Nucleoside Analogs Plus Ritonavir in Stable Antiretroviral Therapy–Experienced HIV-Infected Children a Randomized Controlled Trial

ORIGINAL CONTRIBUTION

Nucleoside Analogs Plus Ritonavir in Stable Antiretroviral Therapy–Experienced HIV-Infected Children A Randomized Controlled Trial

Sharon A. Nachman, MD Context Although protease inhibitors are used routinely in adults with human im- Kenneth Stanley, PhD munodeficiency virus (HIV) infection, the role of these drugs in the treatment of clini- Ram Yogev, MD cally stable HIV-infected children is not clear. Objective To evaluate the safety, tolerance, and virologic response produced by a Stephen Pelton, MD change in antiretroviral therapy in HIV-infected children who were clinically and im- Andrew Wiznia, MD munologically stable while receiving previous therapy. Sophia Lee, MS Design The Pediatric AIDS Clinical Trials Group 338, a multicenter, phase 2, randomized, open-label controlled trial conducted from February 6 to April 30, 1997 (pa- Lynne Mofenson, MD tient entry period); patients were followed up for 48 weeks. Susan Fiscus, PhD Setting Pediatric HIV research clinics in the United States and Puerto Rico. Mobeen Rathore, MD Patients Two hundred ninety-seven antiretroviral-experienced, protease inhibitor– Eleanor Jimenez, MD naive, clinically stable HIV-infected children aged 2 to 17 years. William Borkowsky, MD Interventions Children were randomized to receive zidovudine, 160 mg/m2 3 times per day, plus lamivudine, 4 mg/kg 2 times per day (n = 100); the same regimen plus Jane Pitt, MD ritonavir, 350 mg/m2 2 times per day (n = 100); or ritonavir, 350 mg/m2 2 times per Mary E. Smith, MD day, and stavudine, 4 mg/kg 2 times per day (n = 97). Barbara Wells, BA Main Outcome Measure Plasma HIV-1 RNA levels at study weeks 12 and 48, compared among the 3 treatment groups. Kenneth McIntosh, MD Results At study week 12, 12% of patients in the zidovudine-lamivudine group had for the Pediatric AIDS Clinical Trials undetectable plasma HIV RNA levels (Ͻ400 copies/mL) compared with 52% and 54% Group 338 Study Team of patients in the 2- and 3-drug ritonavir-containing groups, respectively (PϽ.001). Through study week 48, 70% of children continued receiving their ritonavir- N EARLY 1997 MANY CHILDREN IN- containing regimen. At study week 48, 42% of children receiving ritonavir plus 2 nucleo- fectedwithhumanimmunodeficien- sides compared with 27% of those receiving ritonavir and a single nucleoside had un- cy virus (HIV) were receiving single detectable HIV RNA levels (P = .04); however, similar proportions in each group or dual nucleoside analog therapy continuing initial therapy had HIV RNA levels of less than 10 000 copies/mL (58% vs andI were clinically and immunologically 48%, respectively; P = .19). 1,2 stable, despite increasing plasma HIV Conclusions In our study, change in antiretroviral therapy to a ritonavir-containing RNA levels. However, the availability of regimen was associated with superior virologic response at study week 12 compared protease inhibitors and their demon- with change to a dual nucleoside analog regimen. More children receiving ritonavir in strated success in adults raised the ques- combination with 2 compared with 1 nucleoside analog had undetectable HIV RNA tion of how they should be used in chil- levels at study week 48. dren. Pediatric AIDS Clinical Trials JAMA. 2000;283:492-498 www.jama.com Group (PACTG) Protocol 338 was un- dertaken to evaluate the safety, tolerance, dual nucleoside analog therapy or a 2- Author Affiliations, Study Team Members, and Finan- and virologic efficacy of changing from or 3- drug regimen containing a prote- cial Disclosure are listed at the end of this article. current antiretroviral treatment in clini- ase inhibitor, ritonavir. Corresponding Author and Reprints: Sharon A. Nach- man, MD, State University of New York at Stony Brook, cally and immunologically stable, pro- Current antiretroviral guidelines for HSC T11, 080, Stony Brook, NY 11794-8111 (e-mail: tease inhibitor–naive children to either treatment of HIV-infected children and [email protected]).

Downloaded From: https://jamanetwork.com/ on 09/29/2021 RITONAVIR IN HIV-INFECTED CHILDREN

adults recommend initiation of therapy same regimen with 350 mg/m2 of rito- cluded a medical history, physical ex- with a combination of antiretroviral navir (maximal dose, 600 mg/dose) twice amination, a complete blood cell count drugs, including a protease inhibi- per day; or 350 mg/m2 of ritonavir twice with differential and serum chemis- tor.3,4 Preliminary results from PACTG per day and 4 mg/kg of stavudine (maxi- tries. Lymphocyte surface markers were Protocol 3385 were instrumental in mal dose, 40 mg/dose) twice per day. evaluated at preentry; entry; study weeks clarifying the role of protease inhibi- The primary objective of the study 4, 8, 12, and every 12 weeks thereafter tors for children in these guidelines. was to evaluate the safety and toler- and were performed by local laborato- Therapeutic results at 1 year are pre- ance of these 3 treatment regimens and ries participating in the National Insti- sented in this article. compare the change in plasma HIV tute of Allergy and Infectious Diseases RNA copy number between entry and (NIAID) Flow Cytometry Quality As- METHODS study weeks 12 and 48 among the regi- surance Program.7 Specimens for HIV-1 Study Design and Patients mens. The duration of study treat- RNA were obtained at preentry; entry; The PACTG 338 study was a multi- ment for each patient was initially and at study weeks 4, 12, 24, 36, 44, 48, center, randomized, phase 2 clinical trial planned to be 48 weeks but was sub- and every 12 weeks thereafter. The that compared change from current sequently extended to 120 weeks. Only NucliSens Assay (Organon Teknika, therapy to either zidovudine plus lam- 48-week outcomes are presented here. Durham, NC)8 was used in the assess- ivudine or 1 of 2 ritonavir-containing Randomized study treatment was dis- ment of HIV RNA copy number.9 Speci- regimens (a 3-drug regimen of ritona- continued for children who experi- mens from individuals from preentry vir, zidovudine, and lamivudine or a enced a virologic failure, disease pro- through week 12 were assayed for HIV 2-drug regimen of ritonavir and stavu- gression, or persistent grade 3 or higher RNA copy number in batched fashion; dine) in HIV-infected, clinically stable drug-related adverse effects. Such chil- subsequent specimens were run at the children. All subjects were aged 24 dren were offered the best available specified individual time points. The months to 17 years; had stable CD4 cell therapy at the discretion of their clini- lower limit of assay quantification for number or percentage maintained in cian, themselves, or their parents. All RNA was 400 copies/mL. Assay results Centers for Disease Control and Pre- subjects were enrolled for the full study were adjusted using Virology Quality As- vention (CDC) immune category 1 or period of 48 weeks. surance (VQA) standards.10 2 during the 4 months prior to study en- A primary virologic failure at week All adverse events were graded us- try6; had experienced no new CDC clini- 12 was defined as a failure to achieve ing protocol-specified, standard toxic- cal category C diagnosis in the 12 either HIV RNA copy number of at least ity criteria for pediatric populations. months prior to study entry; received 2.0 log10 copies/mL below baseline val- Only those events occurring during continuous antiretroviral therapy in the ues or to maintain at least 10 000 copies/ therapy or less than 60 days after ter- 16 weeks prior to study entry; and were mL. A subsequent virologic failure was mination of initial study medication either zidovudine- and lamivudine- defined as a persistent 0.75 log10 in- were included in the analysis. naive or had received no more than 6 crease above the nadir in HIV RNA copy weeks of zidovudine and lamivudine in number for patients who had an ini- Statistical Analysis the year prior to study entry and none tial HIV RNA copy number decrease A minimum of 80 children were in the 4 months prior to study entry. Ex- of more than 2 log10 but whose RNA planned to be enrolled in each of the 3 clusion criteria included current grade remained at more than 10 000 cop- arms of this study. This sample size 3 or 4 adverse effect (as judged by pro- ies/mL at study week 12, or as a per- would ensure a power of 80% to de- tocol-specified, standard pediatric tox- sistent increase in HIV RNA to more tect a difference of one third in the av- icity criteria); active opportunistic and/or than 10 000 copies/mL for patients who erage change in log10 RNA copy num- serious bacterial infection; docu- had an initial HIV RNA decrease to less ber between 2 therapy arms using a mented hypersensitivity to any of the than 10 000 copies/mL. 2-sided P = .05 level of significance. The therapies under study; prior protease in- Two hundred ninety-eight HIV-infec- primary end point of the study was ini- hibitor therapy; or current diagnosis of ted children from 48 sites were enrolled tially specified to be the change in HIV malignancy or pregnancy. in the study. The institutional review RNA from baseline. However, since the Children were stratified by CD4 cell board at each institution approved the majority of patients in the ritonavir arms percentage (Ͻ25% vs Ն25% of the nor- study, and informed consent was ob- were observed to have HIV RNA copy mal range) and randomized in a bal- tained from all patients or their parents. numbers below the level of assay quan- anced fashion to 1 of 3 open-label treat- tification (Ͻ400 copies/mL) at the week ment arms. Medication doses dispensed Study Evaluations 12 interim analysis, the primary end were 160 mg/m2 of zidovudine (maxi- Evaluations were performed within 14 point for the study was changed to the mal dose, 200 mg/dose) 3 times per day days prior to randomization (the preen- proportion of children with values be- and 4 mg/kg of lamivudine (maximal try visit), at study entry, and every 4 low the level of assay quantification, dose, 150 mg/dose) twice per day; that weeks during the study. These visits in- consistent with other, similar studies.

Downloaded From: https://jamanetwork.com/ on 09/29/2021 RITONAVIR IN HIV-INFECTED CHILDREN

Comparisons among treatment groups ment, a child was classified as having Figure 1. Profile of Patient Enrollment and Follow-up used the Fisher exact test for categorical an undetectable level of RNA at a spe- variables and Wilcoxon/Kruskal-Wallis cific study week only if the child was test for continuous variables.11,12 An RNA still receiving initial randomized study 298 Eligible Patients assay result was categorized as below the treatment at that time. The number of level of quantification (undetectable) if patients on the treatment arm at base- 298 Randomized the VQA-adjusted result was less than line serves as the denominator for this 400 copies/mL. Comparisons of the rates conservative definition of response rate.

100 Received 97 Received 100 Received of undetectable RNA by baseline RNA Periodic interim monitoring of this Zidovudine Stavudine Zidovudine, values used the Fisher exact test for study by an independent study moni- and and Lamivudine, 13 Lamivudine Ritonavir and Ritonavir ordered categorical data. The baseline toring committee was planned at weeks 0 Did Not 1 Did Not 0 Did Not RNA value was defined to be the geo- 12 and 24 using protocol-specified cri- Receive Receive Receive Intervention Intervention Intervention metric mean of the preentry and the entry teria for the stopping or modification values. Patients with undetectable base- of a treatment arm. The criterion for

Evaluated for Evaluated for Evaluated for line RNA values could not contribute to stopping a specific treatment at the RNA Response RNA Response RNA Response the determination of the change from week 12 interim analysis was the de- 95 at Week 24 92 at Week 24 93 at Week 24 detectable to undetectable RNA and so tection of a statistically and medically NA at Week 48 92 at Week 48 93 at Week 48 were excluded from that analysis. significant difference in virologic effi-

Evaluated for Evaluated for Evaluated for Some patients experienced viro- cacy between any 2 treatment arms; a CD4 Cell Count CD4 Cell Count CD4 Cell Count logic failure and discontinued their ran- virologically inferior arm was defined 98 at Week 24 87 at Week 24 86 at Week 24 domized study treatment. These pa- as a one in which the average change NA at Week 48 80 at Week 48 90 at Week 48 tients remained on study follow-up but in HIV RNA copy number was at least received nonprotocol alternative anti- 1.5 log10 less than that in another arm Off Assigned Off Assigned Off Assigned Treatment Treatment Treatment retroviral treatment that could affect or in which the proportion of children 8 at Week 24 21 at Week 24 23 at Week 24 their subsequent virologic evaluation who achieved HIV RNA levels below 62 at Week 48 34 at Week 48 28 at Week 48 during follow-up. To avoid confound- quantification was at least 50% lower ing the study-related evaluations by than that in another arm. All analyses NA indicates not applicable. such nonstudy antiretroviral treat- were based on an intent-to-treat approach.14 All P values were 2-sided, were not adjusted for multiple compari- Table 1. Baseline Patient Characteristics by Treatment Group* sons, and were not adjusted for in- Zidovudine/ terim analyses. Zidovudine/ Lamivudine/ Stavudine/ Lamivudine Ritonavir Ritonavir P Variable (n = 100) (n = 100) (n = 97) Value RESULTS Median age, y 6.7 7.4 7.2 .91 Study Population Male sex, % 48 43 51 .32 Two hundred ninety-eight children en- Race or ethnicity, % tered the study between February 6 and White, non-Hispanic 20 12 14 April 30, 1997. One child was not in- Black, non-Hispanic 42 53 49 .58 cluded in the analysis because the child Hispanic 38 33 32 never started study therapy due to an Prior antiretroviral therapy, % Monotherapy 44 43 36 illness (FIGURE 1). Baseline patient Zidovudine/didanosine 44 43 43 .32 characteristics (TABLE 1) were well bal- Other combinations 12 13 21 anced among the treatment groups. The CD4 percentage, Ͻ25% of 38 33 36 .66 11 children with prior stavudine expe- the normal range, % rience who entered the study were CD4 cell count equally distributed among the treat- Median, ϫ106/L 680 644 693 ment arms. The median duration of fol- Ͻ500 ϫ 106/L, % 22 28 27 .45 low-up was 12.7 months, varying be- Ն1000 ϫ 106/L, % 23 20 19 tween 12.4 and 12.8 months for the 3 HIV RNA copy number, log10 Median 4.40 4.41 4.26 treatment groups. 2.6-3.0, % 7 13 11 3.0-4.0, % 29 18 29 HIV RNA Response to Treatment .63 4.0-5.0, % 42 49 41 The week 12 interim analysis of the first 5.0-6.0, % 22 19 19 136 children enrolled was conducted on *HIV indicates human immunodeficiency virus. Percentages may not add to 100% due to rounding. August 11, 1997. This analysis showed

Downloaded From: https://jamanetwork.com/ on 09/29/2021 RITONAVIR IN HIV-INFECTED CHILDREN

Table 2. Proportion of Children With Undetectable HIV RNA Levels at Baseline Receiving Original Randomized Treatment, by Group* No. (%) of Children

HIV RNA Ͻ400 Copies/mL HIV RNA Ͻ10 000 Copies/mL

Zidovudine/ Zidovudine/ Stavudine/ Zidovudine/ Zidovudine/ Stavudine/ Lamivudine Lamivudine/Ritonavir Ritonavir Lamivudine Lamivudine/Ritonavir Ritonavir Week (n = 95) (n = 93) (n = 92) (n = 95) (n = 93) (n = 92) 0 31 (33) 24 (26) 34 (37) 4 24 (25) 48 (52)† 46 (50)† 52 (55) 71 (76)† 68 (74)† 12 11 (12) 50 (54)† 48 (52)† 40 (42) 56 (60)† 62 (67)† 24 8 (8) 44 (47)† 31 (34)† 32 (34) 56 (60)† 49 (53)† 36 36 (39) 34 (37) 49 (53) 46 (50) 48 39 (42)‡ 25 (27)‡ 54 (58) 44 (48) *HIV indicates human immunodeficiency virus. Blank spaces indicate end of treatment. †PϽ.01 in comparison with zidovudine-lamivudine group. ‡P = .04 for triple therapy vs stavudine/ritonavir.

that the proportion of children whose Table 3. Proportion of Children Receiving Original Randomized Treatment With HIV RNA copy number reached an un- Undetectable HIV RNA Levels Categorized by Baseline HIV RNA* Ͻ detectable level ( 400 copies/mL) in the No. (%) of Children zidovudine and lamivudine group (14% Ritonavir-Containing Arms Zidovudine/ [6/43]) was significantly less than in the Baseline Lamivudine 2 ritonavir-containing treatment groups HIV RNA, Log10 Week 24 Week 48 Week 24 (61% [28/46] in the triple-therapy group 2.6-3.0 9/13 (69) 9/13 (69) 2/2 (100) and 57% [27/47] in the dual-therapy 3.0-4.0 21/46 (46) 20/46 (44) 5/29 (17) ritonavir group; PϽ.001 for both 4.0-5.0 36/89 (40) 28/89 (32) 1/42 (2) comparisons). Based on this interim 5.0-6.0 9/37 (24) 7/37 (19) 0/22 (0) analysis, children in the zidovudine and *In testing for ordered categorical data in the ritonavir-containing arms at week 24, P = .03; in week 48, PϽ.01; and in Ͻ lamivudine group with HIV RNA lev- the zidovudine/lamivudine arm, P .01. HIV indicates human immunodeficiency virus. els greater than 10 000 copies/mL were offered combination ritonavir, nevirap- RNA below the level of assay quantifi- We analyzed whether baseline HIV ine, and stavudine treatment identified cation was 8% (8 of 95) for the zidovu- RNA copy number, CD4 cell count, as the step 2 phase of the study; chil- dine and lamivudine group, 34% (31 of or age were predictors of long-term dren who had HIV RNA levels of up to 92) for the stavudine and ritonavir group, virologic success in the ritonavir- 10 000 copies/mL continued taking their and 47% (44 of 93) for the triple- containing treatment groups. In those original, randomized zidovudine and la- therapy group. The pairwise differ- children continuing their initial treat- mivudine treatment. Forty-eight chil- ences between the zidovudine and lam- ment at study week 48, only lower base- dren enrolled in step 2; 25% (12/48) en- ivudine group and the other 2 treatment line HIV RNA copy number was asso- rolled in step 2 between study weeks 30 groups at study weeks 4, 12, and 24 were ciated with a greater proportion of and 36 of their initial randomization, and statistically significant (PϽ.001). For children achieving HIV RNA levels be- the remainder enrolled after study week children in the 2 ritonavir-containing low quantification (TABLE 3). Sixty- 36. Therefore, analyses of the zidovu- study arms, the differences between these nine percent of children with baseline dine and lamivudine group patients in 2 groups at study weeks 4, 12, 24, and HIV RNA between 2.6 and 3 log10 cop- this article are only included through 36 were not statistically significant. How- ies/mL achieved RNA levels below study week 24 (when all patients were ever, a significant difference was ob- quantification at study week 48 com- still receiving their original, random- served at study week 48. The propor- pared with only 19% of children with ized treatment). tion of children receiving initial treatment baseline HIV RNA between 5 and 6 log10 The overall week 12 analysis results are at study week 48 with HIV RNA below copies/mL. There was no statistically presented in TABLE 2. At study week 12, the level of quantification was 27% (25 significant association of baseline CD4 12% of the zidovudine plus lamivudine of 92) for the stavudine and ritonavir cell count or age at study entry with vi- group had undetectable plasma HIV RNA group and 42% (39 of 93) for the triple- rologic response (data not shown). compared with 52% and 54% of those in therapy group (P = .04). At the same the 2- and 3-drug ritonavir-containing time, the proportion of children with HIV CD4 Cell Response to Treatment arms, respectively (PϽ.001). The pro- RNA less than 10 000 copies/mL was There was no difference in the median portion of children (taking initial treat- similar in both ritonavir-containing treat- CD4 cell count for all 3 treatment ment) at study week 24 who had HIV ment groups (P = .19). groups combined between study en-

Downloaded From: https://jamanetwork.com/ on 09/29/2021 RITONAVIR IN HIV-INFECTED CHILDREN

try and week 24 (671 vs 744 ϫ 106/L) initial therapy had been discontinued RNA copy number after study week 12 (FIGURE 2). In the ritonavir-contain- prematurely for 62% of patients in the was not common in the zidovudine and ing study arms, median CD4 cell counts zidovudine and lamivudine group, 35% lamivudine group until study week 36, at week 48 for the children receiving in the stavudine and ritonavir group, at which time it became the predomi- stavudine and ritonavir and those re- and 28% in the triple-therapy group nant reason. ceiving triple therapy were 767 and (FIGURE 3). While intolerance was the 818 ϫ 106/L, respectively (P = .47). most frequent reason for discontinua- COMMENT However, there was a significant dif- tion of initial ritonavir therapy by study This study was the first large, random- ference in CD4 percentage between the week 4, by study week 48 the rates of ized clinical trial to evaluate the use of 2 ritonavir-containing treatment groups discontinuation due to intolerance were ritonavir in HIV-infected children. Of the at study week 48, with medians of 29% similar for the 3 treatment groups. 297 evaluable children who entered the vs 33% for the 2-drug vs 3-drug com- Thirty-six percent of the 197 children study, 197 received ritonavir-contain- binations (PϽ.01). in the ritonavir-containing arms were ing antiretroviral treatment regimens. Af- receiving full-dose therapy, and 31% ter 12 weeks of therapy, a significant and Adverse Events were no longer receiving ritonavir at sustained decrease in HIV RNA copy Overall, 72% of the children experi- study week 48. Discontinuation of ini- number was observed in children enced a moderate (grade 2) or worse tial treatment due to an increase in HIV changed from their previous nucleo- toxic event while receiving initial therapy, and 21% experienced a severe Figure 2. Median CD4 Cell Count by Treatment Group (N = 297) (grade 3) or worse toxic event. How- ever, there were no significant differ- 1400 Zidovudine + Lamivudine

ences between the treatment groups with /L Stavudine + Ritonavir 6 1200 Zidovudine + Lamivudine + Ritonavir

respect to the overall rates of toxicity. The 10 × overall rate of severe or worse toxic ef- 1000 fects was 22% for those receiving zidovudine and lamivudine; 23% for the 800

stavudine and ritonavir group; and 17% 600 for the triple-therapy group. The most commonly observed adverse events were 400 nausea or vomiting (24%) (grade 2 or Median CD4 Cell Count, 200 higher: Ͼ3 episodes of vomiting per day, duration of Ͼ3 days, nausea resulting in 0 4 8 12 24 36 48 decreased oral intake); rash (19%) (grade Weeks in Study

2 or higher: diffuse maculopapular rash, No. of Children dry desquamation, or worse); fever of Zidovudine + Lamivudine 100 100 100 100 100 38.5°C or more (18%); and neutrope- Stavudine + Ritonavir 97 97 96 96 96 96 96 Zidovudine + Lamivudine + Ritonavir 100 100 100 100 100 100 100 nia (absolute neutrophil count Ͻ750 ϫ 6 10 /L) (18%). Moderate or worse nau- Bars give interquartile range. P values for stavudine plus ritonavir vs zidovudine plus lamivudine plus ritonavir sea or vomiting was more frequent in were .45 at baseline, .84 at 4 weeks, .86 at 8 weeks, .24 at 12 weeks, .95 at 24 weeks, .98 at 36 weeks, and .47 at 48 weeks. children receiving triple therapy, compared with those receiving zidovudine Table 4. Common Moderate or Severe Adverse Effects by Treatment Group and lamivudine alone (TABLE 4). Neu- tropenia of a moderate or worse degree % of Patients was less frequent in children receiving Zidovudine/ Zidovudine/ Stavudine/ stavudine and ritonavir than either of the Lamivudine Lamivudine/Ritonavir Ritonavir Toxic Effect (n = 100) (n = 100) (n = 97) study arms that contained zidovudine. Nausea or vomiting 16* 31 26 Rash 22 17 19 Discontinuation of Study Drugs Fever 21 17 16 Three reasons that children needed per- Neutropenia 22† 22† 8 manent discontinuation of their ini- Gastrointestinal 10‡ 12 21 tial randomized study treatments were: Hepatic 10 16 17 (1) HIV RNA greater than 10 000 cop- Anemia 22† 12 5 ies/mL, (2) adverse events or intoler- *P = .02 in comparison with the zidovudine/lamivudine/ritonavir group. ance, or (3) other nontoxicity or viro- †PϽ.01 in comparison with the stavudine/ritonavir group. ‡PϽ.05 in comparison with the stavudine/ritonavir group. logic failure reasons. By study week 48,

Downloaded From: https://jamanetwork.com/ on 09/29/2021 RITONAVIR IN HIV-INFECTED CHILDREN

side analog antiretroviral therapy to a the effect of therapy on clinical out- There were no significant differ- ritonavir-containing regimen. Fifty- comes, as few outcome events had oc- ences in absolute CD4 cell counts after three percent of children who received curred by 48 weeks. The effect of treat- 48 weeks of therapy between the 2 rito- a ritonavir-containing regimen had HIV ment must therefore be evaluated over navir-containing regimens; however, a RNA levels below the limit of assay quan- extended periods before definitive con- greater increase in CD4 percentage was tification at study week 12 compared clusions can be drawn regarding any seen in children receiving triple therapy. with only 12% of children who had been comparisons of highly active antiretro- This discrepancy between CD4 cell randomized from their previous regi- viral therapies. There were no cases of count and percentage suggests that CD4 men to the zidovudine and lamivudine Pneumocystis carinii pneumonia re- percentage may be a more sensitive or study arm (PϽ.001). Thus, changing ported during the study period. One case accurate measure of immune response from a single or dual nucleoside regi- of cytomegalovirus disease was re- to antiretroviral therapy than absolute men to another dual nucleoside regi- ported in the stavudine plus ritonavir CD4 cell count when nucleoside ana- men had little virologic benefit for clini- treatment group and none in any of the log therapy suppresses white blood cell cally stable pediatric patients. As a other treatment groups. counts over time. Another potential rea- consequence, children randomized to the Several reasons may have contrib- son for the lack of a significant change zidovudine and lamivudine study arm uted to the relatively low success rate. in CD4 cell number between treatment who had HIV RNA levels of more than First, children tend to have higher regimens is that many children entered 10 000 copies/mL after study week 12 plasma HIV RNA levels than adults,15 the study with absolute CD4 cell counts were changed to a protease-containing and the higher the viral load, the lower that were already normal for age (the regimen (stavudine, nevirapine, and rito- the success rate appears. Second, al- median age at entry was approximately navir) in the step 2 phase of this study though more than two thirds of the chil- 7 years, and the median entry CD4 cell and are still undergoing follow-up. dren were still taking their initially as- count in the study arms was 644 to After 48 weeks of therapy, 42% of the signed ritonavir-containing treatment 693 ϫ 106/L; normal for this age group children receiving triple therapy main- at study week 48, only half of them were is Ͼ500 ϫ 106/L). Additional studies are tained HIV RNA levels below the limit still receiving the full dose of ritona- needed to investigate the importance of of assay quantification, compared with vir, and resistant virus may have been more specialized phenotypes within the only 27% of children receiving dual present. Third, all these children were larger CD4 cell group. therapy with ritonavir and stavudine. antiretroviral-treatment experienced Ritonavir was relatively well toler- This difference was not noted at study and often only changed 2 drugs, not 3. ated by the children in this study. Sev- weeks 24 and 36. In addition, there was Fourth, adherence to multiple-drug enty percent continued taking ritonavir no significant difference in the propor- regimens is difficult in children. These for the 48 weeks of the study. While nau- tion of children with HIV RNA viral factors may have contributed to the sea and vomiting appeared to be the most number below 10 000 copies/mL be- overall lower success rates when com- common adverse effect in the ritonavir tween the 2 ritonavir-containing treat- pared with adult antiretroviral trials. arms, taste was the biggest obstacle for ment groups at study week 48. These data suggest that dual therapy includ- Figure 3. Major Reasons for Permanent Cessation of Initial Randomized Treatment ing ritonavir has comparable activity to 70 triple therapy including ritonavir in terms Treatment of moderate reduction in HIV RNA copy 1. Zidovudine + Lamivudine (n =100) Ͻ 60 2. Stavudine + Ritonavir (n = 97) numbers (to levels 10 000 copies/ 3. Zidovudine + Lamivudine + Ritonavir (n =100) mL) but that triple therapy with ritona- 50 vir was superior to dual therapy in main- HIV RNA ≥10 000 Copies/mL taining HIV RNA copy numbers below Intolerance 40 the level of assay quantitation. Al- Other

though initial evaluation at week 24 sug- 30 gested that the 2-drug combination of stavudine and ritonavir was as effective 20 as zidovudine, lamivudine, and ritona- Cumulative Discontinuation Rate, % vir in effecting an undetectable viral load, 10 this outcome did not appear to be fully sustained when a longer interval (48 0 123 123 123 123 123 weeks) was examined. It is important to Week 4 Week 12 Week 24 Week 36 Week 48 note that even triple therapy was less than 50% effective in maximally suppressing HIV indicates human immunodeficiency virus; intolerance, patient was unable to tolerate antiretroviral medication; and other, family and home issues. the viral load. It is difficult to evaluate

Downloaded From: https://jamanetwork.com/ on 09/29/2021 RITONAVIR IN HIV-INFECTED CHILDREN

the health care providers. To increase the in antiretroviral therapy in children McIntosh), Boston, Mass; Division of Infectious Dis- eases, Children’s Memorial Hospital, Chicago, Ill (Dr Yo- palatability of the ritonavir, coating the should be considered when HIV RNA gev); Pediatric Adolescent and Maternal AIDS Branch, mouth with peanut butter or with choco- levels are only moderately elevated, Center for Research for Mothers and Children, Na- tional Institute of Child Health and Human Develop- late or vanilla pudding or applying ice rather than waiting until HIV RNA lev- ment (Dr Mofenson) and Division of AIDS, National In- chips were recommended. In many cases, els become excessively high. stitute of Allergy and Infectious Diseases (Dr Smith), these methods or others devised at in- This study demonstrated that rito- National Institutes of Health, Bethesda, Md; Depart- ment of Microbiology and Immunology, School of Medi- dividual sites were effective in permit- navir-containing treatment regimens cine, University of North Carolina at Chapel Hill (Dr Fis- ting successful drug administration. have potent antiviral effects, and, there- cus); Department of Pediatrics, Health Science Center, University of Florida, Jacksonville (Dr Rathore); Depart- When possible, capsules were substi- fore, children who are nucleoside- ment of Pediatrics, San Juan City Hospital, San Juan, tuted for the liquid formulation. A few experienced should be switched to a Puerto Rico (Dr Jimenez); and Section of Pediatric In- children able to tolerate ritonavir early protease inhibitor–containing treat- fectious Disease, Social and Scientific Systems, Rock- ville, Md (Ms Wells). in the study were unable to tolerate it ment regimen to successfully decrease Additional Members of the Pediatric AIDS Clinical Tri- weeks or months later. Other toxic events their viral load. To extend the durabil- als Group 338 Study Team: Anita Ballow, Frontier Sci- ence and Technology Research Foundation, Amherst, NY; noted in the study included fever and ity of the viral load suppression, 2 Courtney Fletcher, Pharm D, University of Minnesota, skin rash, which appeared at approxi- nucleosides instead of 1 should be part Minneapolis; Molly Nozyce, PhD, Bronx Lebanon Hos- pital Center, Bronx, NY; Stephen Spector, MD, Univer- mately the same rates in all 3 treatment of the combination. Ritonavir was sity of California, San Diego; Steve Douglas, MD, Chil- groups, and neutropenia, which ap- generally well tolerated and associ- dren’s Hospital of Philadelphia, Philadelphia, Pa; Lynette peared more commonly in the zidovu- ated with a toxicity profile commonly Perdue, PharmD, Division of AIDS, National Institute of Allergy and Infectious Diseases, Bethesda, Md; Bethany dine and lamivudine treatment arms. seen with protease inhibitor therapy. Griffin, MA, Social Scientific Systems, Rockville, Md; Ross There were no differences overall in the While there appeared to be some late- McKinney, MD, Duke University Medical Center, Durham, NC; Jeanne Berg, MS, and Rebecca Hoffman, MD, Ab- rate of grade 3 or 4 toxic events across onset problems with intolerance, 70% bott Laboratories, Abbott Park, Ill; Maria Gigliotti, MS, all treatment groups. of children continued their ritonavir- Patrick Robinson, MD, Laurie Reynolds, and Vicki Rut- kiewicz, MS, Boehringer-Ingleheim Pharmaceuticals, Baseline HIV RNA copy number was containing treatment through study Ridgefield, Conn; Barbara Lane, RN, Glaxo-Wellcome, an important prognostic factor in viro- week 48. Protease inhibitor–contain- Research Triangle Park, NC; Kathleen Mohan, ARNP, Chil- logic response to therapy. Sixty-nine per- ing combination therapy should be dren’s Hospital and Medical Center, Seattle, Wash; Moraima Rivera, MPH, San Juan City Hospital, San Juan, cent of children who entered the study viewed as part of the standard therapy Puerto Rico; and Virginia Parks, San Francisco, Calif. with HIV RNA levels under 1000 cop- for children with HIV disease. Financial Disclosures: Drs Nachman, Yogev, Pel- ton, Wiznia, Fiscus, and Rathore have served as ad hoc ies/mL achieved HIV RNA levels below Author Affiliations: Departments of Pediatrics, State consultants or as speakers in programs sponsored by the limit of assay quantitation by study University of New York at Stony Brook (Dr Nachman), Abbott Laboratories, Glaxo-Wellcome, or Bristol- week 48, compared with only 19% of Jacobi Medical Center, Albert Einstein College of Medi- Myers Squibb, the pharmaceutical firms whose prod- cine (Dr Wiznia), New York University Medical Center ucts were studied. children who entered with baseline HIV (Dr Borkowsky), and Department of Pediatric Infec- Funding/Support: This study was supported in part by RNA levels greater than 100 000 copies/ tious Disease, College of Physicians and Surgeons, Co- the Pediatric AIDS Clinical Trials Group of the National lumbia University (Dr Pitt), New York, NY; Center for Institute of Allergy and Infectious Diseases, the Pediatric/ mL. A similar finding has been re- Biostatistics in AIDS Research, School of Public Health, Perinatal HIV Clinical Trials Network of the National In- ported in HIV-infected adults16 but was Harvard University (Dr Stanley and Ms Lee), Section stitute of Child Health and Human Development, Ab- of Pediatric Infectious Diseases, Boston Medical Cen- bott Laboratories, Glaxo-Wellcome, and Bristol-Myers not observed in a smaller study in chil- ter (Dr Pelton), and Department of Medicine, Division Squibb. This study also received support from National dren.17 These data suggest that change of Infectious Disease, Children’s Hospital of Boston (Dr Institutes of Health grant AI-41110.

REFERENCES 1. Englund JA, Baker CJ, Raskino C, et al. Zidovu- ference on Antimicrobial Agents and Chemotherapy; man immunodeficiency virus type 1 RNA in plasma. dine, didanosine, or both as the initial treatment for September 28-October 1, 1997; Toronto, Ontario. Ab- J Clin Microbiol. 1998;36:311-314. symptomatic HIV-infected children. N Engl J Med. stract LB-6. 11. Hollander M, Wolfe DA. Nonparametric Statis- 1997;336:1704-1712. 6. Centers for Disease Control and Prevention. Re- tical Methods. New York, NY: John Wiley; 1973. 2. McKinney RE Jr, Johnson GM, Stanley K, et al. A vised classification system for HIV infection in chil- 12. Lehmann E. Nonparametrics: Statistical Meth- randomized study of combined zidovudine- dren less than 13 years of age. MMWR Morb Mortal ods Based on Ranks. San Francisco, Calif: Holden- lamivudine versus didanosine monotherapy in chil- Wkly Rep. 1994;43(RR-12):1-10. Day; 1975. dren with symptomatic therapy-naive HIV-1 infec- 7. Gelman R, Cheng SC, Kidd P, Waxdal M, Kagan 13. Mehta CR, Patel NR, Tsiatis AA. Exact significance tion. J Pediatr. 1998;133:500-508. J. Assessment of the effects of instrumentation, mono- testing to establish treatment equivalence with ordered 3. Centers for Disease Control and Prevention. Guide- clonal antibody, and fluorochrome on flow cytomet- categorical data. Biometrics. 1984;40:819-825. lines for the use of antiretroviral agents in pediatric ric immunophenotyping. Clin Immunol Immuno- 14. Pocock SJ. Clinical Trials: A Practical Approach. HIV infection. MMWR Morb Mortal Wkly Rep. 1998; pathol. 1993;66:150-162. New York, NY: John Wiley; 1983. 47(RR-4):1-43. 8. Dyer JR, Pilcher CD, Shepard R, Schock J, Eron JJ, 15. Palumbo PE, Kwok S, Waters S, et al. Viral mea- 4. Department of Health and Human Services and Fiscus SA. Comparison of NucliSens and Roche Moni- surement by polymerase chain reaction–based as- Henry J. Kaiser Family Foundation. Guidelines for the tor assays for quantitation of levels of human immu- says in human immunodeficiency virus–infected in- use of antiretroviral agents in HIV-infected adults and nodeficiency virus type 1 RNA in plasma. J Clin Mi- fants. J Pediatr. 1995;126:592-595. adolescents [published correction appears in: MMWR crobiol. 1999;37:447-449. 16. Notermans DW, Goudsmit J, Danner SA, de Wolf Morb Mortal Wkly Rep. 1998;47:619]. MMWR Morb 9. Yen-Lieberman B, Brambilla D, Jackson B, et al. F, Perelson AS, Mittler J. Rate of HIV-1 decline follow- Mortal Wkly Rep. 1998;47(RR-5):43-82. Evaluation of a quality assurance program for quan- ing antiretroviral therapy is related to viral load at base- 5. Yogev R, Stanley K, Nachman S, et al. Virologic ef- titation of human immunodeficiency virus type 1 RNA line and drug regimen. AIDS. 1998;12:1483-1490. ficacy of ZDV + 3TC vs d4T + Ritonavir vs in plasma by the AIDS Clinical Trials Group virology 17. Mueller BU, Nelson RP Jr, Sleasman J, et al. A phase ZDV + 3TC + Ritonavir in stable antiretroviral expe- laboratories. J Clin Microbiol. 1996;34:2695-2701. I/II study of the protease inhibitor ritonavir in chil- rienced HIV-infected children (Pediatric ACTG trial 10. Brambilla D, Leung S, Lew J, et al. Absolute copy dren with human immunodeficiency virus infection. 338). In: Proceedings of the 37th Interscience Con- number and relative change in determinations of hu- Pediatrics. 1998;101(3 pt 1):335-343.

Downloaded From: https://jamanetwork.com/ on 09/29/2021