The two original objectives were to determine in HIV-infected children initiating antiretroviral therapy (ART): 1. Whether clinically driven monitoring (CDM) will have a similar outcome in terms of disease progression or death as routine laboratory and clinical monitoring (LCM) for toxicity (haematology/biochemistry) and efficacy (CD4)? 2. Whether induction with four drugs from two ART classes followed by maintenance with three drugs after 36 weeks be more effective than a continuous non-nucleoside reverse transcriptase inhibitors (NNRTI)-based triple drug regimen in terms of CD4 and clinical outcome? Two secondary objectives were to determine 3. Whether changing from twice daily lamivudine+abacavir to once daily lamivudine+abacavir after 48 weeks on ART will have a similar outcome in terms of virological suppression and will result in improvements in adherence to ART? 4. Whether stopping daily cotrimoxazole prophylaxis in children over 3 years of age who have been on ART for at least 96 weeks has a similar outcome in terms of hospitalisation or death as continuing daily cotrimoxazole?
The ARROW (AntiRetroviral Research fOr Watoto) protocol describes an open-label randomised trial primarily evaluating two strategic approaches for management of antiretroviral therapy (ART) in 1200 symptomatic HIV-infected infants and children initiating ART following WHO guidelines in Uganda and Zimbabwe. The first strategy compares clinically driven monitoring (CDM) with laboratory plus clinical monitoring (LCM). In both groups, tests for toxicity (standard haematology and biochemistry panels) and efficacy (lymphocyte subsets including CD4 count) will be done every 12 weeks. In LCM, all results will be returned for patient management. In CDM, physicians may request results from routine haematology/biochemistry panels if needed for clinical management, but results will not be returned routinely, and lymphocyte subsets will never be returned. Extra laboratory tests may be requested outside of the scheduled visits at any time in either group (except for lymphocyte subsets in CDM). The second strategy compares a continuous WHO-recommended first-line ART three-drug two-class regimen, comprising two Nucleoside Reverse Transcriptase Inhibitors (NRTIs) plus one Non-Nucleoside Reverse Transcriptase Inhibitor (NNRTI), with induction with four drugs (two classes) for 36 weeks followed by maintenance with three drugs. After at least 36 and 96 weeks on ART respectively, two further randomisations will assess simplification strategies which could improve long-term ART adherence (i) once versus twice daily lamivudine+abacavir NRTI drugs (ii) stopping versus continuing daily cotrimoxazole prophylaxis.
Study Type
INTERVENTIONAL
Allocation
RANDOMIZED
Purpose
TREATMENT
Masking
NONE
Enrollment
1,206
Participants were examined by a doctor and had routine full blood count with white cell differential, lymphocyte subsets (CD4, CD8), biochemistry tests (bilirubin, urea, creatinine, aspartate aminotransferase, alanine aminotransferase) at screening, randomisation (lymphocytes only), weeks 4, 8, and 12, then every 12 weeks. Screening results were used to assess eligibility. All subsequent results at and after randomisation were only returned if requested for clinical management (authorised by centre project leaders); haemoglobin results at week 8 were automatically returned on the basis of early anaemia in a previous adult trial as were grade 4 laboratory toxicities (protocol safety criteria). Total lymphocytes and CD4 tests were never returned for CDM participants, but for all children other investigations (including tests from the routine panels) could be requested and concomitant drugs prescribed, as clinically indicated at extra patient-initiated or scheduled visits.
Participants were examined by a doctor and had routine full blood count with white cell differential, lymphocyte subsets (CD4, CD8), biochemistry tests (bilirubin, urea, creatinine, aspartate aminotransferase, alanine aminotransferase) at screening, randomisation (lymphocytes only), weeks 4, 8, and 12, then every 12 weeks. All results were returned to physicians for patient management. Other investigations (including tests from the routine panels) could be requested and concomitant drugs prescribed, as clinically indicated at extra patient-initiated or scheduled visits.
Children received a standard WHO-recommended regimen of open-label lamivudine, abacavir, plus NNRTI continuously. The NNRTI (nevirapine or efavirenz) was chosen by clinicians according to local availability and age.
Children initiated ART using an induction-maintenance approach, starting with open-label four-drug lamivudine, abacavir, NNRTI, plus zidovudine for 36 weeks, then open-label lamivudine, abacavir, plus NNRTI subsequently. The NNRTI (nevirapine or efavirenz) was chosen by clinicians according to local availability and age.
Children initiated ART using an induction-maintenance approach, starting with open-label four-drug lamivudine, abacavir, NNRTI, plus zidovudine for 36 weeks, then open-label lamivudine, abacavir, plus zidovudine subsequently (triple NRTI maintenance). The NNRTI (nevirapine or efavirenz) was chosen by clinicians according to local availability and age.
MRC /UVRI Uganda Research Unit on AIDS
Entebbe, Uganda
Joint Clinical Research Centre
Kampala, Uganda
Baylor College of Medicine Children's Foundation
Mulago, Uganda
University of Zimbabwe Medical School
Harare, Zimbabwe
LCM vs CDM: Disease Progression to a New WHO Stage 4 Event or Death
Number of participants with disease progression to a new WHO stage 4 event or death, to be analysed using time-to-event methods
Time frame: Median 4 years (from randomization to 16 March 2012; maximum 5 years)
LCM vs CDM: New Grade 3 or 4 Adverse Event (AE), Not Solely Related to HIV
Number of participants with a new Grade 3 or 4 adverse event (AE), not solely related to HIV, to be analysed using time-to-event methods
Time frame: Median 4 years (from randomization to 16 March 2012; maximum 5 years)
Induction ART: Change From Baseline in CD4% 72 Weeks After ART Initiation
Time frame: Baseline, 72 weeks
Induction ART: Change From Baseline in CD4% to 144 Weeks From ART Initiation
Time frame: Baseline, 144 weeks
Induction ART: New Grade 3 or 4 Adverse Event (AE), Not Solely Related to HIV
Number of participants with a new grade 3 or 4 adverse event (AE), not solely related to HIV, to be analysed using time-to-event methods
Time frame: Median 4 years (from randomization to 16 March 2012; maximum 5 years)
Once Versus Twice Daily Abacavir+Lamivudine: Suppressed HIV RNA Viral Load 48 Weeks After Randomisation
Number of participants with HIV RNA viral load \<80 copies/ml at 48 weeks. Measured retrospectively on stored plasma specimens: due to low stored volumes from some children, samples had to be diluted and therefore a threshold of \<80 copies/ml was used to indicate suppression.
Time frame: 48 weeks
Once Versus Twice Daily Abacavir+Lamivudine: New Grade 3 or 4 Adverse Event (AE), Not Solely Related to HIV, Judged Definitely/Probably or Uncertain Whether Related to Lamivudine or Abacavir
Number of participants with a new grade 3 or 4 adverse event (AE), not solely related to HIV, judged definitely/probably or uncertain whether related to lamivudine or abacavir, to be analysed using time-to-event methods
Time frame: Median 2 years (from once vs twice daily randomization to 16 March 2012; maximum 2.6 years)
Cotrimoxazole: New Hospitalisation or Death
Number of participants with a new hospitalisation or death, to be analysed using time-to-event methods
Time frame: Median 2 years (from cotrimoxazole randomization to 16 March 2012; maximum 2.5 years)
Cotrimoxazole: New Grade 3 or 4 Adverse Event (AE), Not Solely Related to HIV
Number of participants with a new grade 3 or 4 adverse event (AE), not solely related to HIV, to be analysed using time-to-event methods
Time frame: Median 2 years (from cotrimoxazole randomization to 16 March 2012; maximum 2.5 years)
LCM vs CDM, Induction ART: All-cause Mortality
Number of participants who died from any cause, to be analysed using time-to-event methods
Time frame: Median 4 years (from randomization to 16 March 2012; maximum 5 years)
Induction ART: New WHO Stage 4 Event or Death
Number of participants with a new WHO stage 4 event or death, to be analysed using time-to-event methods
Time frame: Median 4 years (from randomization to 16 March 2012; maximum 5 years)
LCM vs CDM, Induction ART: New WHO Stage 3 or 4 Event or Death
Number of participants with a new WHO stage 3 or 4 event or death, to be analysed using time-to-event methods
Time frame: Median 4 years (from randomization to 16 March 2012; maximum 5 years)
LCM vs CDM, Induction ART: New or Recurrent WHO Stage 3 or 4 Event or Death
Number of participants with a new or recurrent WHO stage 3 or 4 event or death, to be analysed using time-to-event methods
Time frame: Median 4 years (from randomization to 16 March 2012; maximum 5 years)
LCM vs CDM, Induction ART: Weight-for-age Z-score
Age-adjusted change in weight-for-age Z-score at all 12-weekly visits attended over the whole follow-up, no specific timepoint prespecified. Mean and SD are time-averaged area under the change curve calculated using the trapezoidal rule. Z-scores calculated using UK norms which cover the full age range of children (Cole, T. J., J. V. Freeman, and M. A. Preece. 1998. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood. Statistics in Medicine 17(4): 407-29).
Time frame: Baseline and a median of 4 years (maximum 5 years)
LCM vs CDM, Induction ART: Height-for-age Z-score
Age-adjusted change in weight-for-age Z-score at all 12-weekly visits attended over the whole follow-up, no specific timepoint prespecified. Mean and SD are time-averaged area under the change curve calculated using the trapezoidal rule. Z-scores calculated using UK norms which cover the full age range of children (Cole, T. J., J. V. Freeman, and M. A. Preece. 1998. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood. Statistics in Medicine 17(4): 407-29).
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.
Time frame: Baseline and a median of 4 years (maximum 5 years)
LCM vs CDM, Induction ART: Body Mass Index-for-age Z-score
Age-adjusted change in body mass index-for-age Z-score at all 12-weekly visits attended over the whole follow-up, no specific timepoint prespecified. Mean and SD are time-averaged area under the change curve calculated using the trapezoidal rule. Z-scores calculated using UK norms which cover the full age range of children (Cole, T. J., J. V. Freeman, and M. A. Preece. 1998. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood. Statistics in Medicine 17(4): 407-29).
Time frame: Baseline and a median of 4 years (maximum 5 years)
LCM vs CDM: Change From Baseline in CD4% to Week 72
Time frame: Baseline, week 72
LCM vs CDM: Change From Baseline in CD4% to Week 144
Time frame: Baseline, week 144
LCM vs CDM, Induction ART: Change From Baseline in Absolute CD4 to Week 72
Estimated in those \>5 years at enrolment, in whom absolute CD4 is meaningful. (In uninfected children, CD4 decreases with age during early childhood.)
Time frame: Baseline, week 72
LCM vs CDM, Induction ART: Change From Baseline in Absolute CD4 to Week 144
Estimated in those \>5 years at enrolment, in whom absolute CD4 is meaningful. (In uninfected children, CD4 decreases with age during early childhood.)
Time frame: Baseline, week 144
CDM vs LCM, Induction ART: Suppression of HIV RNA Viral Load 72 Weeks After Baseline
Number of participants with HIV RNA viral load \<80 copies/ml 72 weeks after baseline. Threshold for suppression \<80 copies/ml as samples had to be diluted due to low volumes.
Time frame: 72 weeks
CDM vs LCM, Induction ART: Suppression of HIV RNA Viral Load 144 Weeks After Baseline
Number of participants with HIV RNA viral load \<80 copies/ml 144 weeks after baseline. Threshold for suppression \<80 copies/ml as samples had to be diluted due to low volumes.
Time frame: 144 weeks
LCM vs CDM, Induction ART: Cessation of First-line Regimen for Clinical/Immunological Failure
Number of participants stopping their first-line regimen for clinical/immunological failure, to be analysed using time-to-event methods
Time frame: Median 4 years (from randomization to 16 March 2012; maximum 5 years)
LCM vs CDM, Induction ART: New Grade 3 or 4 Adverse Event Definitely/Probably or Uncertainly Related to ART
Number of participants with a new grade 3 or 4 adverse event definitely/probably or uncertainly related to ART, to be analysed using time-to-event methods
Time frame: Median 4 years (from randomization to 16 March 2012; maximum 5 years)
LCM vs CDM, Induction ART: New Serious Adverse Events Not Solely Related to HIV
Number of participants with a new serious adverse events not solely related to HIV, to be analysed using time-to-event methods
Time frame: Median 4 years (from randomization to 16 March 2012; maximum 5 years)
LCM vs CDM, Induction ART: New ART-modifying Adverse Event
Number of participants with a new ART-modifying adverse event, to be analysed using time-to-event methods
Time frame: Median 4 years (from randomization to 16 March 2012; maximum 5 years)
LCM vs CDM, Induction ART: Adherence to ART as Measured by Self-reported Questionnaire (Missing Any Pills in the Last 4 Weeks)
Binary outcome measure: missed any doses of ART in the last 4 weeks by self-report. Mean calculated across all 12-weekly visits attended over the whole follow-up (no specific timepoint prespecified), giving the percentage of visits attended where the carer/participant reported missing any pills in the last 4 weeks.
Time frame: Median 4 years (from randomization to 16 March 2012; maximum 5 years)
Once Versus Twice Daily Abacavir+Lamivudine: Suppression of HIV RNA Viral Load 96 Weeks After Randomisation
Number of participants with HIV RNA viral load \<80 copies/ml at 96 weeks. Threshold for suppression \<80 copies/ml as samples had to be diluted due to low volumes.
Time frame: 96 weeks
Once Versus Twice Daily Abacavir+Lamivudine: Change From Baseline in CD4% to Week 48
Time frame: Randomisation to once vs twice daily, week 48
Once Versus Twice Daily Abacavir+Lamivudine: Change From Baseline in CD4% to Week 72
Time frame: Baseline, week 72
Once Versus Twice Daily Abacavir+Lamivudine: Change From Baseline in CD4% to Week 96
Time frame: Randomisation to once vs twice daily, week 96
Once Versus Twice Daily Abacavir+Lamivudine: Change From Baseline in Absolute CD4 to Week 48
Estimated in those \>5 years at enrolment, in whom absolute CD4 is meaningful. (In uninfected children, CD4 decreases with age during early childhood.)
Time frame: Randomisation to once vs twice daily, week 48
Once Versus Twice Daily Abacavir+Lamivudine: Change From Baseline in Absolute CD4 to Week 72
All participants aged \>5 years at randomization to once versus twice daily alive in follow-up with CD4 measured
Time frame: Baseline, week 72
Once Versus Twice Daily Abacavir+Lamivudine: Change From Baseline in Absolute CD4 to Week 96
All participants aged \>5 years at randomization to once versus twice daily alive in follow-up with CD4 measured
Time frame: Randomisation to once vs twice daily, week 96
Once Versus Twice Daily Abacavir+Lamivudine: All-cause Mortality
Number of participants who died, to be analysed using time-to-event methods
Time frame: Median 2 years (from randomization to 16 March 2012; maximum 2.6 years)
Once Versus Twice Daily Abacavir+Lamivudine: New WHO Stage 4 Event or Death
Number of participants with a new WHO stage 4 event or death, to be analysed using time-to-event methods
Time frame: Median 2 years (from randomization to 16 March 2012; maximum 2.6 years)
Once Versus Twice Daily Abacavir+Lamivudine: New WHO Stage 3 or 4 Event or Death
Number of participants with a new WHO stage 3 or 4 event or death, to be analysed using time-to-event methods
Time frame: Median 2 years (from randomization to 16 March 2012; maximum 2.6 years)
Once Versus Twice Daily Abacavir+Lamivudine: Height-for-age Z-score
Age-adjusted change in height-for-age Z-score over all 12-weekly visits attended over the whole follow-up, no specific timepoint prespecified. Mean and SD are time-averaged area under the change curve calculated using the trapezoidal rule. Z-scores calculated using UK norms which cover the full age range of children (Cole, T. J., J. V. Freeman, and M. A. Preece. 1998. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood. Statistics in Medicine 17(4): 407-29).
Time frame: Baseline and a median of 2 years (from once vs twice daily randomization to 16 March 2012; maximum 2.6 years)
Once Versus Twice Daily Abacavir+Lamivudine: Weight-for-age Z-score
Age-adjusted change in weight-for-age Z-score at all 12-weekly visits attended over the whole follow-up, no specific timepoint prespecified. Mean and SD are time-averaged area under the change curve calculated using the trapezoidal rule. Z-scores calculated using UK norms which cover the full age range of children (Cole, T. J., J. V. Freeman, and M. A. Preece. 1998. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood. Statistics in Medicine 17(4): 407-29).
Time frame: Baseline and a median of 2 years (from once vs twice daily randomization to 16 March 2012; maximum 2.6 years)
Once Versus Twice Daily Abacavir+Lamivudine: Body Mass Index-for-age Z-score
Age-adjusted change in body mass index-for-age Z-score at all 12-weekly visits attended over the whole follow-up, no specific timepoint prespecified. Mean and SD are time-averaged area under the change curve calculated using the trapezoidal rule. Z-scores calculated using UK norms which cover the full age range of children (Cole, T. J., J. V. Freeman, and M. A. Preece. 1998. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood. Statistics in Medicine 17(4): 407-29).
Time frame: Baseline and a median of 2 years (from once vs twice daily randomization to 16 March 2012; maximum 2.6 years)
Once Versus Twice Daily Abacavir+Lamivudine: New Grade 3 or 4 Adverse Event (AE), Not Solely Related to HIV
Number of participants with a new grade 3 or 4 adverse event (AE), not solely related to HIV, to be analysed using time-to-event methods
Time frame: Median 2 years (from once vs twice daily randomization to 16 March 2012; maximum 2.6 years)
Once Versus Twice Daily Abacavir+Lamivudine: New Serious Adverse Events Not Solely Related to HIV
Number of participants with a new serious adverse event not solely related to HIV, to be analysed using time-to-event methods
Time frame: Median 2 years (from once vs twice daily randomization to 16 March 2012; maximum 2.6 years)
Once Versus Twice Daily Abacavir+Lamivudine: Adherence to ART as Measured by Self-reported Questionnaire (Missing Any Pills in the Last 4 Weeks) at 48 Weeks
Number of participants reporting missing any doses of ART in the last 4 weeks by self-report at 48 weeks.
Time frame: 48 weeks after randomization to once- versus twice-daily
Once Versus Twice Daily Abacavir+Lamivudine: Adherence to ART as Measured by Self-reported Questionnaire (Missing Any Pills in the Last 4 Weeks) at 96 Weeks
Number of participants reporting missing any doses of ART in the last 4 weeks by self-report at 96 weeks.
Time frame: 96 weeks after randomization to once- versus twice-daily
Once Versus Twice Daily Abacavir+Lamivudine: Adherence to ART as Measured by Self-reported Questionnaire (Missing Any Pills in the Last 4 Weeks)
Binary outcome measure: missed any doses of ART in the last 4 weeks by self-report. Mean calculated across all 12-weekly visits attended over the whole follow-up (no specific timepoint prespecified), giving the percentage of visits attended where the carer/participant reported missing any pills in the last 4 weeks.
Time frame: Mean over median 2 years (from once vs twice daily randomization to 16 March 2012; maximum 2.6 years)
Cotrimoxazole: New Clinical and Diagnostic Positive Malaria
Number of participants with a new clinical and diagnostic positive malaria, to be analysed using time-to-event methods. Diagnostic positive by either microscopy (thick film) or rapid diagnostic test (RDT)
Time frame: Median 2 years (from cotrimoxazole randomization to 16 March 2012; maximum 2.5 years)
Cotrimoxazole: New Severe Pneumonia
Number of participants with a new severe pneumonia, to be analysed using time-to-event methods
Time frame: Median 2 years (from cotrimoxazole randomization to 16 March 2012; maximum 2.5 years)
Cotrimoxazole: New WHO Stage 3 or 4 Event or Death
Number of participants with a new WHO stage 3 or 4 event or death, to be analysed using time-to-event methods
Time frame: Median 2 years (from randomization to 16 March 2012; maximum 2.5 years)
Cotrimoxazole: New WHO Stage 3 Severe Recurrent Pneumonia or Diarrhoea
Number of participants with a new WHO stage 3 severe recurrent pneumonia or diarrhoea, to be analysed using time-to-event methods
Time frame: Median 2 years (from cotrimoxazole randomization to 16 March 2012; maximum 2.5 years)
Cotrimoxazole: New WHO Stage 4 Event or Death
Number of participants with a new WHO stage 4 event or death, to be analysed using time-to-event methods
Time frame: Median 2 years (from randomization to 16 March 2012; maximum 2.5 years)
Cotrimoxazole: All-cause Mortality
Number of participants who died, to be analysed using time-to-event methods
Time frame: Median 2 years (from cotrimoxazole randomization to 16 March 2012; maximum 2.5 years)
Cotrimoxazole: Weight-for-age Z-score
Age-adjusted change in weight-for-age Z-score at all 12-weekly visits attended over the whole follow-up, no specific timepoint prespecified. Mean and SD are time-averaged area under the change curve calculated using the trapezoidal rule. Z-scores calculated using UK norms which cover the full age range of children (Cole, T. J., J. V. Freeman, and M. A. Preece. 1998. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood. Statistics in Medicine 17(4): 407-29).
Time frame: Baseline and a median of 2 years (from cotrimoxazole randomization to 16 March 2012; maximum 2.5 years)
Cotrimoxazole: Height-for-age Z-score
Age-adjusted change in height-for-age Z-score at all 12-weekly visits attended over the whole follow-up, no specific timepoint prespecified. Mean and SD are time-averaged area under the change curve calculated using the trapezoidal rule. Z-scores calculated using UK norms which cover the full age range of children (Cole, T. J., J. V. Freeman, and M. A. Preece. 1998. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood. Statistics in Medicine 17(4): 407-29).
Time frame: Baseline and a median of 2 years (from cotrimoxazole randomization to 16 March 2012; maximum 2.5 years)
Cotrimoxazole: Body Mass Index-for-age Z-score
Age-adjusted change in body mass index-for-age Z-score at all 12-weekly visits attended over the whole follow-up, no specific timepoint prespecified. Mean and SD are time-averaged area under the change curve calculated using the trapezoidal rule. Z-scores calculated using UK norms which cover the full age range of children (Cole, T. J., J. V. Freeman, and M. A. Preece. 1998. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood. Statistics in Medicine 17(4): 407-29).
Time frame: Baseline and a median of 2 years (from cotrimoxazole randomization to 16 March 2012; maximum 2.5 years)
Cotrimoxazole: Change From Baseline in CD4% to Week 72
Time frame: Baseline, week 72
Cotrimoxazole: Change From Baseline in Absolute CD4 to Week 72
Estimated in those \>5 years at randomization to stop vs continue, in whom absolute CD4 is meaningful. (In uninfected children, CD4 decreases with age during early childhood.)
Time frame: Baseline, week 72
Cotrimoxazole: New Serious Adverse Events Not Solely Related to HIV
Number of participants with a new serious adverse event not solely related to HIV, to be analysed using time-to-event methods
Time frame: Median 2 years (from cotrimoxazole randomization to 16 March 2012; maximum 2.5 years)
Cotrimoxazole: Adherence to ART as Measured by Self-reported Questionnaire (Missing Any Pills in the Last 4 Weeks)
Binary outcome measure: missed any doses of ART in the last 4 weeks by self-report. Mean calculated across all 12-weekly visits attended over the whole follow-up (no specific timepoint prespecified), giving the percentage of visits attended where the carer/participant reported missing any pills in the last 4 weeks.
Time frame: Mean over median 2 years (from cotrimoxazole randomization to 16 March 2012; maximum 2.5 years)