A Phase I/II Study of the Protease Inhibitor Ritonavir in Children With Human Immunodeficiency Virus Infection
Background. Ritonavir, a potent antiretroviral protease inhibitor, has been approved for the treatment of adults and children with human immunodeficiency virus (HIV) infection. In a phase I/II study, we assessed the safety, tolerability, and pharmacokinetic profile of the oral solution of ritonavir in HIV-infected children and studied the preliminary antiviral and clinical effects.
Methods. HIV-infected children between 6 months and 18 years of age were eligible. Four dose levels of ritonavir oral solution (250, 300, 350, and 400 mg/m2 given every 12 hours) were evaluated in two age groups (≤2 years, >2 years). Ritonavir was administered alone for the first 12 weeks and then in combination with zidovudine and/or didanosine. Clinical and laboratory parameters were monitored every 2 to 4 weeks.
Results. A total of 48 children (median age, 7.7 years; range, 0.5 to 14.4 years) were included in this analysis. Dose-related nausea, diarrhea, and abdominal pain were the most common toxicities and resulted in discontinuation of ritonavir in 7 children. Ritonavir was well absorbed at all dose levels, and plasma concentrations reached a peak 2 to 4 hours after a dose. CD4 cells counts increased by a median of 79 cells/mm3 after 4 weeks of monotherapy and were maintained throughout the study. Plasma HIV RNA decreased by 1 to 2 log10 copies/mL within 4 to 8 weeks of ritonavir monotherapy, and this level was sustained in patients enrolled at the highest dose level of 400 mg/m2 for the 24-week period.
Conclusions. The oral solution of ritonavir has potent antiretroviral activity as a single agent and is relatively well tolerated by children when administered alone or in combination with zidovudine or didanosine.
- AIDS =
- acquired immunodeficiency syndrome •
- HIV =
- human immunodeficiency virus •
- RNA =
- ribonucleic acid •
- bDNA =
- branched chain deoxyribonucleic acid •
- CDC =
- Centers for Disease Control and Prevention •
- NCI =
- National Cancer Institute •
- AUC =
- area under the time-concentration curve •
- Cl/F =
- apparent clearance •
- GIMA =
- General Index of Mental Abilities •
- NS =
- not significant •
- Cmax =
- maximal concentration •
- CNS =
- central nervous system
As of December 1996, >7600 children have been diagnosed with acquired immunodeficiency syndrome (AIDS) in the United States.1 Approved treatment options for human immunodeficiency virus (HIV)-infected children include reverse transcriptase inhibitors (zidovudine, didanosine, and lamivudine). Recently, the Food and Drug Administration approved the protease inhibitors ritonavir and nelfinavir for use in children, the former based on the data contained in this report.
Ritonavir acts as a competitive inhibitor of the HIV protease that cleaves the Gag and Pol polyproteins into individual functional proteins. In cultured cells infected with various laboratory strains and clinical isolates (including zidovudine-resistant strains) of HIV-1, ritonavir results in a 90% inhibition of viral replication at concentrations between 0.06 and 0.6 μm.2-4Studies in adults demonstrate decreases in plasma viral ribonucleic acid (RNA) of >1 log10 and increases in CD4 counts of >50 cells/mm3 at doses of ritonavir between 400 and 1200 mg per day (divided into 2 to 4 doses).2 5 At a dose of 600 mg twice daily, a median increase of >100 cells/mm3 was sustained for >32 weeks of ritonavir monotherapy. In adults with a plasma RNA level >25 000 equiv/mL (branched chain deoxyribonucleic acid [bDNA], Chiron), ritonavir therapy produced a rapid reduction of the mean viral load by 2.1 log10 (range, 1.3 to 2.5 log10).6 7
Combination therapies that include two reverse transcriptase inhibitors and a protease inhibitor have been demonstrated to have a significant impact on the level of plasma HIV RNA and clinical outcome in adults.8 This report presents the results of the first 24 weeks of treatment of a phase I/II study in HIV-infected children, designed to determine the safety, tolerability, and pharmacokinetic profile, as well as the preliminary efficacy of ritonavir given as monotherapy and then in combination with two dideoxynucleosides.
HIV-infected children between the ages of 6 months and 18 years were eligible for enrollment, regardless of gender, race, or route of transmission. Previously untreated, asymptomatic children were eligible if their age-corrected absolute CD4 count rendered them at risk for an AIDS-related opportunistic infection (immunologic categories 2 and 3 of the Centers for Disease Control and Prevention [CDC] classification).9 Children with moderate to severe symptomatic HIV infection (CDC classes B and C), including those who had become refractory or intolerant to prior therapy, also were eligible.
Baseline laboratory values required for entry included the following: a total white blood count >1500 cells/mm3; neutrophil plus band count >750 cells/mm3; hemoglobin >8.0 g/dL; platelet count >50 000/mm3; serum creatinine less than two times the upper limit of normal; hepatic transaminases less than three times the upper limit of normal; total bilirubin <1.5 mg/dL; fasting triglyceride level <400 mg/dL; pancreatic amylase less than two times the upper limit of normal. Patients were required to be in stable clinical condition and to have discontinued all antiretroviral therapy at least 2 weeks before study entry. Treatment with immunomodulating agents (except stable steroid therapy for lymphocytic interstitial pneumonitis or an autoimmune process), cytolytic chemotherapy, or radiation therapy within 30 days before entry was considered an exclusion criteria. Stable therapy (same dose for at least 4 weeks before entry) with granulocyte colony-stimulating factor was allowed.
This study was performed by the HIV and AIDS Malignancy Branch, National Cancer Institute (NCI) in collaboration with the Department of Pediatrics of the University of Florida and the University of South Florida/All Children's Hospital. 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 before entry.
Children were stratified by two age groups, which enrolled patients independently into escalating dosage levels. Three to five children between 6 months and 2 years of age, and four to six children >2 years of age (with a minimum of three children between 2 and 12 years of age) were entered into each dose level. At least three patients had to complete 4 weeks of therapy before the first patient could be entered into the next higher dose level. To limit nausea, ritonavir was begun at a dosage of 250 mg/m2 twice daily and then, in patients assigned a higher dose level, increased over a 5-day period to reach the protocol-specific level.
Four dose levels of the liquid formulation of ritonavir were evaluated: 250 mg/m2, 300 mg/m2, 350 mg/m2, and 400 mg/m2, each given orally every 12 hours. Ritonavir was administered as monotherapy during the first 12 weeks, after which zidovudine (90 mg/m2 per dose given every 6 hours) and didanosine (90 mg/m2 per dose given every 12 hours) were added. If a child had a history of severe intolerance to one of these dideoxynucleosides, only the tolerated agent was added. Parents and guardians were instructed to give didanosine 2 to 4 hours apart from ritonavir and on an empty stomach. Chocolate, peanut butter, yogurt, or any other protein- or fat-containing food or drink was used before and after ritonavir administration to mask the unpleasant, long-lasting taste. If ritonavir had to be given through a nasogastric or gastric tube, a protein-containing solution, such as milk or an enteral feeding solution, was flushed through the tube before and after the dose.
Patients who experienced new onset grade III toxicity or persistent grade II toxicity that lasted for >1 month had the study medication(s) discontinued temporarily until the abnormality returned to baseline or symptoms resolved. If the toxicity resolved within 1 month, ritonavir was reintroduced at the next lower dose level, and if the toxicity was not thought to be study-drug related, the dose again could be increased to 100% of the assigned dose at the investigator's and study sponsor's discretion. Patients who experienced any grade IV toxicity had their study medication discontinued permanently.
Clinical and Laboratory Monitoring
All eligible patients were evaluated initially in the outpatient clinic of the NCI. Vital signs were monitored during the first 72 hours of the study. Patients were seen every 2 weeks until week 14 and then every 4 weeks. Skin tests for mycobacteria (purified protein derivative, 5 TU) with Candida albicans (1:100) and tetanus toxoid (1:5) control were performed at entry and at weeks 12 and 24. Echocardiography, electrocardiography, chest radiography, and computer tomography of the head were performed at entry and at 24 weeks. An eye examination was performed by an ophthalmologist every 4 weeks (with photographic documentation if feasible), and an electrooculography was performed, where feasible, in children >6 years old, at entry and at weeks 12 and 24.
Patients were evaluated with an age-appropriate comprehensive neuropsychometric battery at entry and at 24 weeks that included a full standardized intelligence test (McCarthy or Wechsler Intelligence Scale for Children) or the Bayley Revised Test. In addition, a monitor assessment was administered at entry, 12, and 24 weeks, which included the Peabody Picture Vocabulary Test Revised, the Gardner One-word Expressive Language Test, the Beery Visual Motor Integration Test, the colored or standard Raven's Test, and the Digit Span Subtest (McCarthy or Wechsler Intelligence Scale for Children).10
Complete blood counts, coagulation profile, routine biochemistry values (including triglycerides), and urinalysis (including urine pregnancy test in postmenarchal females) were monitored throughout the study. A fluorescence-activated cell sorter analysis of lymphocyte subsets was obtained on days 2, 7, and 14, as well as on day 7 of week 12 and day 4 of week 13, and every 4 weeks between weeks 4 and 24. Plasma HIV RNA by polymerase chain reaction assay (Roche Amplicor; lower limit of detection 200 copies/mL)11 was obtained on days 1, 2, 3, 4, 7, 14, and 28, and at weeks 8, 12 (day 7), 13 (days 1, 2, 3, 4), 14, 16, and 24. An undetectable HIV RNA level was considered to be 199 copies/mL.
A baseline, pretreatment specimen was drawn on day 1. Patients were hospitalized for 24 hours for the pharmacokinetic measurements on day 7, but not on day 28. On days 7 and 28, 3 mL of blood was collected immediately before the morning dose of ritonavir (12 hours after the preceding dose), and then 0.5, 1, 2, 3, 4, 6, 8, and 12 hours after the morning dose. Ritonavir concentrations in plasma were analyzed by reverse-phase high-performance liquid chromatography.4 12 The area under the time-concentration curve (AUC) was calculated using the linear trapezoidal rule, and the apparent clearance (Cl/F) was estimated from the ratio of the administered dose to the calculated AUC.
Criteria for Response to Treatment
Clinical response was defined as a sustained weight gain, evidenced by an increase of at least 10% over baseline, that was not associated with the initiation of intravenous hyperalimentation. Improvement in neurocognitive function was defined as an increase of 10% over baseline in age-appropriate full-scale IQ score and at least eight IQ points, because changes of this magnitude are unlikely to be caused by practice effect.10 Evidence of improvement also could be determined by changes in motor function or improvement in neural imaging studies. Immunologic response was defined as an ≥10% increase of over the baseline in absolute CD4 cell count with a minimum increase of 50 CD4 cells on two consecutive measures at least 4 weeks apart. Response in CD4 percentage was defined as an ≥25% increase in the percentage of CD4 cells to total lymphocytes on two successive measures at least 4 weeks apart. A response in viral load was defined as an ≥1 log10 copies/mL decrease in patients with a measurable plasma HIV RNA level at entry.
The comparison of baseline values among dose groups was performed using the Kruskal–Wallis test. Comparison of changes from baseline between-dose levels or other groups as described was performed with the Wilcoxon rank sum test. All P values are two-sided.
Changes in neuropsychologic function were evaluated using repeated-measures analysis of variance, and a type-1 α level of 0.05 was used. Change from baseline was calculated using methods published previously.13 Fisher's exact test was used to evaluate possible differences in number of patients improving or declining in two compared groups.
A two-way analysis of covariance was conducted to test whether dose group, gender, age, body weight, and baseline CD4 count had an effect on time to maximal concentration (T max) and logarithmically transformed dose-normalized maximal concentration (C max) and AUC 12 h. Because age and body weight are highly correlated (R 2 = 0.75), their effects were modeled separately. A significant level of 0.05 was used for all tests.
Between June 1995 and December 1996, 51 children with HIV infection were enrolled into this study. This report includes results from the 48 children (3 previously untreated, 1 treated with 6 weeks of zidovudine postpartum, and 44 with refractory disease or intolerance to prior antiretroviral therapy) who completed at least 12 weeks of ritonavir therapy on protocol (as of January 16, 1997) or who have come off study before 12 weeks.
As of January 16, 1997, 30 of the 48 children had completed 24 weeks of therapy. None of the patients remained on monotherapy after week 12. Thirty of 38 patients who had reached the combination phase received triple therapy with ritonavir, zidovudine, and didanosine after week 12. Four children received only didanosine, and 4 received only zidovudine in combination with ritonavir because of a history of prior toxicity that was attributed to the other dideoxynucleoside. Five of 48 children were <2 years of age. Baseline patient characteristics are summarized in Table 1.
The 45 previously treated children all had been treated with zidovudine, 40 with didanosine, 18 with lamivudine, 7 with stavudine, 3 with zalcitabine, and 1 with nevirapine. Four patients had been enrolled previously in a 12-week study of the protease inhibitor KNI-272 (one patient for only 2 weeks); their data have been included in the pharmacokinetic and safety analyses but excluded from the efficacy analysis.
Twenty-five children (52%) had advanced (category C) HIV infection as defined by the CDC classification system.14 No statistically significant differences were observed in the baseline parameters for the children enrolled into the four dose levels.
Ten patients were removed from study for toxicities or intolerance, and one patient discontinued study drug because of an embolic stroke. This patient had a history of cerebral infarcts related to a preexisting coagulopathy. Seven patients who were still on study had not yet reached 24 weeks of follow-up (protocol weeks 12 to 20). No death occurred during this study.
Safety and Tolerance
Mild and transient nausea, diarrhea, and abdominal pain were experienced at some point during the study by 98% of the patients, but these symptoms usually subsided within 2 to 8 weeks of starting ritonavir monotherapy. Seven children (14%) withdrew from study because of gastrointestinal toxicity. Four children (1 at 250 mg/m2, 2 at 350 mg/m2, and 1 at 400 mg/m2, respectively) had intolerable nausea and vomiting despite dose reduction, necessitating their removal from study after 24, 3, 13, and 5 weeks, respectively. Two children (1 at 350 mg/m2 and 1 at 400 mg/m2, respectively) refused to continue therapy after 4 weeks, following an infectious episode (bacteremia and pulmonary cryptococcal infection, respectively), despite the offer to reduce the dose. One child (350 mg/m2), whose dose had been reduced after 2 weeks to 300 mg/m2 because of gastrointestinal complaints, refused to continue after 16 weeks on study, because of excessive burping and mild malaise.
Therapy was interrupted in four patients (one at 250 mg/m2, one at 300 mg/m2, and two at 400 mg/m2, respectively) because of elevations of hepatic transaminases to more than five times the upper limit of normal (grade III toxicity). In three of these patients, abnormal values persisted for more than 1 month, necessitating their removal from study after 8 (250 mg/m2), 23 (300 mg/m2), and 5 weeks (400 mg/m2) of treatment, respectively. Each of these patients had a history of intermittently elevated hepatic enzymes. The hepatic transaminases in the fourth patient returned to normal after interruption of ritonavir monotherapy for a few days (week 5), and reintroduction of the study drug at a lower dose level (350 mg/m2) has been well tolerated.
Dose reductions were necessary in two other patients at the 400 mg/m2 dose level. One patient with preexisting frequent hematemesis and bloody stools of unknown etiology experienced a worsening of gastrointestinal bleeding during the first 12 weeks of treatment on study drug. Reduction of the dose of ritonavir to 350 mg/m2 was better tolerated, and the frequency of gastrointestinal bleeding episodes decreased to less than the baseline frequency. A second patient developed isolated thrombocytopenia (platelets <50 000) at week 16, and all three study drugs were temporarily discontinued. (This patient was later removed from study because of persistent thrombocytopenia thought to be related to either disease progression or therapy.)
There was no evidence of an increase in the bleeding tendency of patients without baseline clotting disorders. 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 statistically significant increase in activated partial thromboplastin time occurred during the study. In addition to the child with (preexisting) gastrointestinal bleeding described above, six other children experienced bleeding episodes, but none of them were different from normal childhood experiences (three episodes of epistaxis, two hematomas, and one bleed of the oral mucosa). One of the three adolescent males with Factor VIII deficiency had a renal bleed, and one reported an increase in bleeding episodes from three to four per month to one to two per week with no change in localization or severity.
Nine patients complained of a decreased appetite, whereas 14 reported an increase. Fifteen patients (31%) complained of one or more episodes of headache. One patient had moderately severe headaches that subsided after stopping zidovudine. Two patients had side effects thought to be attributable to didanosine therapy: 1 patient developed a clinically asymptomatic elevation in amylase and lipase and the other patient appeared to have an allergic reaction with increased salivation, wheezing, and fever. In both instances, didanosine was stopped, the symptoms resolved, and the other study drugs were well tolerated. No patient developed peripheral retinopathy, a toxicity observed with didanosine.15
Statistically significant changes occurred in several laboratory values, albeit none of them clinically significant. An increase in alkaline phosphatase was observed from a mean of 207 U/L at baseline to 281 U/L at week 12 and to 318 U/L at week 24 (P= .0001, each). The mean gamma glutamyl transpeptidase increased from 71 U/L at baseline to 145 U/L after 24 weeks (P= .017). However, hepatic transaminases exhibited a nonsignificant trend toward lower values. As expected, mean serum triglyceride levels increased from a baseline of 130 mg/dL (maximum 397 mg/dL) to 206 mg/dL at 12 weeks (maximum 460 mg/dL; P = .0001) and 209 mg/dL (maximum 646 mg/dL; P = .0005) at week 24. This increase was more pronounced at the higher dose levels. Mean serum cholesterol levels increased from 121 mg/dL at baseline to 184 mg/dL at 12 weeks and 191 mg/dL at 24 weeks (P = .0001, each). None of the patients developed diabetes mellitus or a significant increase in blood glucose levels.
At least one set of pharmacokinetic samples was available from 44 patients. Pharmacokinetic parameters on day 28 (at steady state) for 37 children >2 years of age are shown in Table2. Peak plasma concentrations (C max) were achieved within 2 to 4 hours of the dose and exceeded 2.1 μg/mL (range, 3.6 to 32.8 μg/mL) at all dose levels. This concentration is predicted to inhibit HIV replication by 90% (EC90) based on preclinical data and accounting for protein-binding of ritonavir in plasma.3-5The median steady-state (day 28) plasma concentrations of ritonavir for the 400 mg/m2 dose group were above the protein binding-corrected EC90 of ritonavir throughout the 12-hour dosing period.
The day 28 plasma concentrations appeared to be lower than or similar to the day 7 values (Fig 1). However, the changes in AUC and C max with time were not statistically significant. Although there was considerable interpatient variability, statistical analysis did not show a significant effect of age, gender, or body weight on steady-state pharmacokinetics in these children. In addition, the dose-normalized C maxand AUC values were not statistically significantly different among dose groups, indicating that the increase of steady-state ritonavir plasma concentrations were proportional to dose.
Pharmacokinetic data for four children <2 years of age were available. Two 6-month-old infants treated at 250 mg/m2 had very high Cl/F values of 56.5 L/m2/h and 82 L/m2/h on day 7. The Cl/F values on day 28 in these infants were 36.3 L/m2/h (at 250 mg/m2) and 16.4 L/m2/h (at 300 mg/m2). Poor bioavailability may in part account for the high Cl/F values in these infants. The Cl/F values in two 18-month-old children were comparable with the values in older children.
After excluding the 4 patients who had been treated previously with another protease inhibitor (KNI-272), 44 patients were evaluable for response to ritonavir.
Infectious episodes during the study period were uncommon in this patient population with advanced HIV infection. One patient developed pulmonary cryptococcosis 2 weeks after starting ritonavir, one patient developed Pneumocystis carinii pneumonia after 16 weeks on therapy, and one patient was diagnosed with osteomyelitis of the tibia after 12 weeks on protocol. Four patients had respiratory tract infections, two patients were diagnosed with central line-related infections, and one patient had a recurrence of herpes zoster. Sixteen patients had mild oropharyngeal thrush. None of the patients progressed to a more advanced stage as graded by the CDC classification.
No change in neuropsychometric scores was observed during the first 24 weeks of the study. Baseline neuropsychometric evaluations included 39 patients with a mean age of 7.2 ± 0.7 years and a General Index of Mental Abilities (GIMA) of 88.7 ± 2.4 (range, 57 to 123). After 12 weeks of ritonavir monotherapy, the estimated GIMA (EsGIMA) was unchanged (n = 37; mean EsGIMA 0 = 91.7, mean EsGIMA 12 = 92.5; F = 0.88; not significant [NS]), and comparison of baseline with the evaluation at 24 weeks (n = 21; meanGIMA 0 = 87.6, mean GIMA 24= 90.7; F = 0.39; NS) did not demonstrate a change. There was no association with dose level (F = 0.87; NS). The dose groups were relatively well matched with respect to GIMA score (mean GIMA values of 87.3, 92.9, 90.3, and 81.0 for dose levels 1 to 4, respectively). Similarly, there was no significant change in the verbal IQ and performance IQ for the children >30 months of age, and there were no differential effects for the two measures (F = 2.11; P > .16).
Baseline CD4 counts ranged from 0 to 2022 cells/mm3, and 19 of 43 (44%) patients had < 50 cells/mm3 at study entry. The median change in CD4 counts was similar at all dose levels, when all 43 patients with baseline data were included. (One patient did not have a baseline CD4 count and has been excluded from this analysis.) A median increase in CD4 counts of 79 cells/mm3 (n = 34) was observed after 4 weeks of ritonavir monotherapy. This increase was sustained throughout the monotherapy phase. The median increase between baseline and week 12 was 75 cells/mm3 (n = 35, P = .0001) and 203 cells/mm3 over baseline at week 24 (n = 27, P = .0001).
There was no apparent dose–response effect for the increases observed in CD4 counts if all patients are included. However, if patients <2 years of age were excluded from the analysis, the CD4 increment at the 250 mg/m2 dose level appeared to be less pronounced than at higher dose levels (Fig 2). Overall, 19 of 35 (54%) patients experienced a response in absolute CD4 counts by week 12 (ie, during ritonavir monotherapy), and 18 of 26 (69%) by week 24 (during combination therapy). The increase in CD4 cells was not restricted to patients with a relatively well-preserved immune status, and neither the baseline CD4 count nor the CD4 percentage was predictive of response. Among the 18 patients with a CD4 percentage of <5%, the median percentage increased from 2% at baseline (n = 18) to 4% at week 12 (n = 15) and 11% at week 24 (n = 11). For some patients, the increase was quite marked (Fig 3).
Thirty-seven of 43 evaluable patients (86%) were anergic (defined as no induration measurable) to C albicans skin testing at baseline. Thirty-two patients had results available at weeks 0 and 12, and 6 (18.7%) of these patients became newly skin test-positive after 12 weeks of ritonavir monotherapy. For tetanus toxoid, 1 of 43 patients (2.3%) had a positive skin test at baseline, compared with 4 of 32 (12.5%) after 12 weeks and 7 of 23 (30%) after 24 weeks of therapy. None of the patients had a positive purified protein derivative value at baseline, but 1 patient developed a 13 × 20-mm area of induration after 12 weeks, and another had a 6 × 6-mm area of induration at 24 weeks. The first patient had been diagnosed withMycobacterium avium complex bacteremia shortly before entry on study, and his chest x-ray was normal. However, his CD4 cell count increased from 0 cells/mm3 (0%) at baseline to 215 cells/mm3 (24%) after 12 weeks of ritonavir monotherapy. The second patient did not have a known exposure to M tuberculosis, had a normal chest x-ray, and was not bacteremic with mycobacterium avium complex during the study period. Her CD4 count was relatively well preserved (583 cells/mm3) at baseline and did not change markedly while on study. Both patients had isoniazid added to their treatment regimen.
As described in other HIV-infected pediatric populations,16 the viral load as measured by plasma HIV RNA copies was higher than those observed in adults (Table 1), and 61% of the children in the present study had a baseline viral load of >100 000 copies/mL (>5.00 log10). A maximum decrease in HIV RNA copy numbers of 1 to 2 log10 copies/mL was achieved after 4 to 8 weeks of ritonavir monotherapy (Fig4), after which the viral load rebounded to ∼0.5 log10 copies/mL below baseline at the three lowest dose levels. Patients enrolled at the highest dose level (400 mg/m2) experienced a sustained median decrease in plasma HIV RNA levels of 2 log10 copies/mL during the study period. The addition of zidovudine and didanosine after 12 weeks of ritonavir monotherapy resulted in another decrease of 0.5 to 1 log10 copies/mL at all dose levels; however, this was sustained for the study period only at the highest dose level.
The HIV RNA level at baseline was not predictive for response (P = .94), and CD4 count at baseline was significant for an association with response in HIV RNA level after 12 weeks (P = .043) and was significantly associated with a response at 24 weeks (P = .015), as was CD4 percentage (P = .0079).
In seven (15.9%) patients, (one each at the three lower dose levels and four at the highest dose level), the HIV RNA level decreased to undetectable levels (<200 copies/mL) and remained such for at least 2 consecutive measurements taken 4 weeks apart. Not only was this more common at the highest dose level of ritonavir, but the duration of time with undetectable levels was longer in patients at 400 mg/m2 per dose (>3 months compared with 1 to 2 months on the lower levels).
Ritonavir oral solution was relatively well tolerated by most children enrolled, despite the frequent occurrence of nondose-limiting side effects. The unpleasant taste of the liquid formulation was a minor and transient problem; however, as in adults, mild to moderate gastrointestinal complaints were common, leading to discontinuation of study medication in 14% of children. Interestingly, six of the seven children who discontinued ritonavir therapy because of gastrointestinal side effects were > 10 years of age, although this age group comprised only 33% (n = 16) of the total study population. In three children, gastrointestinal symptoms became unacceptable only after the addition of zidovudine and didanosine. Didanosine can cause diarrhea and has to be given in a fasting state, whereas ritonavir should be given with a meal. Therefore, ritonavir may be more tolerable in combination with dideoxynucleosides other than didanosine.
Although the elevation in serum transaminases observed in three patients was not clearly caused by ritonavir, the potential for hepatotoxicity is of concern and necessitates caution when using this drug in children with preexisting liver function abnormalities. The liquid formulation of ritonavir contains 43% ethanol and, theoretically, this could impact liver function tests. However, serum transaminases declined in most patients, and serum GGT and alkaline phosphatase remained unchanged. An increase in triglyceride and cholesterol levels was common but did not reach clinical significance. None of the five children <2 years of age experienced any side effects, although the numbers are too small to make a general statement about tolerance in this age group.
Oral ritonavir was readily absorbed in pediatric patients. Peak concentrations were achieved within 2 to 4 hours, similar to results obtained in HIV-infected adults.2 5 The increase inC max and AUC was proportional to dose and the Cl/F value was dose-independent, and age showed no effect on ritonavir pharmacokinetics in this study. The variability in the pharmacokinetic parameters observed in pediatric patients (especially at the 400 mg/m2 dose level) was substantially larger than that observed in adults, but considering the wide range of age and weight in the present study, this is not totally unexpected. Ritonavir induces its own metabolism; however, the changes in AUC andC max between days 7 and 28 were not statistically significant. Because the turnover half-life of cytochrome P450 3A has been estimated to be ∼3 days, 80% of the induction is expected to be accomplished by day 7 for a fixed regimen.17 18 Given the variability observed in these children, a larger population would be required to detect a difference of this magnitude between days 7 and 28. Because ritonavir concentrations obtained in the 350 mg/m2 and 400 mg/m2 twice daily dose groups are comparable with those obtained in adults receiving 600 mg twice daily, children between 2 to 14 years of age can be given 350 to 400 mg/m2 twice daily. The number of children <2 years old enrolled in this study is too small to draw any firm conclusions regarding pharmacokinetics in this age group.
The increase in CD4 counts was independent of dose level and was usually sustained even if the HIV RNA level was increasing again. In contrast to treatment with reverse transcriptase inhibitors such as didanosine, for which increases in CD4 cell count were predominantly found in children with pretreatment CD4 counts >100 cells/mm3, ritonavir resulted in marked and sustained improvements in CD4 cell numbers even in children with very low counts at study entry.19 20
Comparable with the results reported in HIV-infected adults, ritonavir therapy resulted in a marked and rapid decline of plasma HIV RNA levels.2 5 The reduction in plasma HIV RNA copy numbers was more pronounced than that described previously for children and adults treated with reverse transcriptase inhibitors.21-25In adults, HIV RNA levels at baseline are predictive for development of clinical disease.26 27 Children, especially those <4 years of age, tend to have higher plasma HIV RNA levels than adults, independent of CD4 cell count.16 28 29 The extent of viral response did not appear to be related to baseline viral load in our study.
For safety reasons, ritonavir was given as monotherapy for the first 12 weeks, which at the time was an unusually short period for a pediatric phase I study. However, as is frequently observed with monotherapy, a gradual increase of plasma HIV RNA levels toward baseline values, presumably because of the development of resistance, occurred after a few weeks of treatment at potentially subtherapeutic dosages (Fig4).30-32 Despite extensive prior exposure to zidovudine and didanosine, several children experienced a second decrease in plasma HIV RNA levels when these same dideoxynucleosides were added to ritonavir after 12 weeks of monotherapy. These data suggest that the magnitude of the response is likely to be greater when such combination regimens are used as initial therapy, especially in antiretroviral naive infants and children. Clearly, the duration of monotherapy (if indicated at all) with new antiretroviral agents should be kept as short as possible in the future to avoid the emergence of resistance.
Despite the significant improvements in immunologic and virologic markers, we found no statistically significant positive or negative change in measures of central nervous system (CNS) function in this study population. This could be a reflection of a limited CNS penetration of ritonavir or the fact that the suppression of viral load was not optimal in this phase I/II study. The sample size is currently too small to conduct subgroup analyses to investigate whether especially younger children with CNS compromise may show differential treatment related gains.
In summary, therapy with the protease inhibitor ritonavir is feasible in children >2 years of age and results in a sustained increase in CD4 counts and a marked decrease in plasma HIV RNA levels. A limited number of children <2 years of age were enrolled, and they appeared to tolerate 250 mg/m2 and 300 mg/m2 of ritonavir; however, data for the higher dose levels are not yet available. Data in neonates and young infants (ie, <6 months of age) also are lacking. Limitation of such data reflect impediments to conducting clinical trials with experimental therapies in neonates. New regulations and a more proactive approach by clinicians, the pharmaceutics industry, and regulatory agencies are necessary to overcome such problems in the future.
Although a maximum tolerated dose has not been defined, the frequency of gastrointestinal side effects appeared to be higher at the 350 mg/m2 (equivalent to the adult recommended dose of 600 mg, bid) and 400 mg/m2 dose levels. Therefore, an additional increase in dose may not be tolerated. It is currently recommended to try to achieve as high a dose as tolerated with a maximum of 400 mg/m2 per dose. Although the 400 mg/m2 dose level appeared to be more beneficial statistically, individual children at lower dose levels achieved comparable reductions in viral load and increases in CD4 counts. Although studies combining ritonavir with other antiretroviral agents in children, as well as pharmacokinetic studies in neonates will define further the role of ritonavir in the treatment of HIV-infected children, these preliminary results, based on a phase I/II study, are promising and have led to Food and Drug Administration approval of ritonavir for pediatric use in children >2 years of age. Although encouraging, these results also underscore the need for more prompt study and the availability of appropriate formulations for newborns and young infants.
We thank the clinical support provided by the Nursing and Social Services Staff of the Pediatric Branch, NCI, and the collaborating centers, especially Christine Rosko, RN; Margie Sullivan, Anne-Marie Boler, PNP; Pam Wolters, PhD; Christine Wood, RN; Sheryl Zwerski, PNP; Gayl Selkin-Gutman, MA; Linda Lewis, MD; Stacy Shiflett, PharmD; Ellen Townley, PNP; Claire Walsek, PNP; Lori Wiener, PD; Shirley Jankelevich, MD; and Lori Perez, PD. We also thank Shalini Vora, Kimberly Mills, and Kim Mitchell for data management support; Kirsta Waldon, Dr Michael Baseler, and Dr David Waters for their laboratory support; and Richard Rode, PhD, for his critical review of the manuscript.
- Received July 21, 1997.
- Accepted October 30, 1997.
Reprint requests to (B.U.M.) Department of Medicine, Hunnewell 302, HU-215 Children's Hospital, 300 Longwood Ave, Boston, MA 02115.
This work was presented in part at the Fourth Conference on Retroviruses and Opportunistic Infections, January 22–26, 1997, in Washington, DC.
- Kempf DJ,
- Marsh KC,
- Paul DA,
- et al.
- Kempf DJ,
- Marsh KC,
- Denissen JF,
- et al.
- Perelson AS,
- Neumann AU,
- Markowitz M,
- Leonard JM,
- Ho DD
- ↵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
- Mulder J,
- McKinney N,
- Christopherson C,
- Sninsky J,
- Greenfield L,
- Kwok S
- ↵Marsh KC, Eiden E, McDonald E. Determination of ritonavir, a new protease inhibitor, in biological samples using reversed-phase high-performance liquid chromatography. J Chromatogr. 1998. In press
- Centers for Disease Control and Prevention
- Mueller BU,
- Butler KM,
- Stocker VL,
- et al.
- Kline MW,
- Fletcher CV,
- Federici ME,
- et al.
- Husson RN,
- Mueller BU,
- Farley M,
- et al.
- Mellors JW,
- Rinaldo CR,
- Gupta P,
- White RM,
- Todd JA,
- Kingsley LA
- Markowitz M,
- Mo H,
- Kempf DJ,
- et al.
- Copyright © 1998 American Academy of Pediatrics