Several studies have recommended starting and maintaining antiretoviral therapy in HIV-infected patients with three drugs[1-4]. Treatment in adult and paediatric patients older than 6 years should be initiated when patients enter category B or category 2 according to the Centers for Disease Control (CDC) classification  or when patients in category A1 show repeated plasma HIV RNA > 10 000-20 000 copies/ml[2,6,7]. Whereas much therapeutic experience in adult patients has been published, until now there are almost no clinical data of combination therapy including a protease inhibitor in treatment-naive children.
In phase I/II studies, safety, tolerance and pharmacokinetic profiles of the protease inhibitors indinavir, ritonavir and nelfinavir were assessed in HIV-infected children[8-10]. Main side-effects were gastro-intestinal problems, such as vomiting and abdominal pain (ritonavir), renal complications, such as interstitial nephritis (indinavir) and diarrhoea (nelfinavir). Because of intolerance and non-compliance a high percentage of paediatric patients on ritonavir and indinavir discontinued therapy[9,11,12], whereas nelfinavir appears to be tolerated better by children[10,13].
Studies with pretreated HIV-infected children demonstrated a viral load reduction of 0.9-1.5 log10 after 12 months on triple therapy with indinavir or ritonavir[14-16]. Nineteen out of 50 pretreated children (38%) receiving highly active antiretroviral therapy with ritonavir showed RNA HIV copy numbers < 1000/ml after 3 months‚ treatment. Good therapeutic effects were seen in treatment-naive, adult patients on therapy with two nucleoside reverse transcriptase inhibitors (NRTI) and nelfinavir (2250 mg three times a day). After 6 and 12 months of treatment, mean reduction of viral load was 2.3 and 2.9 log10 in patients on triple therapy respectively and 1.3 log10 in patients on two NRTI plus placebo. Viral load was below detection (< 500 copies/ml) in 81% of nelfinavir-treated patients after 6 months and 76% after 12 months of treatment.
Another therapeutic strategy includes a non-nucleoside reverse transcriptase inhibitor (NNRTI) in combination with NRTI with or without protease inhibitors. Luzuriaga et al.  demonstrated a median reduction of 2.9 log10 in 12 small children (age < 25 months et entry) after 12 weeks of treatment with a combination of NRTI and nevirapin. First data of NNRTI- and protease inhibitor-naive, but NRTI-experienced paediatric patients have shown a reduction in HIV RNA of < 500 copies/ml in 63% of all children on a combination of nelfinavir and efavirenz for a period of 20 weeks. Other paediatric studies with nelfinavir/ NRTI combinations versus treatment with placebo/ NRTI combinations are ongoing (PENTA 5). We report our preliminary results of using triple antiretroviral therapy that includes nelfinavir in paediatric patients.
Patients and method
In an open prospective study, 16 out of 20 perinatally HIV-infected children (eight female and eight male) in CDC category A or B2 received antiretroviral triple therapy. Informed consent was obtained from the parents or legal guardians. Four out of 20 patients were not included because their parents refused to start with one of the recommended therapeutic regimens. Data were evaluated by an intent-to-treat analysis. None of the children had prior antiretroviral treatment. According to American and German paediatric guidelines [7,20] antiretroviral therapy was started when patients entered category B or 2 or when patients in category A1 had a plasma HIV RNA > 20 000 copies/ml for children > 2 years of age and > 100 000 copies/ml during the first 2 years of life on at least two separate samples within 3 months. Patients were not randomized to therapeutic regimens. Combinations were chosen after presenting the four NRTI to parents and children and after test doses were given. One patient was put on a zidovudine/lamivudine/nelfinavir combination because of hepatitis B co-infection. Eight patients in group I received zidovudine together with lamivudine and nelfinavir, and eight patients in group II received stavudine, didanosine and nelfinavir.
Dosage of antiretroviral agents
In group I dosages were: nelfinavir, 60-90 mg/kg body weight three time a day (taken with food); zidovudine, 12 mg/kg body weight three times a day; lamivudine, 8 mg/kg body weight three times a day. In group II dosages were: didanosine, 180 mg/m2 (taken 1 hour before food) twice a day; stavudine, 2 mg/kg twice a day.
Adherence to therapy was questioned at every visit. Compliance was considered poor when more than 10% of the dose was missed. Non-compliance was also defined as repeated insufficient nelfinavir levels despite adequate prescribed nelfinavir dosage 8 weeks after the therapeutic regime had been initiated.
As shown in Table 1, eight children were in category A and eight children in category B; five children (all in category B) had CD4 cell counts < 200 × 106/l or below 15% of lymphocytes and were classified as category 3 patients. CD4 cell counts of all 16 patients were in the range 15-2250 × 106/l (median, 656 × 106) before treatment was initiated. Before onset of treatment viral load of all patients was in the range 36 000-5 000 000 copies/ml (median, 203 000 copies/ml). Median age at study onset was 77 months (range, 14-152 months), three patients were younger than 30 months of age (Table 1).
The most common symptoms of patients in category A were lymphadenopathy and recurrent upper respiratory infections or otitis media (six patients); hepatosplenomegaly was less common (two patients). Three patients in category B had severe bacterial infections such as pneumonia or meningitis, two patients had a persisting thrombocytopenia, in each of whom herpes zoster infection, recurrent diarrhea or lymphoid interstitial pneumonitis (LIP) was seen.
Viral load was measured by the Amplicor test (version 2; Hoffman LaRoche, Grenzach Whylen, Germany) and was expressed in HIV-1 RNA copies/ml plasma (limit of detection < 500 copies/ml, later on < 50 copies/ml, after availability of Amplicor test, version 1.5). CD4 cell count was measured by flow cytometry (monoclonal antibodies from Becton-Dickinson, Heidelberg, Germany), expressed in cells/l plasma, and was measured every 4-8 weeks. SGOT, SGPT, triglycerides and cholesterol were monitored every 4-8 weeks to record protease inhibitor-related side-effects.
Plasma nelfinavir level
Quantification of nelfinavir in plasma was carried out by tandem mass spectrometry, in conjunction with high-speed on-line high-performance liquid chromatography (Dr M. Kurowski, Berlin, Germany). Evaluation was performed according to a recently described method. Maximal plasma nelfinavir levels (Cmax) were measured 2-4 h after morning drug administration. As shown by Shetty et al.  in adult patients sufficient Cmax values range from 3000 to 4000 ng/ml after giving 2250 mg nelfinavir three times a day. In patients with Cmax values < 3000 ng/ml Cmin values were also measured; in these children nelfinavir dosages were inreased to achieve Cmin values > 1000 ng/ml and Cmax values > 3000 ng/ml. Plasma nelfinavir levels were assessed once in all patients at study onset and repeated in patients with increasing viral load without informing patients and parents.
Comparison between surrogate markers of both groups was made with the Wilcoxon rank sum test. HIV-1 RNA copies and CD4 cell count in groups I and II were evaluated and compared before onset and after 3, 6, 9 and 12 months of treatment.
Lymphadenopathy and hepatosplenomegaly improved or disappeared in seven out of eight category A patients after starting therapy. Patients with inital recurrent upper respiratory infections or otitis media showed a decreased infection rate. Two out of three patients in category B3 with severe bacterial infections at the start of treatment had no further infections when the CD4 cell count increased to > 200 × 106cells/l or 15% of lymphocytes. In one patient with repeated episodes of pneumonia under triple therapy a gastro-oesophageal reflux was found. Herpes zoster infection, recurrent diarrhea, manifestation of LIP or opportunistic infections such as Pneumocystis carinii pneumonia (PCP) were not seen in category B patients during the 12 months of treatment.
No significant differences in viral load were found between the two treatment groups at any of the time-points investigated (P > 0.1). Therefore results of viral load reduction were considered together for all 16 patients. The median viral load in both groups was 5.3 log10 (203 000 copies/ml; range, 36 000-5 000 000 copies/ml) 1 month before treatment (Table 2, Fig. 1). After 3 months of treatment, the median reduction of viral load was 2.5 log10, after 6 months it was 2.9 log10 and after 12 months it was 2.8 log10. Viral load ranged between < 500 and 15 000 copies/ml (median, 500 copies/ml) in the third month and between < 50 and < 80 000 copies/ml (median, 350 and 250) after 6 and 9 months of treatment respectively (Table 2, Fig. 1). Because of non-compliance one patient discontinued therapy and was scored as a non-responder. After 12 months of treatment viral load ranged between < 50 and 21 000 copies/ml (median, 270 copies/ml or 2.5 log10). HIV-1 RNA < 500 copies/ml was found in three of 16 patients (18%) after 3 months, in 12 patients after 6 months (75%) and in 11 patients after 9 and 12 months (69%). Five patients on triple therapy had > 5000 HIV-1 RNA copies/ml, two of them were category A1 patients, two were category B1 and one was a category B3 patient. Before beginning treatment the median viral load of these patients was 1 200 000 copies/ml (range, 70 000-5 000 000 copies/ml). Using the ultra sensitive assay, seven out of 16 patients (44%) had a viral load < 50 copies/ml after 12 months of treatment.
CD4 cell count
No significant differences were observed between the treatment groups in terms of increased CD4 cell counts at any time of observation (P > 0.1); therefore results were considered together. Before the beginning of therapy the median CD4 cell count of all patients was 656 × 106 cells/l (range, 15-2250 × 106 cells/l) which increased to 850 × 106 cells/l (range, 56-3050 × 106cells/l) after 3 months and decreased to 813 × 106 cells/l after 12 months of treatment (Table 3). In five patients (category B3) median CD4 cell count increased significantly (P < 0.01) from 35 × 106 cells/l (range, 15-205 × 106 cells/l) to 267 × 106 cells/l (range, 56-397 × 106 cells/l) after 3 months and was maintained at 654 × 106 cells/l after 12 months of treatment (range, 482-1450 × 106 cells/l). In 11 patients (category A1 or B1) median CD4 cell count increased from 850 × 106 cells/l (range, 567-2250 × 106 cells/l) to 1240 × 106 cells/l (range, 56-397 × 106 cells/l) within 3 months and decreased to 956 × 106 cells/l after 12 months of treatment (range, 482-1450 × 106/l).
Plasma nelfinavir level
Maximal plasma nelfinavir level (Cmax) was measured 2-4 hours after administration. Cmax values ranged from 1798 to 7900 ng/ml (median, 4027 ng/ml). Because of a nelfinavir level below 3000 ng/ml in three patients (1789, 2205 and 2396 ng/ml), nelfinavir Cmin was measured: two of these patients had nelfinavir < 1000 ng/ml. The nelfinavir dosage was then increased from 60 to 90 mg/kg body weight per day.
In general, triple therapy was well tolerated. During the initial phase of treatment, diarrhoea, lack of concentration and rash were the most common side-effects in patients under both therapeutic regimens. Transient or persisting elevation of serum triglycerides (maximum 420 mg/dl) were detected in eight out of 16 patients and elevated cholesterol values (maximum 308 mg/dl) in five out of 16 patients after starting nelfinavir. However, peripheral lipodystrophy or elevation of liver enzymes were not observed.
Administration of nelfinavir and didanosine was difficult in the whole paediatric group, especially in small children. In order to achieve compliance, an appropriate formula and individual setting had to be found during the first weeks of treatment. Initially 12 patients refused drug ingestion because of the unpleasant taste and were therefore switched from nelfinavir powder to tablets (dissolved in liquid). Because of the large tablet size, six of the eight patients were switched from didanosine tablets to powder (dissolved in liquid).
To our knowledge this is the first report of preliminary experience with a nelfinavir-based triple therapy in antiretroviral therapy-naive HIV-infected children. Our data show a profound and lasting antiretroviral effect in most of the patients studied. No significant differences were seen in terms of CD4 cell count increase or viral load reduction at any time of treatment between the therapeutic regimens used in groups I and II.
In a group of infants < 25 months of age, Luzuriaga et al.  demonstrated a viral load reduction of 2.9 log10 after 12 weeks of treatment with NNRTI and NRTI. Starr et al.  found a reduction of HIV RNA of < 500 copies/ml in 63% of NRTI-experienced paediatric patients on combination therapy with nelfinavir, efavirenz and NRTI over a period of 20 weeks. Although it is unknown whether NNRTI show such a lasting effect as that seen in the present study of children taking nelfinavir, our data appear to be comparable with these short-term results. Our results were also confirmed by clinical studies in treatment-naive, adult HIV-infected patients on triple therapy with nelfinavir; there was a mean reduction in viral load of 2.9 log10 after 12 months of treatment. In these patients viral load was below the limit of detection (< 500 copies/ml) in 76% after 12 months of treatment.
Children in categories 1 and 2 had a moderate increase in median CD4 cell count after 12 months of treatment (from 850 to 956 × 106 cells/l), whereas patients with an initial CD4 cell count < 15% (CDC category 3) showed a greater benefit in both treatment groups, increasing from 35 to 654 × 106 cells/l (median values) during the treatment period. PCP prophylaxis was stopped in these children because they changed from category 3 to category 2. Various studies have demonstrated a high recovery rate of naive (CD45RA), memory (CD45RO) and total CD4 T lymphocytes especially in small children, and a recovery of lymph node architecture in patients on highly active antiretroviral therapy[23,24]. Despite these results it remains to be shown whether the increase in CD4 cell count under triple therapy means a restoration of immunological responsiveness.
Evaluation of plasma nelfinavir levels showed Cmax values of 3000-4000 ng/ml in most children, indicating therapeutic plasma levels. In two children low Cmax and low Cmin values (< 3000 ng/ml and < 1000 ng/ml) were found and nelfinavir dosages were increased in order to achieve sufficient plasma levels. These results demonstrated a good absorption of the drug, although nelfinavir tablets were dissolved in liquid.
Besides transient elevation of serum triglycerides and mild diarrhoea, both regimens were well tolerated by most children. Most difficulties arise in the beginning of treatment, explaining the moderate reduction in viral load during the first month. Especially in small children it was important to find the right formula and an optimum form of administration. A smaller size and fewer tablets could ease administration and increase compliance. Pharmacokinetic studies in adults, and recently also in children, have demonstrated sufficient plasma levels and similar supression of viral replication when nelfinavir was given three times a day or twice a day[25-27]. Taking nelfinavir in two instead of three doses per day would facilitate the therapeutic regimen for most of these children.
Our data show a stronger and a more sustained reduction of viral load in children on triple therapy with nelfinavir when compared with data combining two nucleoside analogues[17,28]. In our opinion antiretroviral therapy should be started in both adult and paediatric patients as a triple therapy. Dual therapy may provoke therapeutic failure because of early emergence of resistant HIV variants and leads to higher disease progression rate, as shown for patients with CD4 cell count < 200 × 106/l. As this study cohort is small, data from larger trials such as the ongoing PENTA 5 study are urgently needed. Despite national recommendations the right moment to start or change the therapeutic regimen is still a matter of debate. Levy  saw a great danger in ‚using up‚ all antiretroviral drugs and then having no further options when treatment is started too early. Therefore he recommended reservation of current therapeutic regimens for those patients with symptoms or whose CD4 cell counts have declined substantially. Present studies focus mainly on the short-term effects of antiretroviral therapy[31,32]. Long-term effects, disease progression and emergence of (cross)-resistance in patients with early versus late onset of triple therapy including a protease inhibitor needs to be studied in prospective clinical trials. Particularly for children, an easy to manage and well-tolerated therapeutic regimen is of great importance to achieve good compliance over a long treatment period.
Triple therapy wih nelfinavir appears to meet these criteria at least in part. Our data show clinical benefit, viral load reduction in both category A and category B patients and moderate adverse reactions. Nevertheless, further studies are needed to determine the optimal first-line combination and the right moment to initiate and change therapy in HIV-infected children.
1. Bowersox J. ACTG 343: three drugs better than two for maintaining HIV supression.
NIAID AIDS agenda 1998; p 1-2: 10-1.
2. Carpenter CC, Fischl MA, Hammer SM, et al. Antiretroviral therapy for HIV infection in 1998: updated recommendations of the International AIDS Society - USA Panel. JAMA
3. Gazzard B, Moyle G. 1998 revision to the British HIV Association guidlines for antiretroviral treatment of HIV seropositive individuals. BHIVA Guidlines Writing Committee. Lancet
4. Havlir DV, Marschner JC, Hirsch MS, et al. Maintenance antiretroviral therapies in HIV-infected subjects with undetectable HIV RNA after triple therapy. N Engl J Med
5. Centres for Disease Control and Prevention. 1994 revised classification system for human immunodeficiency virus infection in children less than 13 years of age. MMWR
6. Centers for Disease Control and Prevention. Guidelines for the use of antiretroviral agents in pediatric HIV infection. MMWR
7. Working Group on Antiretroviral Therapy and Medical Management of Infants, Children and Adolescents with HIV Infection. Antiretoviral therapy and medical management of pediatric HIV infection. Pediatrics
8. Mueller BU, Sleasman J, Nelson RP, et al. A phase I/ II study of protease inhibitor ritonavir in children with HIV infection. Pediatrics
9. Melvin AJ, Mohan KM, Arcuino LA, et al. Clinical, virologic and immunologic responses of children with advanced HIV 1 disease treated with protease inhibitors. Pediatr Infect Dis J
10. Krogstad P, Wiznia A, Luzuriaga K, et al. Phase I study of HIV-protease inhibitor nelfinavir mesylate in HIV+ children. Fourth Conference on Retroviruses and Opportunistic Infections.
Washington DC, January 1997 [abstract 721].
11. Rutstein RM, Feingold A, Meislich D, Word B, Rudy B. Protease inhibitor therapy in children with perinatally acquired HIV infection. AIDS
12. Watson DC, Farley JJ, Lovelace S, et al. Efficacy and adherence to highly active antiretroviral therapy (HAART) in HIV-1 infected children. Fifth Conference on Retroviruses and Opportunistic Infections.
Chicago, February 1998 [abstract 230].
13. Funk M, Linde R, Wintergerst U, et al. Viral load and CD4-cell count under triple therapy with nelfinavir and two NRTIs in previously untreated HIV-infected children. XII World AIDS Conference
. Geneva, June 1998 [abstract 12 256].
14. Kline MW, Flechter CV, Harris AT, et al. One year follow-up of HIV-infected children receiving combination therapy with indinavir, stavudine and didanosine. Fifth Conference on Retroviruses and Opportunistic Infections.
Chicago, February 1998 [abstract 232].
15. Pelton SI, Yogev R, Johnson D, et al. Changes of viral load and immunologic markers in children treated with triple therapy including ritonavir. Fifth Conference on Retroviruses and Opportunistic Infections.
Chicago, February 1998 [abstract 234].
16. Wintergerst U, Hoffmann F, Solder B, et al. Comparison of two antiretroviral triple combinations including the protease inhibitor indinavir in children infected with human immunodeficiency virus. Pediatr Infect Dis J
17. Clendeninn N, Quart B, Anderson R, et al. Study 511: Analysis of long-term virologic data from nelfinavir (NFV) 511 protocol using three HIV-RNA assays. Fifth Conference on Retroviruses and Opportunistic Infections.
Chicago, February 1998 [abstract 372].
18. Luzuriaga K, Wu H, McManus M, et al. Dynamics of human immunodeficiency virus type 1 replication in vertically infected infants
. J Virol
19. Starr SE, Fletcher CV, Spector SA, et al. Efavirenz in combination with nelfinavir and nucleoside reverse transcriptase inhibitors (NRTIs) is safe and virologically effective in HIV-infected children. 38th Interscience Conference on Antimicrobial Agents and Chemotherapy.
San Diego, September 1998 [abstract LB-6].
20. Wahn V, Bialek R, Boehler T, et al. Empfehlungen zur antiretroviralen Therapie bei HIV-infizierten Kindern. Monatsschr Kinderheilkd
21. Knebel NG, Sharp SR, Madigan MJ. Quantification of the Anti-HIV drug saquinavir by high-speed on-line high-performance liquid chromatography/ tandem mass spectrometry. J Mass Spectrometry
22. Shetty BV, Kosa MB, Khalil DA, et al. Preclinical pharmacokinetics and distribution to tissue of AG 1343, an inhibitor of human immunodeficiency virus type 1 protease. Antimicrob Agents Chemother
23. Cohen Stuart JW, Slieker WA, Rijkers GT, et al. Early recovery of CD4+ T lymphocytes in children on highly active antiretroviral therapy. AIDS
24. Visser CJ, Cohen Stuart JW, Dejong D, et al. Lymph node reconstitution of HIV patients after four weeks. 38th Interscience Conference on Antimicrobial Agents and Chemotherapy.
San Diego, September 1998 [abstract I-50].
25. Johson M, Petersen A, Winslade J, Clendeninn N. Comparison of BID and TID dosing Viracept (Nelfinavir, NFV) in combination with stavudine (d4T) and lamivudine (3TC). Fifth Conference on Retroviruses and Opportunistic Infections.
Chicago, February 1998 [abstract 373].
26. Sension M, Elion R, Farthing C, et al. Open label pilot studies to assess the safety and efficacy of bid dosing regimens of Viracept (nelfinavir mesylate) combined with NRTI‚s in HIV-infected treatment-naive patients. Fifth Conference on Retroviruses and Opportunistic Infections.
Chicago, February 1998 [abstract 387a].
27. Kline MW, Fletcher CV, Brundage RC, et al. Dual HIV protease inhibitor (protease inhibitor) therapy of children. Fourth International Congress on Drug therapy in HIV Infection.
Glasgow, November 1998 [abstract 431].
28. Horneff G, Adams O, Wahn V. Pilot study of zidovudine-lamivudine combination therapy in vertically HIV-infected antiretroviral-naive children. AIDS
29. Egger M, Hirschel B, Francioli P, et al. Impact of new antiretroviral combination therapies in HIV infected patients in Switzerland: prospective multicentre study
30. Levy JA. Caution: should we be treating HIV infection early? Lancet
31. Markowitz M, Cao Y, Hurley A, et al. Virologic and immunologic response to triple therapy with nelfinavir in combination with AZT/3TC in antiretroviral naive HIV-infected subjects at 20 months. Fifth Conference on Retroviruses and Opportunistic Infections.
Chicago, February 1998 [abstract 371].
32. Notermans DW, Jurriaans S, de Wolf F, et al. Decrease of HIV-1 RNA levels in lymphoid tissue and peripheral blood during treatment with ritonavir, lamivudine and zidovudine. AIDS