AVT040204-Gatell

Antiviral Therapy 4: 79-86 AVANTI 1: randomized, double-blind trial to evaluate the efficacy and safety of zidovudine plus lamivudine versus zidovudine plus lamivudine plus loviride in HIV-infected antiretroviral-naive patients

J Gatell2, J Lange3, M Gartland1* and The AVANTI study group

1Glaxo Wellcome Research and Development, Middlesex, UK 2Hospital Clinico i Provincial Casanova, Barcelona, Spain 3Academic Medical Centre, Amsterdam, The Netherlands

*Corresponding author: Tel: +44 181 966 3135; Fax: +44 181 966 3310; E-mail: [email protected]

The objective of this randomized double-blind, placebo- plasma HIV-1 RNA as measured by RNA PCR and the controlled trial was to investigate the effect of development of drug-related toxicities sufficiently combination antiretroviral therapy on plasma HIV-1 severe to warrant dose modification, interruption or RNA as measured by HIV RNA PCR and to assess the permanent discontinuation. A mild, though statistically safety and tolerability of such regimens. The trial was significant difference in favour of zidovudine plus carried out in seven European countries, Australia and lamivudine plus loviride for log10 plasma HIV-1 RNA Canada and involved antiretroviral-naive patients (P=0.022), as compared to zidovudine plus lamivudine, (n=106) with CD4 counts between 150–300 cells/mm3 was observed using area-under-the-curve minus base- (CDC group A) and 150–500 cells/mm3 (CDC group B/C). line (AUCMB). An increase in CD4 cell count in the Patients were randomly assigned to zidovudine (200 mg zidovudine plus lamivudine plus loviride group was three times daily) plus lamivudine (300 mg twice daily) observed with a median improvement of 124 cells/mm3 or to zidovudine plus lamivudine plus loviride (100 mg at week 52 compared with 70 cells/mm3 in the zidovu- three times daily) for 52 weeks. The main outcome dine plus lamivudine group (P=0.06). Both treatment measures were degree and duration of reduction of regimens were well tolerated.

Introduction

It is now universally recognized that the treatment of Although such results are encouraging, these data, HIV-1 infection requires the use of a combination of together with data from earlier studies of the combina- two or more antiretroviral drugs. This is supported by tion of zidovudine with the nucleoside analogues a number of recent clinical studies in which significant didanosine and zalcitabine, indicate that complete reductions in the rate of disease progression and suppression of plasma HIV-1 RNA for extended mortality were observed in patients receiving combina- periods can not be achieved with a double nucleoside tion therapy as compared to those receiving single analogue combination [11,12]. agent treatment [1–5]. AVANTI 1 is the first of a series of three studies A series of four studies investigating the combina- designed to investigate triple drug combinations in tion of zidovudine and lamivudine in both naive patients, utilizing the three main different classes antiretroviral-naive and -experienced HIV-1-infected of antiretroviral drugs currently available, nucleoside patients have been reported [6–9]. These demonstrated analogues, non-nucleoside analogues and protease a pronounced and sustained increase in CD4 cell count inhibitors (PI). Each study uses zidovudine plus and reduction in plasma HIV-1 RNA. A subsequent lamivudine as the basic building block to which either meta-analysis indicated that these changes in CD4 cell a non-nucleoside analogue or PI is added. The second count and plasma HIV-1 RNA were associated with a and third studies in the series, named AVANTI 2 and reduction in disease progression [10]. In a large clinical 3, investigate zidovudine plus lamivudine in combina- end point study (CAESAR), the addition of lamivudine tion with the PIs indinavir and nelfinavir, respectively. to zidovudine-containing regimens reduced disease This paper describes the results from the initial study progression by over 50% [5]. in which zidovudine plus lamivudine was compared

with zidovudine plus lamivudine plus loviride. This paper deals only with the first 52 weeks double- blind treatment period. Patients and Methods Treatment regimens Patients Patients were randomized to either zidovudine 200 mg HIV-infected patients ≥18 years old were eligible for the three times daily plus lamivudine 300 mg twice daily study if their CD4 cell count was ≥150 cells/mm3 and plus placebo three times daily or zidovudine 200 mg ≤300 cells/mm3 (CDC classification group A) or ≥150 three times daily plus lamivudine 300 mg twice daily cells/mm3 and ≤500 cells/mm3 (CDC classification group plus loviride 100 mg three times daily using a B/C) on two separate occasions, 72 h apart and within 4 computer-generated scheme. Zidovudine was supplied weeks of the anticipated entry date. For the purpose of as 100 mg capsules, lamivudine as 300 mg tablets and this study, a CD4 count <200 cells/mm3, in the absence loviride as 100 mg capsules. Notification of patient of any other AIDS-defining indicator conditions, was entry and allocation of treatment was by fax. not considered AIDS-defining. Women of childbearing One level of dose reduction was permitted for drug- age must have had a negative β-HCG test within 14 days related toxicity: zidovudine 100 mg three times daily, of entry and all participants were encouraged to utilize lamivudine 150 mg twice daily and loviride 100 mg adequate contraception throughout the study period. twice daily. Investigators were not permitted to reduce Patients were excluded if they had received any previous the dose of individual drugs. If dose reduction was treatment with nucleoside analogue reverse transcriptase required, then all three drugs or placebo were reduced. inhibitors (NRTIs), non-nucleoside analogue reverse For grade 1 and 2 toxicities, investigators were transcriptase inhibitors (NNRTIs) or PIs. In addition, permitted to use their own judgement as to whether a patients were not permitted to have any active or dose reduction was necessary. In the case of a grade 3 ongoing AIDS-defining opportunistic infections or toxicity, investigators were required to temporarily diseases, or have a history of lymphoma or Kaposi’s stop medication until the toxicity returned to grade 2 sarcoma. or less, for a period of 2 weeks; reduced dose medication could then be restarted. Patients who had a Study design recurring grade 3 toxicity within a 4 week interval This was a multicentre trial run in Australia, Belgium, despite the dose modification, or a non-resolving toxi- Canada, Denmark, France, Germany, Holland, Italy city of ≥6 weeks duration had all study medications and Spain. It was a randomized, double-blind compar- permanently discontinued. ative study designed to compare the efficacy and safety Those patients who experienced a grade 4 toxicity of the combination regimens zidovudine plus lamivu- had all study medications permanently discontinued. dine versus zidovudine plus lamivudine plus loviride A Data Safety Monitoring Board (DSMB) was administered to HIV-infected patients with baseline established to closely observe any safety issues arising CD4 cell counts between 150 cells/mm3 and 500 from the study. This body met approximately every 3 cells/mm3. Patients satisfying the inclusion criteria were months throughout the duration of the trial. randomized to one of the two combination antiretroviral regimens, employing a stratification procedure Protocol amendments based on CDC HIV classification groups A and B/C to In the early part of the study, five patients experienced balance both groups for the stratification factors. grade 4 anaemia and the DSMB recommended the Patients were screened for eligibility and then evalu- introduction of extra assessments at weeks 20 and 24 ated prior to entry, at entry, at weeks 2, 4, 8, 12, 16 for patients who experienced a drop in haemoglobin and every 12 weeks thereafter. Dosing with study value of >2 g/dl and which took the value below 11 medication and scheduled visits continued to week 52. g/dl. In addition, patients who experienced an adverse Patients permanently discontinuing study medication, event of anaemia were required to undergo a series of prior to completion of 52 weeks, were assessed at the investigations to better understand the underlying regular study intervals for development of AIDS (or cause of the condition. new AIDS-defining events), survival, virological and immunological status, haematological and biochemical Statistical analysis parameters, antiretroviral therapy and adverse events. Patients completing the 52 week period had the Primary end points. Degree and duration of reduction of option to continue treatment. If patients chose to plasma HIV-1 RNA as measured by RNA PCR (Roche continue treatment, they could either continue on Molecular Systems assay); development of drug-related randomized treatment for another 24 weeks or switch toxicities sufficiently severe to warrant dose modifica- to open-label zidovudine plus lamivudine plus loviride. tion, interruption or permanent discontinuation.

Table 1. Demographic data Table 2. Percentage change over 52 week treatment period as measured by AUCMB Zidovudine/ Zidovudine/ Zidovudine/ lamivudine/ Zidovudine/ lamivudine/ Parameter lamivudine loviride Parameter lamivudine loviride P value Number 52 54 Viral RNA Age (mean) 38 years 37 years Sex 85% male 76% male Median decrease* –1.3 –1.6 0.022 Race 92% Caucasian 98% Caucasian (log10 copies/ml) CDC group A 15 16 CDC group B/C 37 38 CD4 count

Plasma HIV-1 RNA (log10, median) 4.83 4.95 Median increase* 54.8 84.9 0.060 6 CD4 count (x10 /l, median) 270 270 (cells/mm3) *Over 52 weeks as measured by AUCMB.

Secondary end points. Degree and duration of increase in CD4 cell counts from baseline; development of or the investigator classified compliance with each study delay in development of AIDS-defining indicator drug as either 100%, 99–76%, 75–26% or 25–0%. diseases or death; development of or delay in development of genotypic and phenotypic markers of reduced Adverse events. Each adverse event (AE) was classified viral sensitivity to reverse transcriptase inhibitors used by body system and the number and percentage of in dosing regimens; changes in quality of life as patients who reported an AE were tabulated by treat- measured using the EuroQol questionnaire. ment group, causality and maximum intensity or All patients who initiated treatment and returned seriousness. for a clinical visit were included in the evaluation of efficacy. There were no provisions to impute missing Laboratory safety data. Individual measurements were data. The safety analysis included all patients who compared to the laboratory specific range and values initiated treatment (intent-to-treat population). All outside the normal range were compared to toxicity statistical tests were interpreted at the 5% significance grading definitions specified in the protocol. Summary level (two-tailed). statistics for each measured haematology and clinical chemistry test were determined over time by treatment Efficacy. The primary efficacy parameter was the area- group. under-the-curve minus baseline (AUCMB) of the log10 transformed HIV RNA PCR which is the average area- Results under-the-curve (AAUC) divided by baseline and then multiplied by 100. Area-under-the-curve (AUC) was Subject population calculated according to the trapezoidal method. AAUC The trial was run from 30 March 1995 to 31 August was calculated as the AUC divided by the number of 1996 and twenty centres from Australia, Belgium, days on the treatment phase. Baseline was defined as Canada, Denmark, France, Germany, Holland, Italy the mean of the log10 plasma HIV-1 RNA collected at and Spain participated. In total, 106 patients were screen and baseline. recruited; 52 were assigned to the zidovudine plus The limit of detection of the RNA PCR assay was lamivudine treatment group and 54 were assigned to

400 copies/ml (2.6 log10 copies/ml). The AUCMB was the zidovudine plus lamivudine plus loviride treat- analysed using an analysis of covariance with factors ment group. Eighteen patients (17%) dropped out for treatment and CDC classification. The interaction before trial completion, 11 in the zidovudine plus between the two factors was tested: if the corre- lamivudine group and seven in the zidovudine plus sponding P value was less than 0.1, the interaction lamivudine plus loviride group. Of the 11 patients term was included in the model. who withdrew in the double arm, however, one Summary statistics were calculated for the CD4 cell patient died, one was lost to follow-up and one with- count (absolute data and change from baseline) over drew consent. Table 1 shows the demographic data of time and for each treatment group. The AUCMB of the the 106 patients recruited to the trial. CD4 cell count was analysed using the same analysis of Overall, 80% of the patients were male, 95% were covariance model as for the AUCMB of the log10 Caucasian and the mean age was 38 years. The plasma HIV-1 RNA. majority of patients had become infected through homosexual contact (65% in the zidovudine plus Compliance. Patients were requested to return used and lamivudine group and 61% in the zidovudine plus unused study medication packs at each clinic visit and lamivudine plus loviride group). Demographic data

Antiviral Therapy 4:2 81 M Gartland et al.

Table 3. Most commonly reported adverse experiences and 0.85 for zidovudine plus lamivudine plus loviride. The most common reason for adjustments Zidovudine/ Zidovudine/ was due to a clinical adverse event (47% in the lamivudine lamivudine/loviride n (%) n (%) zidovudine plus lamivudine group versus 42% in the Nausea 16 (31%) 26 (48%) zidovudine plus lamivudine plus loviride group). The Headache 11 (21%) 11 (20%) mean number of days before the dose was reduced was 126 days in the zidovudine plus lamivudine Malaise and fatigue 10 (19%) 15 (28%) group versus 70 days in the zidovudine plus lamivu- Nausea and vomiting 9 (17%) 8 (15%) dine plus loviride group. Diarrhoea 6 (12%) 3 (6%) Anaemia 5 (10%) 5 (9%) Efficacy Feeding problems 8 (15%) 1 (2%) Abdominal discomfort 5 (10%) 3 (6%) Plasma HIV-1 RNA. The median baseline plasma HIV-1 and pain RNA, as measured by quantitative HIV RNA PCR Leukopenia 6 (12%) 4 (7%) (lower limit of detection 400 copies/ml) was compa-

rable in the two treatment groups, with 4.83 log10 copies/ml in the zidovudine plus lamivudine group and and baseline disease characteristics were comparable in 4.95 log10 copies/ml in the zidovudine plus lamivudine the two treatment groups. plus loviride group (Table 1). Both groups had the greatest reduction in plasma HIV-1 RNA at week 4;

Treatment compliance and exposure to study drug the median reduction (log10 copies/ml) was –1.88 in the Measurements of compliance were performed by zidovudine plus lamivudine arm and –2.13 in the group and by time interval. Overall compliance was zidovudine plus lamivudine plus loviride arm. The acceptable with the majority of the patients taking largest difference between the treatment groups for this more than 76% of each prescribed study drug in both parameter was noted at week 8 (0.49 log10 copies/ml) groups. (Figure 1). Most patients had no dose modification during the The AUCMB calculation was used to determine study (60% in zidovudine plus lamivudine versus degree and duration of reduction in plasma HIV-1 72% in zidovudine plus lamivudine plus loviride). RNA. Over the 52 week treatment period, this analysis

The average number of dose adjustments over the indicated a median decrease from baseline of –1.3 log10 study period was 1.06 for zidovudine plus lamivudine copies/ml for the zidovudine plus lamivudine group

Figure 1. Median change from baseline in viral load

-0.5

-1 HIV-1 RNA 10

Log -1.5

-2

-2.5 0102030405060 Study week

Wk 0 Wk 2 Wk 4 Wk 8 Wk 12 Wk 16 Wk 28 Wk 40 Wk 52 Zidovudine/lamivudine (n) 47 41 44 44 37 42 37 35 37 Zidovudine/lamivudine/loviride (n) 53 43 45 49 49 50 45 39 44

Figure 2. Patients with HIV-1 RNA 500 copies/ml: Missing=Failure analysis

30 atients below 500 copies/ml (%) P 20

0 0102030405060 Study week

Wk 0 Wk 2 Wk 4 Wk 8 Wk 12 Wk 16 Wk 28 Wk 40 Wk 52 Zidovudine/lamivudine (n/N) 1/52 12/52 24/52 17/52 10/52 12/52 7/52 5/52 5/52 Zidovudine/lamivudine/loviride (n/N) 1/54 10/54 19/54 28/54 19/54 12/54 7/54 7/54 9/54

and –1.6 log10 copies/ml for the zidovudine plus and two in the zidovudine plus lamivudine plus lamivudine plus loviride group; a statistically signifi- loviride group) and one patient (zidovudine plus cant difference (P=0.022, Table 2). lamivudine group) died during the study. To allow comparison with previous work [13], the proportion of patients below 500 copies/ml is detailed Safety evaluation as a Missing=Failure analysis in Figure 2. The proportion of patients with plasma HIV-1 RNA levels below Adverse events. 45/52 (87%) patients in the 500 copies/ml was higher in the zidovudine plus zidovudine/lamivudine group and 45/54 (83%) of the lamivudine plus loviride group (52%) compared with patients in the zidovudine plus lamivudine plus loviride the zidovudine plus lamivudine arm (33%) at week 8. group had at least one adverse event and there were no At week 52, 17% of the zidovudine plus lamivudine statistical differences between the groups. The most plus loviride patients had plasma HIV-1 RNA levels frequently reported adverse events were nausea, below 500 copies/ml compared with 8% in the zidovu- headache, nausea and vomiting, diarrhoea, anaemia, dine plus lamivudine group. gastrointestinal infections, leukopenia, malaise and fatigue. The incidence of adverse events was similar in Effect on CD4 count. CD4 cell count increased signifi- both groups, except for nausea, which occurred more cantly for both groups and was sustained for the 52 frequently in the zidovudine plus lamivudine plus loviride week treatment period. The median increase over base- group (26 patients, 48%) compared with the zidovudine line was more pronounced in the zidovudine plus plus lamivudine group (16 patients, 31%). This differ- lamivudine plus loviride group with an increase of 124 ence was only noted in the first weeks of the study. cells/mm3 at week 52 compared to 70 cells/mm3 for the Adverse events reported as possibly or reasonably zidovudine plus lamivudine group (Figure 3). AUCMB related to study drug in at least five patients in either analysis over the 52 week period for CD4 cell count group are given in Table 3. In total, six (11%) patients indicated a median increase from baseline of 54.8 in the zidovudine plus lamivudine group and five (9%) cells/mm3 in the zidovudine plus lamivudine group and patients in the zidovudine plus lamivudine plus loviride 84.9 cells/mm3 in the zidovudine plus lamivudine plus group permanently stopped treatment because of an loviride group (P=0.060; Table 2). adverse event.

Development of AIDS or death Severity. The incidence of grade 3 and 4 adverse events Three patients were diagnosed with an AIDS-defining was low, with a maximum of two cases (4% of event (one in the zidovudine plus lamivudine group patients) of malaise and fatigue, anaemia or leukopenia

Antiviral Therapy 4:2 83 M Gartland et al.

Figure 3. Median change from baseline in CD4 cell count

250

200 ) 3 150

100 CD4 cell count (cells/mm 50

0 0102030405060 Study week

Wk 0 Wk 2 Wk 4 Wk 8 Wk 12 Wk 16 Wk 28 Wk 40 Wk 52 Zidovudine/lamivudine (n) 52 44 48 50 45 46 40 43 39 Zidovudine/lamivudine/loviride (n) 54 49 53 52 48 51 48 49 48

in the zidovudine plus lamivudine group and nausea or is then followed by a return towards baseline. The use nausea and vomiting in the zidovudine plus lamivudine of AUCMB in this study allows assessment of the plus loviride group. All of the adverse events were progress of the marker over the 52 week period, taking rated by the investigator as possibly or reasonably into account the value at baseline. Such AUCMB caused by the study drug. analysis indicated a small but statistically significant difference in favour of zidovudine plus lamivudine plus Deaths. One patient in the zidovudine plus lamivudine loviride for plasma HIV-1 RNA. The increase of the group died due to relapse of carcinoma. This adverse absolute CD4 cell count over baseline was more event was assessed by the investigator as not related to pronounced in the zidovudine plus lamivudine plus the study drug. loviride group. The proportion of patients with plasma HIV-1 RNA below 500 copies/ml was generally higher Serious adverse events. The serious AE reported as in the zidovudine plus lamivudine plus loviride group. related to the study drug were anaemia, three patients In terms of clinical outcome, AVANTI 1 was not (6%) in the zidovudine plus lamivudine plus loviride powered to detect differences in this parameter and group and four patients (7%) in the zidovudine plus only three patients experienced an AIDS-defining lamivudine plus loviride group; leukopenia, one illness and one patient died. The CAESAR study, which patient (2%) in the zidovudine plus lamivudine group compared current therapy versus current therapy plus and primary malignant male reproductive neoplasm, lamivudine versus current therapy plus lamivudine plus one patient (2%) in the zidovudine plus lamivudine loviride, was not powered to detect differences in clin- plus loviride group. ical outcome between the latter two arms, and in the study as a whole the difference was marginal. Discussion However, in a subset of naive patients with higher CD4 counts (175–250 cells/mm3) there was some limited In the present trial, 106 antiretroviral-naive HIV-1 evidence of a benefit from the addition of loviride to positive patients with CD4 cell counts between 150 lamivudine plus current therapy as compared to and 500 cells/mm3 were randomized in a double-blind lamivudine plus current therapy alone in the number of study to receive either zidovudine plus lamivudine or clinical endpoints [15 of 251 (6%) versus 3 of 135 zidovudine plus lamivudine plus loviride. The median (2%)], without reaching significance [5]. These data baseline CD4 cell count was 270 cells/mm3 in each would therefore seem to correlate with the CD4 and group. In many antiretroviral regimens, treatment plasma HIV-1 RNA findings presented here from results in an initial effect on a surrogate marker which AVANTI 1. Similarly, the zidovudine plus lamivudine

data presented here is within the range observed in AVANTI 1. This will be difficult to determine, other studies [6–9]. however, as loviride is no longer in clinical develop- The combination of two nucleoside analogues with ment. a non-nucleoside analogue was also investigated in the INCAS study [14], where naive patients received either Acknowledgements zidovudine plus didanosine, zidovudine plus nevirapine or zidovudine plus didanosine plus nevirapine. The AVANTI 1 Study Group would like to thank all More than 50% of those who had taken the triple the patients who participated for their time and effort, combination in the INCAS study had plasma HIV-1 together with the many staff members who were RNA levels below 200 copies/ml at 52 weeks and the involved in the study. Without their involvement, the median change from baseline for CD4 count was 125 study would not have gone ahead. cells/mm3. In comparison to the patients entered in the The AVANTI 1 Study Group: Writing Committee: AVANTI 1 study the patients receiving the triple M Gartland, J Gatell and J Lange; Steering combination in the INCAS study had less advanced Committee: N Clumeck, DA Cooper, J Gatell, B disease (baseline median CD4 387 cells/mm3; median Gazzard, J Gerstoft, F Goebel, J Lange, J Montaner, plasma HIV-1 RNA at baseline 4.54 log10 copies/ml) W Rozenbaum, P Reiss and S Vella; Centres and [14]. An excess incidence of rash attributable to nevi- Investigators: Australia: DA Cooper (St Vincents rapine (15%) was reported in the first 6 months of Hospital, Sydney and NCHECR); J Gold (Albion treatment for those patients in the INCAS study and Street Centre, NSW); M Haberl (Glaxo Wellcome); the dropout rate at week 52 in the zidovudine plus Belgium: N Clumeck (Hospital St Pierre, Bruxelles); didanosine plus nevirapine arm was 20% (10/51 D Luyts (Glaxo Wellcome); Canada: J Montaner (St patients). More recently, data from the DMP-006 study Paul’s Hospital, Vancouver); A Rachlis (Sunnybrook has been presented in which efavirenz was used in Health Science Centre, Ontario); R Marina (Glaxo combination with zidovudine and lamivudine and, in a Wellcome); Denmark: P Andersen (Marselisborg Missing=Failure analysis, 65% of patients had HIV-1 Hospital, Arhuus); J Gerstoft (Rigshospitalet, RNA PCR values of <50 copies/ml at 52 weeks [15]. Copenhagen); L Mathiesen (Hvidovre Hospital, Of the patients receiving zidovudine/lamivudine/ Copenhagen); B Elbrond (Glaxo Wellcome); France: efavirenz, 29% experienced grade 1 or 2 rash and 55% J-M Molina (Hôpital Saint Louis, Paris); G Pialloux reported nervous system symptoms. (Hôpital Pasteur, Paris); W Rozenbaum (Hôpital Both treatment regimens in AVANTI 1 were well Rothschild, Paris); L Beauvais (Glaxo Wellcome); tolerated; the incidence of dropouts related to an Germany: FD Goebel (Klinikum Innenstadt Ludwig adverse experience was 11% (6/52 patients) in the Maximilians Universität Pettenkofer, München); S zidovudine plus lamivudine group and 9% (5/54 Staszewski (Zentrum der Inneren Medizin der patients) in the zidovudine plus lamivudine plus Johann-Wolfgang-Goethe-University, Frankfurt); I loviride group. The most common adverse event, Bruns, M Hug (Glaxo Wellcome); Holland: PHJ nausea, occurred more frequently in the zidovudine Frissen (Onze Lieve Vrouwe Gasthuis Eerste, plus lamivudine plus loviride group than in the zidovu- Amsterdam); P Reiss, J Lange (AMC, Amsterdam); M dine plus lamivudine group. Only two patients Bosboom, C Baas (Glaxo Wellcome); Italy: F Milazzo receiving zidovudine plus lamivudine plus loviride and (I Divisione Malattie Infettive Ossedale, Milan); M one patient receiving zidovudine plus lamivudine Moroni (Clinica Malattie Infettive Universita di reported a rash that was considered related to study Milano, Milan); R Panebianco (Glaxo Wellcome); drug and the overall incidence of rash was low in both Spain: B Clotet (Hospital Germans Trias i Pujol, groups (6% in zidovudine plus lamivudine and 4% in Barcelona); JM Gatell Artigas (Hospital Clinico i zidovudine plus lamivudine plus loviride). The inci- Provincial Casanova, Barcelona); J Gonzalez-Lahoz dence of other adverse events was similar in both (Instituto de Salud Carlos III Sinesio, Madrid); M groups. Leal (Hospital Virgen del Rocio Avda Manuel Siurot, In conclusion, the safety profile of zidovudine/ Seville); C Rodriguez-Lopo, B Gandarias (Glaxo lamivudine/loviride was generally better than that Wellcome); Data and Safety Monitoring Board: E observed for similar combinations of two nucleoside Belsey, J Darbyshire, E Sandstrom; Janssen Research analogues and a non-nucleoside analogue containing Foundation: P Stoffels, R Van Den Broeck, W nevirapine and efavirenz. However, in terms of efficacy Verbiest; Glaxo Wellcome Research and Develop- both nevirapine and efavirenz are superior. In view of ment: M Coughlan, P Fiddian, M Gartland, R the tolerability observed with loviride, it is possible Harrigan, B Larder, M Maguire, J Millard, K Patel, D that a more significant antiviral effect would have been Shortino, R Athisegaran, I Vafidis, SA Moore and observed with a higher dose than that studied in J Yeo.

Antiviral Therapy 4:2 85 M Gartland et al.

References patients. Journal of the American Medical Association 1996; 276:111–117. 1. Delta Coordinating Committee. Delta: a randomised 9. Bartlett JA, Benoit SL, Johnson VA, Quinn JB, Sepulvedal double-blind controlled trial comparing combinations of GE, Ehmann WC, Tsoukas C, Fallon MA, Self PL & zidovudine plus didanosine or zalcitabine with zidovudine Rubin M. Lamivudine plus zidovudine compared with zalcitabine plus zidovudine in patients with HIV infection. alone in HIV-infected individuals. Lancet 1996; Annals of Internal Medicine 1996; 125:161–172. 348:283–291. 10. Staszewski S, Hill AM, Bartlett Jeron JJ, Katlama C, 2. Hammer SM, Katzenstein DA, Hughes MD, Gundacker H, Johnson J, Sawyer W & McDade H. Reductions in HIV-1 Schooley RT, Haubrich HH, Henry WK, Lederman MM, disease progression for ZDV/3TC relative to control treat- Phair JP, Hirsch MS & Merigan TC. A trial comparing ments: a meta-analysis of controlled trials. AIDS 1997; nucleoside monotherapy with combination therapy in HIV- 11:477–483. infected adults with CD4 cell counts from 200 to 500 per cubic millimetre. New England Journal of Medicine 1996; 11. Collier AC, Coombs RW, Fischl MA, Skdnik PR, Nothfelt 335:1081–1090. D, Boutin P, Hooper CJ, Kaplan LD, Volberding PA & Davis LG. Combination therapy with zidovidine and 3. Cameron DW, Heath-Chiozzi M, Kravcik S, Mills R, didanosine compared with zidovudine alone in HIV-1 Potthoff A & Henry D. Prolongation of life and prevention infection. Annals of Internal Medicine 1993; 119:786–793. of AIDS complications in advanced HIV immunodeficiency with ritonavir. XI International Conference on AIDS. July 12. Fischl MA, Stanley K, Collier AC, Arduino JM, Stein DS, 1996, Vancouver, Canada. Abstract MoB411. Feinberg JE, Allan JD, Goldsmith JC & Powderly WG. Combination and monotherapy with zidovudine and 4. Salgo MP, Beattie D, Bragman K, Donatacci L, Jones M & zalcitabine in patients with advanced HIV disease. The Montgomery L. Saquinavir (invirase, SQV) vs Hivid NIAID AIDS Clinical Trials Group. Annals of Internal (zalcitabine, ddC) vs combination as treatment for Medicine 1995; 122:24–32. advanced HIV infection in patients discontinuing/unable to take Retrovir (zidovudine, ZDV). XI International 13. Hirsch M, Meibohm A, Rawlins S & Leavitt R. Indinavir Conference on AIDS. July 1996, Vancouver, Canada. (IDV) in combination with zidovudine (ZDV) and lamivu- Abstract MoB410. dine (3TC) in ZDV-experienced patients with CD4 cell counts ≤50 cells/mm3. 4th Conference on Retroviruses and 5. Caesar Coordinating Committee. Randomised trial of addi- Opportunistic Infections, January 1997, Washington, USA. tion of lamivudine or lamivudine plus loviride to Abstract LB7. zidovudine-containing regimens for patients with HIV-1 infection: the CAESAR trial. Lancet 1997; 349:1413–1421. 14. Montaner JSG, Reiss P, Cooper D, Vella S, Harris M, Conway B, Wainberg MA, Smith D, Robinson P, Hall D, 6. Eron JJ, Benoit SL, Jemsek J, MacArthur RD, Santana J, Myers M & Lange JMA for the INCAS Study Group. A Quinn JB, Kuritzkes DR, Fallon MA & Rubin M. randomised, double-blind trial comparing combinations of Treatment with lamivudine, zidovudine or both in HIV- nevirapine, didanosine and zidovudine for HIV-infected positive patients with 200-500 CD4+ cells per millimetre. patients. Journal of the American Medical Association New England Journal of Medicine 1995, 333:1662–1669. 1998; 279:930–937. 7. Katlama C, Ingrand D, Loveday C, Clumeck N, Mallolas J, 15. Tashima K, Staszewski S, Stryker R, Johnson P, Nelson M, Staszewski S, Johnson M, Hill AM, Pearce G & McDade Morales-Ramirez J, Manion DJ, Farina D, Labriola D, H. Safety and efficacy of lamivudine-zidovudine combina- Ruiz N & the Study 006 Investigator Team. A phase III, tion therapy in antiretroviral naive patients. Journal of the multicentre, randomised open-label study to compare the American Medical Association 1996; 276:118–125. antiretroviral activity and tolerability of efavirenz (EFV) + 8. Staszewski S, Loveday C, Picazo JJ, Dellarnonica P, Skinhoj indinavir (IDV), versus EFV + zidovudine (ZDV) + lamivu- P, Johnson MA, Danner SA, Harrigan PR, Hill AM, Verrity dine (3TC), versus IDV + ZDV + 3TC at 48 weeks. 6th L & McDade H. Safety and efficacy of lamivudine-zidovu- Conference on Retroviruses and Opportunistic Infections, dine combination therapy in zidovudine-experienced February 1999, Chicago, USA. Abstract LB16.

Received 23 October 1998; accepted 14 April 1999