Four drug regimen for infants

Polly Clayden, HIV i-Base

Triple drug regimens are insufficiently potent to achieve undetectable viral load in more than 50% of sick infants. Dr Gareth Tudor Williams presented results at a median of 48 weeks (and up to 108 weeks) in an oral abstract from a small cohort of 18 symptomatic infants treated with a four drug, nevirapine –containing regimen [1].

The infants presented with opportunistic infections, including PCP and CMV and had a median baseline viral load of >750, 000 copies/ml. They received a regimen of: zidovudine (360 mg/m2/day), lamivudine (8mg/kg/day), abacavir (16mg/kg/day) and nevirapine [NVP] (120 mg/m2/day, then 300mg/m2/day from Day 14 if no rash). This was given BID and the investigators reported that these are ‘…4 palatable suspensions that require no food restrictions.’

Of the group, two children developed NVP associated hepatitis (at weeks 1 and 10) and one child vomited abacavir consistently. Their treatment was changed to lopinavir/r (data excluded). Three children developed NVP associated rash, but continued treatment. There were no cases of abacavir hypersensitivity. 1 child discontinued abacavir at 14 weeks and 1 child stopped zidovudine at week 24, each due to GI intolerance, but continued treatment on 3 drugs (data included).

The 15 children achieved undetectable VL (<400 copies) by a median of 16 weeks, and all were consistently undetectable from week 48 and the median time to reach normal CD4 counts and weight for age = 48 weeks.

The investigators reported that amongst this cohort of infants with advanced disease, ‘15 of 18 children achieved sustained viral suppression and subsequent normalisation of CD4 counts and weight for age on a twice-daily, protease-sparing 4 drug combination.’ Initiating more potent therapy upfront, particularly for children with fairly advanced symptoms may well achieve more satisfactory results than standard triple drug combinations.

Abacavir clearance

Dose finding in children is notoriously difficult to achieve. An oral poster presentation evaluated whether defining abacavir’s (ABC) metabolic profile may offer insights into known differences in dose requirements and pharmacokinetics between children and adults [2]. Additionally the authors speculated that variable metabolite formation could be relevant for patients at risk for ABC hypersensitivity reactions.

ABC systemic clearance (CLabc) is primarily determined by the extent of formation of glucuronide (GLU) and carboxylate (CAR) metabolites.

The pharmacokinetics of ABC, GLU, and CAR were evaluated after administration of a dose of 8 mg/kg of ABC given to HIV infected children stratified by both Tanner stage and gender. Plasma concentrations (n=9) of ABC, GLU, and CAR were determined by HPLC. A pharmacokinetic model for parent and metabolites was fit to each patients data and a metabolite phenotype was defined by ratio of GLU and CAR to ABC.

The investigators reported that abacavir systematic clearance is highly variable (CLabc varied > 5 fold with median (range) of 523 (245-1360) ml/min/m2), and found glucuronidation to be an important determinant. A single blood sample is informative for glucuronidation phenotype and distinguishes among patients with low or high CLABC. CAR metabolite profiles were less informative but further studies may identify subjects with important differences in this metabolic pathway. Genotypic studies in progress will clarify the basis for the phenotypic differences in CLABC. They concluded that ‘ABC glucuronidation phenotype warrants further study to identify patients with high CLABC and patients who experience ABC hypersensitivity reactions.’


Adherence has been shown repeatedly to be a strong determinant of treatment success but there are few studies of adherence to highly active antiretroviral therapy (HAART) in children.

In an oral presentation by Dr Di Gibb she reported results from adherence questionnaires completed at 4, 12, 24 and 48 weeks by carers of children participating in the PENTA 5 trial (PENTA 5 was designed to evaluate different dual nucleoside reverse transcriptase inhibitor (NRTI) therapy combinations with and without the protease inhibitor (PI) nelfinavir (NFV)) [3].

The questionnaires included questions about the number of doses missed in previous 7 days, about difficulties taking individual drugs, which doses were hard to remember and how HAART interfered with everyday life. 266 questionaires were returned, at least one for 108 (84%) of children participating in the trial.

NFV was reported to be the most difficult drug to take – reported in 38% of questionnaires, but this decreased over time, p=0.02. Difficulties in taking and remembering drugs reported were related to fear of disclosure and to unpleasant characteristics of the drugs – taste or smell of tablets or liquid and pill size.

Full adherence was reported in 74% of questionnaires which did not change over time, and was reported more frequently in older children (>10 years) and those with symptomatic HIV. More children reporting full adherence achieved HIV RNA <400 copies/ml – at 24 weeks 75%/48 weeks 79% vs 50%/50% of children reported to be not fully adherent: overall p=0.01). These results are similar to those reported in adults.

The authors concluded that ‘Relatively high adherence levels were reported and were significantly associated with virological response. Social factors were important in explaining non adherence. Further research including randomised studies are required to evaluate ways of improving adherence in HIV infected children.’

Carnitine deficiency and HAART

Carnitine deficiency appears to be associated with antiretroviral use particularly zidovudine (ZDV), unlike adults there are few data reporting on carnitine status in children. A poster from Dr Claudia Fortuny and colleagues looked at possible carnitine deficiency in children and tried to find a relationship between serum carnitine, it’s amino acid precursors, nutritional status and antiretroviral therapy [4].

79 HIV-positive children were monitored. Reference values were established in this pediatric population by measurement of carnitine profile (N=31) and amino acids (N=88) in apparently healthy children who underwent pre-operative laboratory tests for minor surgery.

The investigators reported that serum free and total carnitine, acylcarnitines, methionine and lysine were significantly lower in HIV-positive children compared with the reference values for similar ages (p<0.0001 and carnitine deficiency was observed in 37% of HIV-positive children. No relationship was observed between serum carnitine and severity of symptoms, immunological or nutritional status or lipodystrophy. Free and total carnitine were significantly lower (p=0.002 and p=0.033, respectively) in HIV-infected patients on protease inhibitors (N=56) compared with those on other treatments (N= 23). 80% of HIV-infected children with carnitine deficiency were on protease inhibitors therapy.

They concluded that ‘Carnitine deficiency is frequent in HIV-infected children. Malabsorption or defective synthesis may account for the carnitine deficiency detected in these patients.’ And also noted that ‘The identification of the causes of carnitine deficiency in HIV-infected children are complex. And most probably a combination of multiple factors are involved. Additionally ‘Carnitine measurement seems advisable in HIV-infected children so as to be able to normalize decreased serum and probably tissue levels.’

Treatment interruptions and CD4 decline

Adult data reports a median CD4 decline of approximately 20 cells per month during a structured treatment interruption (STI) [5]. An analysis from the Collaborative HIV Paediatric Surveillance (CHIPS) study evaluated CD4 decline in HIV-positive children stopping HAART therapy for at least one month, after at least 3 months before stopping. CHIPS is a cohort of HIV positive children from 14 centres in UK and Ireland under follow-up from 1996. The investigators emphasized that these were unstructured treatment interruptions.

Data on clinical events, T cell subsets, HIV RNA viral load and ART history were collected on HIV-infected children from 14 centres in the UK and Ireland participating in the CHIPS study (n=43) and 3 children from Rotterdam .

The slope of decline of CD4 cell count and CD4% per month off HAART, adjusting for age, was calculated and compared to changes within the 6 months preceding interruption.

The most common reasons for interruption were ‘request of parents’ (n=10) and poor adherence (n=7). The same HAART was restarted after 11 interruptions, was changed after 18, and 8 children remained off HAART. CD4 count and CD4% were stable prior to interruption and only 11 children had HIV RNA <10,000 copies/ml (of whom 3 <400 copies/ml). Only 6 children were fully suppressed before stopping. During 22 interruptions with sufficient data, average CD4 decline was 18 cells/mm3/month – a median fall similar to that of adults.

The authors concluded that ‘CD4 cell decline rate varied considerably following interruption of HAART in this group of children with high viral load before interruption (possibly implying poor adherence). An approach to paediatric STI trials may be to base the length of STI on the rate of CD4 decline rather than having fixed STI periods.’

CD4 and HIV RNA response following HAART

Another CHIPS evaluation looked at predictors of CD4% and HIV RNA change 6 months after initiation of HAART (taking 3 or more drugs with at least a PI, NNRTI or ABC) [6]. Adult data has suggested that CD4 response to HAART may be relatively independent of HIV RNA or CD4 count at HAART initiation. This has been poorly evaluated in children. In this study the investigators aimed to describe predictors of reponse to HAART in children in clinical practice.

HAART was defined as a regimen containing 3 or more drugs of >=2 classes or ABC. Of 627 children 178 were drug naïve at initiation of HAART and had CD4 and HIV-1 RNA values up to 6 months before and 6 months after HAART initiation. The median age was 4.1 years, median CD4% was 16% (CD4 z-score 3.4) and median HIV-1 RNA was 5.3 log10 copies/ml.

Of the regimens used 51% were PI-containing, 39% NNRTI-containing, 19% PI and NNRTI and 8% ABC. NFV was the most commonly used PI and NVP the most common NNRTI.

At 6 months after initiation of therapy 154 (87%) children had higher CD4 than baseline. There was a median increase of 9% CD4 and a median decrease of HIV-1 RNA of 2.6 log. 102 (57%) children were <400 copies.

Adjusting for confounders CD4 rise >10% was inversely correlated with baseline CD4 and age at initiation of HAART. There were no significant additional effects of AIDS status or HIV RNA at start of HAART, calendar year of starting HAART or type of HAART, on CD4 or RNA response. The investigators noted that higher odds may be associated with 4-drug regimens but the number of children taking these was small (n=9)

The investigators concluded that children respond well to HAART irrespective of baseline HIV RNA or clinical status. Adherence may be poorer in younger and asymptomatic children with higher CD4%, whereas older children may be less able to reconstitute their immune system.’

Lactic acidosis in newborns

Over recent years in the industrialised world strategies to reduce mother to child transmission (MTCT) have virtually eliminated this mode of transmission. There is however concern that feotal exposure to nucleoside analogues may cause lactic acidosis in some newborn infants.

A poster from Noguera and colleagues clinical symptoms of mitochondrial dysfunction were analysed in a group of 78 children (40 girls and 38 boys) born to HIV-infected women during 24 months (January 2000 – September 2001) [7]. None of the children were HIV positive.

Most of the mothers had received zidovudine (ZDV) in different regimens during pregnancy, 17 mothers received therapies without ZDV and 4 mothers were untreated. Most of the women also received ZDV alone or combined with other antiretrovirals during labour; five mothers received no treatment. All newborns were treated with ZDV. Hyperlactatemia with hyperalaninemia was detected in 39 (50%) children, accompanied with mild acidosis in three cases. Only 3 infants have shown axial hypotonia with a slight delay in psychomotor development.These symptoms did not persist in all cases by the age of six months. In 18 patients, lactate plasma levels have progressed spontaneously to normality within the first six months of life.

The authors concluded that ‘Half of the children exposed in utero or in the neonatal period to NRTIs develop hyperlactatemia, normally benign and self-limited.’ And recommended that ‘A close follow-up of these children is required.’

Metabolic changes

A number of posters looked at lipodystrophy, bone disorders and metabolic changes in children and one poster evaluated the use of bone ultrasound to evaluate bone strength in children.

Dr Brambilla and colleagues evaluated the correlation between metabolic abnormalities, fat mass and visceral adiposity in HAART-treated HIV-positive children [8]. They assessed a group of 37 receiving d4T/3TC+PI, between the ages of 6 and 18 (mean age 12.2 yrs) of which 8 had and 29 had no clinical signs of lipodystrophy. The children were were assessed for: 1 lipids, glucose, insulin levels; 2 fat content and distribution (by DXA); 3 visceral fat content (by lumbar MRI). As a group the children had an excellent suppression rate – 35/37 HIV-RNA<50 copies/ml

The investigators reported hypertriglyceridemia present in 19 (51%), hypercholesterolemia in 19 (51 %), elevated LDL cholesterol in 7 (19%), elevated insulin levels in 5 (18%). Triglycerides and insulin levels were more elevated in children exhibiting clinical signs of lipodystrophy than in those without (p=0.009 and p=0.01 respectively). Children with clinical signs of lipodystrophy had higher trunk fat (5.7 + 2.4 vs 3.7 + 2.3 kg, p 0.04) and higher visceral fat [86 (34) vs 31 (16) cm2, p= 0.0001], but similar total body fat percent [22 (6) vs 17 (8), p=0.08] when compared to those without. They also found insulin levels were significantly correlated with total body fat (p=0.0009) and with trunk fat (p=0.0006) in all cases; only children with signs of lipodystrophy showed a significant correlation between visceral fat and insulin.

They concluded that ‘HAART treated children show frequent metabolic abnormalities. Fat mass is the best predictor for hyperinsulinism and insulin resistance. Visceral fat, very high in children with clinical signs of lipodystrophy seems to be the major determinant for impaired insulin metabolism.’

Another poster from the same group reported by Alessandra Vigano and colleagues assessed whether progressive harmful changes in body composition are associated with increased exposure to highly active antiretroviral therapy in HIV-infected children [9].

Risk of clinically assessed lipodystrophy appears to increase with the duration of exposure to HAART in HIV-infected adults. This study aimed to assess longitudinal body composition in HIV-positive children receiving HAART using DXA and magnetic resonance imaging (MRI) to estimate regional body composition and visceral fat content (VF).

Thirty seven children as described above were enrolled in the study. DXA scans were performed in all of them at study entry and after 12 months. MRI scans were performed in 14/37 at study entry and after 12 months.

At study entry and at 12 months follow-up mean HAART exposure was 39.3 (4.1) and 50.9 (6.7) months, number of children with clinical signs of lipodystrophy increased from 6 to 8, whereas mean body mass index CD4% and % of children with HIV RNA< 50 copies/ml (100vs 97) and diet remained unchanged. DXA scans showed that: Lean mass increased in all HIV positive children according to physiological growth pattern for age; Fat mass did not change significantly and consequently fat % decreased significantly (p=0.04); Fat distribution changed showing a significant increase at trunk level (p=0.006) with a slight decrease in arms; therefore limbs/trunk fat ratio was significantly reduced (p 0.0001) as well as arms fat/lean ratio (p= 0.0003). All changes were observed independently from clinical signs of lipodystrophy. In children with lipodystrophy, increase in trunk fat was more pronounced. MRI scans showed an increase in VF in HIV+ with as well as in HIV+ without lipodystrophy.

The investigators reported that abnormal fat mass deposition occurs frequently and are a progressive phenomenon in HIV-infected children receiving HAART. They conclude that ‘The prevention as well as an effective and safe treatment of such adverse events need to be urgently considered’ and that ‘These changes of body composition expose children to a high risk for severe psychological repercussions and metabolic abnormalities.’

In a poster presentation Jim Oleske’s group analysed bone disorders in HIV-infected and -exposed children using data from PACTG 219 (this is a long-term study, since 1991, to examine outcomes in 2,695 HIV-positive children and 1,352 HIV negative children exposed to HIV and antiretroviral drugs) [10]. The database was examined for reports of osteopenia, osteonecrosis and any other bone disorders. 38 HIV-positive children were diagnosed with bone disorders.

Fractures were the most common diagnosis, seen in 33 (30 infected, 3 uninfected)/82 (40%), 39% of whom had comorbid conditions such as growth or developmental delay. Three children had rickets (2 infected, 1 uninfected). The 2 HIV infected children with rickets had comorbid malabsorption syndromes. The following conditions were only reported in infected children: osteomyelitis, osteonecrosis and osteopenia. Osteomyelitis was reported in 22 children, 9 female and 13 male, 8 of whom had comorbid conditions. Osteonecrosis was reported in 7 children, 5 male, 2 female involving the hip in 6 and the humerus in 1. Osteopenia was reported in 4 children, 1 girl and 3 boys; 2 had bone fractures as well, and 3 had extensive comorbid conditions, including congential toxoplasmosis, active CMV, and renal disease.

The investigators found that bone disorders were seen in 2.9% of infected children followed in PACTG 219 as opposed to 0.07% HIV-exposed but uninfected children. They recommend that ‘HIV-infected children need to be monitored for bone disorders in order to determine the role of disease or adverse treatment effects on bone health’.

Stephen Arpadi and colleagues looked at the relationship between bone mineral accrual in HIV-positive children with or without lipodystrophy [11].

This relationship had not been previously assessed using objective criteria for lipodystrophy and longitudinal measurements of bone. The study objective was to compare accrual of total body bone mineral content (TBBMC) in HIV positive children using antiretrovirals with and without lipodystrophy (LD).

They found that eight children (29%) were classified as LD+. The mean unadjusted TBBMC was lower in LD+ children at both visits 1 and 2 but these differences were not significant. No significant differences in bone mineral accrual from visit 1 to visit 2 were observed At both visits, the mean unadjusted total bone density (TBD) was lower in LD+ versus LD- children but the differences were not significant The differences in log TBBMC and log TBD remained insignificant after adjusting for age, sex, race, height, and weight.

The investigators found that ‘TBBMC accrual does not differ in HIV+children with and without LD. Using objective measurements and criteria are necessary to determine the relationships between LD and bone abnormalities and associated risk factors’.

Finally a poster from Drs Desai and Mathur perfomed an assessment of bone strength in children with HIV on HAART using a novel non-invasive bone ultrasound

Bone density is generally studied using complicated procedures such as DXA or indirectly assessed using hormone levels. The authors presented the use of a novel Quantitative ultrasound (US) as a simple, non-invasive method of assessing bone strength.

58 children age 3-17 years had bone ultrasounds performed. 38/58 (65.5%) had bone density below the 50th percentile, 26 (45%) below the 25th, 18 (31%) below the 10th percentile and 12 (21%) below the 5th percentile. 29 patients are on one & 18 on two protease inhibitors and 11 on no PIs. There was no correlation between bone strength & HIV viral load, CD4 count or PI therapy. Patient & family acceptance of bone US was high. Nutritional counseling was reinforced based on US results.

They concluded ‘Decreased bone strength is common in children on HAART, with a significant percentage having bone strength below the 10th percentile for age and sex. Bone ultrasound is a reliable, non-invasive & easily performed bedside test of bone strength and should be incorporated in the routine assessment of patients on HAART.’


All references from the Program and Abstracts XIV International AIDS Conference Barcelona 2002.

  1. Tudor-Williams G, Head S, Weigel R et al Baby Cocktail! A protease-sparing 4 drug combination for symptomatic infants. Abstract MoOrB1129
  2. Rodman J H, Cross S J, Dangelo L J et al. Abacavir systemic clearance in children is highly influenced by glucuronidation phenotype Abstract MoPpB2010
  3. Giacomet V, Gibb D M, Goodall R et al. Adherence to HAART in children: results from a questionnaire study of children in PENTA 5 trial. Abstract TuPpB2050
  4. Fortuny C, Vilaseca M A, Artuch R et al. Carnitine deficiency in HIV-infected children on antiretroviral treatment. Abstract TuPeB4533.
  5. Leclezio V, DuongT, McGee L et al. Treatment interruptions of HAART in paediatric HIV: effect on CD4 count. Abstract TuPeB4630
  6. Walker A S, Gibb D M, McGee L et al. CD4 and HIV RNA response following HAART in ART naive children. Abstract TuPeB4631
  7. Noguera A, Perez-Duenas B, Tello L M et al. Lactic acidosis in newborns exposed to HIV and antiretrovirals. Abstract WePeB5951
  8. Brambilla P, Mora S, Monti L et al. Correlation between metabolic abnormalities, fat mass and visceral adiposity in HAART-treated HIV-infected children. Abstract TuPeB4512
  9. Vigano A, Mora S, Beccio S et al. Progressive harmful changes in body composition are associated with increased exposure to highly active antiretroviral therapy in HIV-infected children. Abstract TuPeB4647
  10. Oleske J M, Gaughan D M, Mofenson L M et al. Bone disorders in HIV-infected and -exposed children. Abstract TuPpB2053
  11. Arpadi S M, Horlick M, Verret W et al. Bone mineral accrual does not differ in HIV-infected children with and with out lipodystrophy. Abstract ThPeB7233
  12. Desai N, Mathur M. Assessment of bone strength in children with HIV/AIDS on HAART using a novel non-invasive bone ultrasound. Abstract TuPeB4638

Links to other websites are current at date of posting but not maintained.