A Phase I/II Study of the Protease Inhibitor Indinavir in Children With HIV Infection
Background. Indinavir, an inhibitor of the human immunodeficiency virus type 1 (HIV-1) protease, is approved for the treatment of HIV infection in adults when antiretroviral therapy is indicated. We evaluated the safety and pharmacokinetic profile of the indinavir free-base liquid suspension and the sulfate salt dry-filled capsules in HIV-infected children, and studied its preliminary antiviral and clinical activity in this patient population. In addition, we evaluated the pharmacokinetic profile of a jet-milled suspension after a single dose.
Methods. Previously untreated children or patients with progressive HIV disease despite antiretroviral therapy or with treatment-associated toxicity were eligible for this phase I/II study. Three dose levels (250 mg/m2, 350 mg/m2, and 500 mg/m2 per dose given orally every 8 h) were evaluated in 2 age groups (<12 years and ≥12 years). Indinavir was initially administered as monotherapy and then in combination with zidovudine and lamivudine after 16 weeks.
Results. Fifty-four HIV-infected children (ages 3.1 to 18.9 years) were enrolled. The indinavir free-base suspension was less bioavailable than the dry-filled capsule formulation, and therapy was changed to capsules in all children. Hematuria was the most common side effect, occurring in 7 (13%) children, and associated with nephrolithiasis in 1 patient. The combination of indinavir, lamivudine, and zidovudine was well tolerated. The median CD4 cell count increased after 2 weeks of indinavir monotherapy by 64 cells/mm3, and this was sustained at all dose levels. Plasma ribonucleic acid levels decreased rapidly in a dose-dependent way, but increased toward baseline after a few weeks of indinavir monotherapy.
Conclusions. Indinavir dry-filled capsules are relatively well tolerated by children with HIV infection, although hematuria occurs at higher doses. Future studies need to evaluate the efficacy of indinavir when combined de novo with zidovudine and lamivudine.
- AIDS =
- acquired immunodeficiency syndrome •
- CDC =
- Centers for Disease Control and Prevention •
- HIV =
- human immunodeficiency virus •
- RNA =
- ribonucleic acid •
- NCI =
- National Cancer Institute •
- AUC =
- area under the time-concentration curve •
- FSIQ =
- full scale intelligence quotient •
- VIQ =
- verbal intelligence quotient •
- PIQ =
- performance intelligence quotient •
- PPD =
- purified protein derivative
As of December 1997, >7900 children with an acquired immunodeficiency syndrome (AIDS)-defining illness have been reported to the Centers for Disease Control and Prevention (CDC) in the United States.1 An estimated 10 million children worldwide will be infected with human immunodeficiency virus (HIV)-1 by the year 2000. As of May 1998, four reverse transcriptase inhibitors (zidovudine, didanosine, lamivudine, and stavudine) and two protease inhibitors (ritonavir and nelfinavir) have been approved for use in HIV-infected children.
Indinavir is a synthetic protease inhibitor that competitively inhibits the HIV-1 aspartyl protease that cleaves the Gag and Pol polyproteins into their functional components. In cultured cells infected with a variety of laboratory strains and clinical isolates of HIV-1, indinavir results in 95% inhibition of viral replication at concentrations 25 to 100 nM. Marked synergism was observed in vitro when indinavir was combined with zidovudine, didanosine, or a nonnucleoside reverse transcriptase inhibitor.
In adults, an indirect hyperbilirubinemia without concurrent increase in hepatic transaminases, and hematuria, sometimes associated with nephrolithiasis, are two adverse events that have been clearly associated with the administration of indinavir. At a dose level of 2.4 g daily of indinavir sulfate, divided into four doses, a decrease in serum p24 antigen and viral ribonucleic acid (RNA) (>1 log10) and an increase in CD4 counts (>50 cells/mm3) was observed in a small study of 5 adult patients.2 The antiviral effects were most pronounced within the first few weeks of treatment. After 12 weeks of monotherapy at daily doses of 1.6 g to 2.4 g, the majority of patients sustained 0.5 to 1 log10 declines in viral RNA levels; however, after 24 to 48 weeks of monotherapy, viral RNA levels generally returned to near baseline levels in most patients. The return of viral RNA toward baseline correlates with the appearance of drug-resistant viral isolates. Recently published data regarding the combination therapy with indinavir, zidovudine, and lamivudine in adults demonstrated a decrease of plasma RNA levels to undetectable levels (<500 copies/mL) in 28 of 31 patients after 24 weeks of therapy, and a markedly slower rate of progression to AIDS.3 4
It is important to develop safer and more potent antiretroviral agents for children. Unfortunately, it has been difficult to formulate a solution or suspension of some of the protease inhibitors because of their poor solubility and unpleasant taste. Here we report the evaluation of a liquid formulation of indinavir (free-base suspension) in children with HIV infection as well as experience with the indinavir capsule formulation (sulfate salt dry-filled capsules).
Children between the ages of 6 months and 18 years were eligible for enrollment. There were no restrictions with regard to gender, ethnicity, or route of HIV acquisition. Children who had become refractory to prior antiretroviral therapy or who had experienced toxicity to prior therapy were eligible for the study, as well as previously untreated, asymptomatic children whose age-corrected absolute CD4 count rendered them at risk for an AIDS-related opportunistic infection (immunologic categories 2 and 3 of the CDC classification).5 Children with moderate to severe symptomatic HIV infection (CDC class B and C) were also eligible. Progressive disease on prior therapy was defined as evidence of neurologic deterioration, progressive weight loss, or failure to thrive; new or recurrent serious HIV-related infections (as listed under CDC category C); a decline in CD4 cell count (absolute count or percentage) by one third or more, confirmed on at least two measurements 4 weeks apart; or a decline in the CD4 percentage to <15% (the critical value for an increased risk for opportunistic infections) sustained throughout two measurements 4 weeks apart.
Patients were required to have a total white blood count >1500 cells/mm3, a neutrophil plus band count >750 cells/mm3, hemoglobin >8.0 g/dL, platelet count >50 000/mm3, a serum creatinine <2 times the upper limit of normal for age, hepatic transaminases <3 times the upper limit of normal, and total bilirubin <1.5 mg/dL. Patients with hematuria or a history of kidney stones were excluded, as were patients with severe recurrent or persistent diarrhea because of the increased risk of dehydration. Patients were required to be in clinically stable condition, to have discontinued all antiretroviral therapy at least 14 days before study entry, and not to have required any new therapeutic interventions for acute infections within 7 days of entry.
The study was performed by the HIV and AIDS Malignancy Branch (previously part of the Pediatric Branch) of the National Cancer Institute (NCI, Bethesda, MD), in collaboration with the Department of Pediatrics of the University of Florida (Gainesville, FL), and the University of South Florida at All Children's Hospital (St. Petersburg, FL). The protocol was approved by the NCI's institutional review board and the institutional review boards of the two collaborating centers. Written informed consent was obtained from the parent or legal guardian of each child.
This study had a time-lagged, dose-escalation design. The initial study was designed to treat children at 3 different dose levels of indinavir suspension (250 mg/m2, 350 mg/m2, and 500 mg/m2 per dose given every 8 h), and, after 12 weeks of monotherapy, to add zidovudine at a dose of 120 mg/m2per dose given every 8 hours and lamivudine at a dose of 4 mg/kg given twice daily. However, pharmacokinetic data from the first two dose levels demonstrated that the suspension was poorly absorbed. We therefore modified the protocol to permit those children able to swallow whole capsules to be switched at week 12 to a dose of 250 mg/m2 of indinavir capsules, rounded down to the nearest multiple of 100 mg, and to add the dideoxynucleosides after a total of 16 weeks of monotherapy. We then proceeded to enroll additional children to receive 16 weeks of indinavir monotherapy in the capsule formulation at 250 mg/m2, 350 mg/m2, and 500 mg/m2, followed by the addition of zidovudine and lamivudine. In September 1996, all children treated at 500 mg/m2 had their dose decreased to 350 mg/m2after 16 weeks of monotherapy (see below). As a result, only 5 children have data available for the combination phase including zidovudine, lamivudine, and the highest dose level of indinavir (500 mg/m2 per dose).
Patients were stratified for enrollment into two age groups (<12 years and ≥12 years), and dose escalations were performed independently in these two groups. Six children, who had previously been enrolled in a 12-week trial of the protease inhibitor KNI-272, were entered in addition to the other patients but escalated together with the whole group.
Indinavir was given every 8 hours orally on an empty stomach. Because inadequate hydration may be a risk factor for kidney stone formation, patients were encouraged to drink 4 fluid ounces or more of water after ingestion of indinavir.
Patients who experienced new onset of grade III toxicity or persistent (>1 month duration) grade II toxicity had their study medication discontinued until the laboratory abnormality returned to baseline or symptoms resolved. If the toxicity did not return to baseline or grade I within 1 month, the patient was removed from study. If the toxicity resolved within 1 month, indinavir was reintroduced at the next lower dose level. If the toxicity was not thought to be study-drug related, the patient's dose was escalated to the pretoxicity level at the investigator's and study sponsor's discretion. Recurrence of the same toxicity (at grade III or IV), despite dose reduction, required withdrawal of the patient from the study. Patients who experienced any grade IV toxicity had their study medication permanently discontinued.
Mild isolated indirect bilirubin elevations (<5 mg/dL), which would be classified as grade III or grade IV toxicities, are commonly observed in adults receiving indinavir and did not necessitate interruption of therapy in this study. Patients with signs or symptoms suggestive of indinavir nephrolithiasis, including new hematuria (>15 red blood cells per high power field) or pain apparently originating from the urinary tract, had indinavir interrupted until the signs and symptoms disappeared.
Clinical and Laboratory Monitoring
All eligible patients were initially evaluated in the outpatient clinic of the NCI. Vital signs were monitored during the first 72 hours of the study, and patients were hospitalized for 48 hours for the pharmacokinetic sampling on Days 1 and 2 (described below). Patients were evaluated every 2 weeks until week 8, and then at weeks 12, 16, 18, 20, 24, and 28. Skin tests for mycobacteria (purified protein derivative [PPD]; 5 tuberculin units [TU]) with Candida albicans (1:100) and tetanus toxoid (1:5) as controls were performed at entry, week 16, and week 28. Echocardiogram, electrocardiogram, chest radiograph, and a computer tomogram of the head were obtained at entry and at 28 weeks.
Detailed, age-appropriate neuropsychometric testing was performed at entry and repeated at weeks 16 and 28. Patients were evaluated with a comprehensive neuropsychologic battery at baseline and at 28 weeks that included a full standardized intelligence test (McCarthy or WISC-III) or the Bayley Revised. In addition a monitor assessment was administered at entry, 16 weeks, and 28 weeks that could include the Peabody Picture Vocabulary Test-Revised, the Gardner one-word expressive language test, the Beery VMI, the colored or standard Raven's, and the Digit Span subtest (of the McCarthy or WISC-III).
A complete blood count, coagulation profile, routine biochemistry values, urinalysis (including urine pregnancy test in postmenarchal females), fluorescence activated cell sorter analysis of lymphocyte subsets, and plasma HIV RNA quantitation by polymerase chain reaction assay (Amplicor, lower limit of detection 200 copies/mL)6were monitored throughout the study.
Pharmacokinetic evaluations were performed using the free-base suspension, the dry-filled capsule formulation, and a modification of the free-base formulation with smaller particle size, a jet-milled suspension. The latter was only administered for pharmacokinetic studies (as a single dose). Plasma samples for pharmacokinetic analysis were obtained through 8 hours after a dose at the following time points: before the dose and at 0.5, 1, 1.5, 2, 3, 4, 6, and 8 hours after the dose on days 1 and 2. The plasma concentrations of indinavir were measured using a previously reported high pressure liquid chromatography assay.7 The area under the concentration-time curve (AUC0–8) was calculated using the linear trapezoidal rule.
Criteria for Response to Treatment
The visits at weeks 16 and 28 were considered to be the main evaluation points. Clinical response was defined as a sustained weight gain, evidenced by an increase of 10% greater than baseline or upward crossing of percentile growth curve that was not associated with the initiation of intravenous hyperalimentation. Improvement in neurocognitive function was defined as >10% increase over baseline in age-appropriate full-scale IQ score and ≥8 IQ points, because changes of this magnitude are unlikely to be caused by practice effect.8 Evidence of improvement could also be determined by changes in motor function or improvement in neural imaging studies. Immunologic response was defined as an increase in absolute CD4 cell count of 10% of the baseline with a minimum increase of 50 CD4 cells on two consecutive measures at least 4 weeks apart. CD4 percentage response was defined as an increase of ≥25% in the percentage of CD4 cells on two or more successive measurements at least 4 weeks apart. A response in viral load was defined as a decrease of at least one log in virus particle number in patients with a measurable level at entry.
Baseline values among dosage groups and changes from baseline among dose levels were compared with the Kruskal-Wallis test. Comparison of results at week 16 and 28 to baseline was performed by Wilcoxon signed-rank test. All P values are two-sided.
Neuropsychologic changes between 2 or 3 time points were assessed with repeated measures analysis of variance with a type 1 α level of 0.05. Change from baseline was calculated with previously published methods.9 Fisher's exact test was used to evaluate possible differences in number of patients improving or declining in two groups.
Fifty-four HIV-infected children were enrolled between July 1995 and August 1996, and the evaluation of the first 28 weeks on protocol is included in this report. The baseline patient characteristics for all children are listed in Table 1. The patients were heavily pretreated with antiretroviral agents and only 3 children were previously untreated. The data of 6 children who had been enrolled in a 12-week study of KNI-272, another protease inhibitor, are included in the safety and pharmacokinetic analysis, but not in the efficacy evaluation. Fifty of the 51 previously treated children had received zidovudine in the past, 14 lamivudine, 44 didanosine, 10 zalcitabine, 6 stavudine, and 3 nevirapine. Twenty-three children each were classified as having clinical CDC category B or C disease, but 35 (65%) had low CD4 counts (immunologic category 3).10No statistically significant differences were observed in the baseline parameters for the children enrolled at different dose levels, except for the baseline HIV RNA level, which was 31 873 copies/mL for the combined suspension groups and 103 556 copies/mL for the combined capsule groups (P = .0049). As of January 16, 1997, 43 of the 54 children had completed 28 weeks of therapy, 7 children had therapy discontinued before week 28, and 4 patients had not yet reached 28 weeks on protocol. No deaths occurred during this study.
Safety and Tolerance
Seven children discontinued treatment, 2 because of progressive disease and 5 because of toxicity. One patient (250 mg/m2 capsules) had rapidly progressive neurologic deterioration (onset probably before initiation of therapy) and was therefore taken off the study after 4 weeks. He improved markedly on open label indinavir, zidovudine, and lamivudine. Another patient (at 250 mg/m2 capsules) developed a progressive loss of vision of unclear etiology after 21 weeks on study. This patient had very pale optical disks with no other features of optic neuritis, and a salt-and-pepper fundus, without positive serology for syphilis. Although most likely HIV-associated, the patient's therapy with the investigational agent was discontinued, without subsequent change in status.
Study medication was discontinued in 5 patients because of adverse experiences thought to be possibly drug-related. This included neutropenia in 1 patient (week 16, 250 mg/m2 capsules), increased hepatic transaminases (grade 3 toxicity) in a patient with known fluctuations in liver enzymes (week 12, 250 mg/m2suspension), and recurrent episodes of hematuria (350 mg/m2capsules, details below) in another patient. One patient discontinued study medication at the parent's request because of hyperactivity (week 26, 250 mg/m2 capsules), and another stopped therapy because of mild gastrointestinal complaints (week 16, 350 mg/m2 capsules).
The dose of indinavir was reduced for hematuria in 3 children, 1 each on 350 mg/m2 suspension, 350 mg/m2 capsules, and 500 mg/m2 capsules. The reduced dose was tolerated without recurrence of hematuria in all 3 patients. Zidovudine dose was reduced in 2 patients because of intermittent neutropenia. One patient (week 6, 350 mg/m2 suspension) who was receiving rifabutin for Mycobacterium avium complex prophylaxis, developed uveitis and keratitis which resolved with symptomatic treatment after stopping rifabutin, and 7 patients had a trend toward lower absolute neutrophil counts when they received rifabutin concurrently with indinavir.
Despite instructions for generous hydration with the administration of indinavir, hematuria was the most common toxicity occurring in 7 (13%) patients during the first 28 weeks on study (Table2). One patient passed a kidney stone and all 7 patients had unidentified crystals in their urine (presumed to be indinavir crystals). All episodes resolved with adequate hydration and without impairment of renal function. The presence of urinary crystals was not predictive for the development of hematuria, because most patients had unidentified crystals in their urinalysis. The stones were recovered and analyzed and found to be a mixture of organic and inorganic material (carbon, oxygen, sodium, zinc, silicon, phosphorus, calcium, and iron) with a trace of indinavir. When we tried later in the study (after 48 weeks) to increase the dose of indinavir capsules in all patients to 500 mg/m2, we observed a cluster of 4 more episodes of hematuria, including another patient with nephrolithiasis. Because of this, all patients treated at 500 mg/m2 dose level had their dose deescalated to 350 mg/m2 after initiation of combination therapy (week 16).
Another drug-related toxicity that has been associated with indinavir is an indirect hyperbilirubinemia. Most patients experienced some increase in bilirubin levels, but the degree of hyperbilirubinemia was clinically insignificant. The median total bilirubin at baseline was 0.4 mg/dL (range, 0.1 to 1.0 mg/dL), compared with 0.5 mg/dL (range, 0.2 to 1.6 mg/dL) at 16 weeks, and 0.6 mg/dL (range, 0.3 to 1.7 mg/dL) at 28 weeks, mainly because of an increase in indirect bilirubin. This increase did not seem to be dose-related but was more pronounced after longer duration of therapy. Median serum alanine aminotransferase and aspartate aminotransferase decreased at all dose levels, whereas the alkaline phosphatase increased slightly from a median of 197 U/L (range, 94 to 356 U/L) at baseline to 220 U/L (range, 102 to 407 U/L) at week 16 and 233 U/L (range, 97 to 465 U/L) at week 28.
The combination of indinavir, zidovudine, and lamivudine was generally well tolerated and no new or unknown toxicities were observed during the combination phase (week 16 to 28). Furthermore, except for the above described episodes of hematuria, none of the patients (including the 8 hemophiliacs or the 2 patients with chronic thrombocytopenia) experienced a change in the incidence, severity, or localization of bleeding manifestations. The mean prothrombin time remained unchanged and, excluding patients with factor VIII deficiency or a central venous catheter (that could be contaminated with heparin), no clinically significant change in activated partial thromboplastin time occurred during the study; in fact, the median activated partial thromboplastin time was 1.6 seconds shorter at 28 weeks compared with baseline (P = .026).
The pharmacokinetics of indinavir administered as free-base suspension, jet-milled suspension, or as dry-filled sulfate salt capsules is summarized in Table 3. Indinavir was rapidly absorbed after the administration of the suspension and capsule formulation (Tmax = 0.8 hours; range, 0.5 to 2 hours; see Fig 1), and the terminal half-life for suspension and capsule formulation was 0.9 hours. The relative bioavailability of the free-base suspension was substantially lower than that of the sulfate salt capsules and was associated with extensive interpatient variability. The mean AUCs(0–8h) of the free-base suspension were 29% and 10% of the AUC of the sulfate salt capsules for the 250 mg/m2and 350 mg/m2 dose levels, respectively. The liquid formulation with smaller particle size (jet-milled suspension) resulted in improved bioavailability but was still inferior to the capsule formulation, especially at the higher dose levels. This prompted us to change therapy in all children to the dry-filled capsule formulation.
After excluding patients who had previously been treated with KNI-272, 48 patients were evaluable for an assessment of efficacy. Because therapy was changed after 12 weeks to the capsule formulation, the data from the children receiving the suspension are only included up to 12 weeks. A new baseline was set at week 16 to be able to evaluate the combination therapy phase in all children.
Serious bacterial or opportunistic infections were relatively uncommon in the study group, despite advanced HIV disease. One patient developed meningococcemia while receiving indinavir (500 mg/m2, capsules, week 16), another was diagnosed with disseminated cryptococcal disease (500 mg/m2, capsules, week 16), 4 patients had a recurrence of herpes zoster, 4 patients were treated for presumed bacterial pneumonia, and 2 each had a urinary tract infection or bacteremia with Staphylococcus epidermidis.
Forty-one of 52 patients (79%) had an increase in weight after 16 weeks on therapy, and in 11 patients it increased ≥10%. After 28 weeks on study, 38 of 42 evaluable patients (90%) had experienced a weight gain and in 18 the increase was >10% greater than baseline.
Comparison of the patients at the 250 mg/m2(n = 7) and 350 mg/m2(n = 6) suspension formulation and the 250 mg/m2 (n = 6) capsule formulation at 28 weeks did not reveal any overall difference between the groups for the neurocognitive measures. Moreover, there was no statistically significant difference in the magnitude of change with therapy between these 3 groups for the Full Scale Intelligence Quotient (FSIQ) (P > .10) or for the Verbal IQ (VIQ) or Performance IQ (PIQ) (P > .10) and the groups were therefore combined as a 250 mg/m2 capsule group (the dosage received after week 12).
A statistically significant (F = 10.14;P < .005) improvement in FSIQ score from 92.8 ± 2.7 to 96.3 ± 2.9, ie, an improvement of 3.5 ± 1.0 (range, −7 to +16), was found in our study population. There was no statistically significant difference in the magnitude of therapy related changes among the 250, 350, and 500 mg/m2 groups. For the component parts of the FSIQ, we found no difference at baseline between mean PIQ at 95.3 ± 2.8 and mean VIQ at 92.3 ± 2.5 (P > .10). However, therapy resulted in a significantly larger improvement in PIQ (to 100.8 ± 3.2;F = 8.34; P < .01) compared with VIQ (increase to 92.9 ± 2.6), but there was no differential effect of dose on the change in these components (F = 0.37; NS).
After 28 weeks on study, 9 of 38 (24%) patients demonstrated a significant improvement of >10% change from baseline and ≥8 points for FSIQ, and/or a change >15% from baseline and ≥12 points for VIQ or PIQ. Two each were treated with 250 and 350 mg/m2suspension, 1 at 350 mg/m2 capsules, and 2 each at 250 mg/m2 and 500 mg/m2 capsules. The majority of patients (27/38) did not show a significant change and 2 patients (at 350 mg/m2 suspension and 500 mg/m2 capsules, respectively), experienced a decline of their FSIQ by 12 and 12 points.
Baseline CD4 counts varied widely in our study population (Table1). One third of the patients (16 of 48) had <50 CD4 positive T cells/mm3 at study entry and more than half of the patients (28/48, 54%) had <200 cells/mm3. Fifteen patients (31%) had a relatively well preserved CD4 count of >500 cells/mm3.
A sustained increase in absolute CD4 cell count was observed at all dose levels (Fig 2). The median increase after 2 weeks was 64 cells/mm3 (n = 46), at 4 weeks 31 cells/mm3 (n = 48), at 12 weeks 58 cells/mm3 (n = 47), and at 16 weeks 60 cells/mm3 (only patients on capsules included, n = 31; P = .0001). There was no statistically significant relationship between change in CD4 count and dose level (P = .57, Kruskal-Wallis test) and no significant difference between the baseline CD4 values between patients who responded or did not respond at 16 weeks (P = .23, Wilcoxon rank sum test). Even patients with severely depressed CD4 counts at entry experienced an increase of similar magnitude. In the 15 patients who had <5% CD4 counts at entry and had a result available after 12 weeks, the median percentage increased from 1% at baseline to 4% after 12 weeks and to 5% after 16 weeks (only patients on capsules included,n = 13). In the 22 patients with <100 cells/mm3 at entry, the median CD4 count increased from 34 cells/mm3 at baseline to 85 cells/mm3 after 12 weeks of indinavir monotherapy and to 88 cells/mm3 after 16 weeks (only patients on capsules included, n = 16).
The addition of zidovudine and lamivudine at week 16 resulted in an additional median increase of 25 cells/mm3 by week 20, and 37 cells/mm3 by week 28. A separate evaluation by dose level is not possible, because only 5 children were actually treated at 500 mg/m2 per dose after week 16.
Forty-nine of the 53 (92%) patients who had a result available at baseline, were anergic to C albicans skin testing. After 16 weeks of indinavir monotherapy, 6 of 49 (12%) previously anergic patients had a positive skin test, and after 28 weeks 5 of 28 (18%) reacted positively. Four of 52 patients (8%) had a positive skin test to tetanus toxoid at baseline and 2 patients with negative baseline tests had a positive result after 16 weeks of indinavir monotherapy. None of the patients had a positive PPD at baseline but 1 patient had a 9 × 9-mm induration after 28 weeks. This patient had previously been treated with isoniazid for PPD conversion and experienced a moderate increase in his CD4 cell count from 9 cells/mm3 to 45 cells/mm3 after 28 weeks of therapy.
The median baseline plasma HIV RNA level was relatively high (4.76 log10 copies/mL, Table 1), consistent with other studies.11-13 Seventeen of 48 evaluable patients (35%) had a baseline plasma HIV RNA level >100 000 copies/mL (>5.0 log10). Patients receiving 250 mg/m2 of the capsule formulation of indinavir experienced a maximal decrease in plasma HIV RNA levels after 2 weeks (−0.72 log10copies/mL) but then it started to rise again (Fig 2). At the 350 mg/m2 and 500 mg/m2 dose levels the maximal response was reached after 4 weeks on therapy and was more pronounced (−1.22 log10 copies/mL and −1.35 log10copies/mL, respectively). However, an increase to approximately 0.5 log10 copies/mL below baseline was observed in both the 350 mg/m2 and 500 mg/m2 patient cohorts by week 8. Only 6 of 31 patients (2 on 250 mg/m2 and 4 on 500 mg/m2 capsules) met the definition of virologic response (as defined by a sustained decrease in HIV RNA of 1 log10copies/mL or more) by week 16. The median decrease in HIV RNA at week 16 was 0.76 log10 copies/mL in patients receiving 500 mg/m2 of indinavir compared with 0.07 log10copies/mL for patients at the 250 mg/m2 and 350 mg/m2 dose levels. However, the differences between the value at baseline and at 16 weeks were not statistically significant (P = .06), and there was not a statistically significant relationship between dose level and change from baseline to week 16 or 28 (P = .14 and P = .68, respectively, Kruskal Wallis test).
The addition of zidovudine and lamivudine resulted in an additional decrease in viral load of 1.36 log10 copies/mL between week 16 and 18 (n = 34). At week 28 (n = 35) the level was still 0.70 log10 copies/mL below the week 16 level. At week 28, 16 of 35 patients (45%) had achieved a viral response. Although the HIV RNA level became undetectable in only 1 patient during the monotherapy phase (between week 4 to 12, at 250 mg/m2 capsules), 3 patients (1 each at 250, 350, and 500 mg/m2 capsules) sustained a level of <200 copies/mL after the initiation of triple therapy for >2 months.
Standard antiretroviral therapy currently consists of the combination of two reverse transcriptase inhibitors and a protease inhibitor.14 15 However, most children who might benefit from protease inhibitors are not able to swallow capsules or tablets but nonetheless should have equal access to potent antiretroviral therapies. Accordingly, the development of formulations that can be dosed based on body weight or body surface area and which can be administered to infants and young children must be considered a high priority in the development of new agents, and recently legislation has been proposed to require manufacturers to include pediatric data whenever the drug is likely to be used in children.
In 1996, indinavir as sulfate dry-filled capsules was approved by the Food and Drug Administration for the treatment of HIV-infected adults at a dose of 800 mg every 8 hours. In this pediatric phase I/II study we evaluated a free-base suspension of indinavir with an acceptable taste. The indinavir suspension was well tolerated at the two dose levels tested (250 and 350 mg/m2) and was rapidly absorbed in the children studied with a time to peak concentration similar to what has been described in adults.2 Unfortunately, comparative analysis of the pharmacokinetics obtained after the same dose of the liquid formulation and the capsule formulation revealed not only a marked interpatient variability but also a substantially lower absorption of the indinavir free-based suspension, raising the possibility of subtherapeutic exposure and the subsequent emergence of resistant virus strains. Accordingly, we changed the protocol to the evaluation of the capsule formulation, although we continued to perform pharmacokinetics with different liquid formulations that the manufacturer prepared in an attempt to overcome the deficits noted. A jet-milled suspension seemed to be better absorbed than the standard free-base suspension, but was still less bioavailable than the capsule formulation.
The area under the time concentration curve in 3 children treated with indinavir capsules at 500 mg/m2 was comparable to values obtained after a dose of 800 mg (roughly equivalent to 460 mg/m2) of the dry-filled capsule formulation in adults, but the half lives seemed to be shorter, resulting in lower trough concentrations. The capsule formulation was relatively well tolerated. Hematuria was the major complication, occurring in 7 patients and associated with nephrolithiasis in 1 of them (Table 2). Of concern, when we increased (after 48 weeks of study) the dose of indinavir in all patients to 500 mg/m2, we observed a cluster of 4 more episodes of hematuria (including another patient with nephrolithiasis), which led the study group in September 1996 to reduce the maximum indinavir dose in this trial to 350 mg/m2. Adjusted for body weight, doses of indinavir in children may result in higher Cmax than in adults. Whether this fully explains why hematuria and nephrolithiasis were relatively common in our pediatric patients is unknown. At least one other group observed a significantly lower incidence of hematuria at similar or even higher doses of indinavir.16
Central nervous system compromise is a common and often early symptom of HIV infection in children.17-19 The trend toward improvement in neuropsychometric function that was observed in this study is promising; however, no children with severe encephalopathy were enrolled. Indinavir is approximately 60% protein-bound and the positive effect observed in this study could possibly be attributable to passage of indinavir through the blood-brain barrier (not measured). Furthermore, a decrease in the amount of circulating virus, perhaps combined with a change in cytokine milieu, may have led to this observed improvement. These findings need to be confirmed in a larger study population, and especially in children with encephalopathy.
Indinavir monotherapy resulted in a rapid dose-dependent decline in plasma HIV RNA levels, but viral RNA levels began to increase at all dose levels after a few weeks of therapy, reaching a level of approximately 0.6 log10 below baseline values at week 16 (Fig 2). This increase in viral load after an initial response is presumably attributable to the emergence of resistant virus and it seemed to occur more rapidly in children than in adults. The children enrolled in this study were heavily pretreated (median length of prior therapy 5 years), and most of them had previously been exposed to zidovudine. However, viral load decreased again after the addition of zidovudine and lamivudine at week 16. It is likely that de novo treatment with the combination of indinavir, zidovudine, and lamivudine will be more effective, as demonstrated in adults.3 4
In summary, although indinavir is a potent and generally well-tolerated antiretroviral agent (except for the relatively common occurrence of reversible hematuria and occasional nephrolithiasis at higher doses), the bioavailability of the currently available pediatric formulations is not yet satisfactory. However, the capsules used in adults do seem to offer benefits in children who are able to swallow them. There may be differences in the incidence of side effects (especially hematuria) in children compared with adults. Accordingly, the combination of indinavir, zidovudine, and lamivudine seems to be generally well-tolerated in children >4 years of age. Whether indinavir will prove to be as efficacious in children as in adults when given in combination with other antiretroviral agents as primary therapy remains to be evaluated. With the current knowledge about the rapid emergence of resistance during monotherapy, it will be of utmost importance to considerably shorten the time of monotherapy required for safety evaluations (probably to a maximum of 2 to 4 weeks) in future trials. Our trial underscores the problems that can hamper and delay the development of pediatric formulations and dosing recommendations. It clearly emphasizes the urgency of developing appropriate pediatric formulations and of initiating careful pharmacokinetic and safety evaluations of new therapeutic modalities in children in tandem with phase I studies in adults. Such an approach will help to assure equal access to promising new agents for children and adults.
The authors would like to acknowledge the data management support of Carol Boss, MPH, Carol Meyer, and Sanju Johri; the clinical support provided by the Nursing and Social Services Staff of the Pediatric Branch, NCI, and especially Colleen Higham, PNP, Ellen Townley, PNP, Anne-Marie Boler, PNP, Gayl Selkin-Gutman, MA, Lori Perez, PhD, Claire Walsek, PNP, Pam Wolters, PhD, Linda L. Lewis, MD, Lori Wiener, PhD, Stacy Shiflett, PharmD, Shirley Jankelevich, MD, Leslie Serchuck, MD, Christine Rosko, RN, Margie Sullivan, and all the physicians involved in the care of the study patients; the laboratory help of Kirsta Waldon, Gene Shearer, PhD, and Michael Baseler, PhD, and David Waters, PhD.
- Received October 29, 1997.
- Accepted January 28, 1998.
Reprint requests to (B.U.M.) Department of Medicine, Hunnewell 302, HU-215, Children's Hospital, 300 Longwood Ave, Boston, MA 02115.
- ↵Centers for Disease Control and Prevention. US HIV and AIDS cases reported through December 1997. HIV/AIDS Surveillance Report. 1997;9:1–40
- Mulder J,
- McKinney N,
- Christopherson C,
- Sninsky J,
- Greenfield L,
- Kwok S
- ↵Brouwers P, Belman A, Epstein L. Central nervous system involvement: manifestations, evaluation, and pathogenesis. In: Pizzo PA, Wilfert CA, eds. Pediatric AIDS. The Challenge of HIV Infection in Infants, Children, and Adolescents. Baltimore, MD: Williams & Wilkins; 1994:433–455
- Centers for Disease Control and Prevention
- Copyright © 1998 American Academy of Pediatrics