other scheduled visits, or ad hoc as stated in line 122. As it reads now, it implies that data on treatment disruptions for each drug were collected covering 3 days prior to every 24 week visit when the questionnaire was administered and barriers to adherence experienced in the 2 weeks prior to this visit. This is a very narrow window to collect this data over 4 years.”

We have clarified the more frequent clinic visits during which the treatment records were recorded and the less frequent, 24-weekly visits on which the adherence questionnaire were collected (pages 7-8). We combined measures to balance the narrow window of data collection identified by the reviewer.

7. “Line 146 implies that the adherence questionnaire was administered strictly every 24 weeks. As the visit schedule meant volunteers were seen at least once every 12 weeks, please clarify if it was permissible for an adherence questionnaire to e administered in the visit immediately following a 24 week visit when it was missed.”

Yes, if the visitor adherence questionnaire was missed at 24 weekly intervals, the adherence questionnaires may have been, and often were, administered at the following visit. We have included this information in the manuscript (page 7).

8. “Lines 167 and 173; please confirm whether the classification of race was based on self-reporting at all sites.”

Yes, race/ethnicity classification was based on self-report or report of the parent/caregiver.

9. “Line 220; please confirm if the 25% of events reported, involved a substitution of either the PI or NNRTI or if these cases include instances of substitution of one or more of the NRTIs without interference with the administration of the PI or NNRTI in the regimen.”

Due to the lack of granularity in the adherence questionnaire, we could not distinguish whether questionnaire-reported treatment disruptions were only due to PI or NNRTI vs. NRTI backbone with confidence. Such could be distinguished on the treatment record. Thus, we performed the primary analysis using the least common denominator, which was disruption of any component of the initial ART regimen (page 8, 19). To account for potential influence of changing only the NRTI backbone, we also performed sensitivity analyses including only changes in the PI or NNRTI, as per the treatment records, recognizing that we would be mis-specifying events from the adherence questionnaire (page 9). The results were largely the same as the primary analysis.

10. “Line 221; please clarify if stoppage or suspension of the entire first line regimen resulted in starting the second line regimen in all cases. Please also specify how the group that had the first line regimen stopped differs from the group that had switches to a second-line regimen.”

A stoppage or suspension of the entire first line regimen may have led to either a permanent discontinuation or a subsequent switch to a 2nd line regimen (page 8, line 142), and we explored different definitions for duration of time off therapy that qualified as a treatment discontinuation (pages 9, line 160; 15, lines 264-266). Permanent discontinuations would also include such events as loss-to-follow-up or withdrawal from study. As a result, we do not have measurement of potential future re-starting of ART for patients where were lost or withdrew.

11. “Table 3; Please clarify if caregiver requests were distinguished from adverse events. Please state in line 24 if the caregiver requests listed in the table were all those that were for reasons other than adverse events.”

The treatment record allowed listing of only one reason for stops or changes in the initial treatment regimen. Although subsequent information from future visits may suggest that some caregiver requests were related to adverse events, differentiating caregiver requests that were or were not related to adverse events requires the authors to read into the records, rather than reporting on the records. Thus, we chose to use the reason first listed at the treatment disruption event, which we expect would introduce less author bias in interpretation. We have added a comment in the Methods on page 8 to clarify.

12. “Line 332; please clarify in the methods section if the adherence questionnaire was standardized across study sites and age groups. The statement on line 332 implies that the questionnaires differed across sites and age groups.”

Thank you for this comment. We now address this issue on page 7 of Methods. Each research network (PENTA or PACTG) had its own standardized questionnaires. Key information was harmonized across questionnaires, but each network used their own standardized questionnaire with some remaining differences.

We are willing to provide copies of the adherence questionnaires as Supplementary Material if the Editor deems this necessary.

13?. “Figure 1; what was the reason for volunteers who withdrew consent after initiation of ART? There were more cases of withdrawal of consent in the NNRTI group than the PI group; please state whether any of these withdrawals were related to adverse events or issues of tolerability of the drugs?”

Unfortunately, we do not have access to these reasons for withdrawal of consent.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

3 Nov 2020

Time to treatment disruption in children with HIV-1 randomized to initial antiretroviral therapy with protease inhibitors versus non-nucleoside reverse transcriptase inhibitors

PONE-D-20-04892R1

Dear Dr. Yin,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Patricia Evelyn Fast, MD, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Thank you for your careful attention to the reviewers' comments.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: (No Response)

Reviewer #2: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Greg Fegan

Reviewer #2: Yes: Vincent Muturi-Kioi

12 Nov 2020

PONE-D-20-04892R1

Time to treatment disruption in children with HIV-1 randomized to initial antiretroviral therapy with protease inhibitors versus non-nucleoside reverse transcriptase inhibitors

Dear Dr. Yin:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Patricia Evelyn Fast

Academic Editor

PLOS ONE