Azithromycin as Adjunctive Treatment for Uncomplicated Severe Acute Malnutrition (AMOUR)

Amoxicillin is recommended by the World Health Organization (WHO) as adjunctive therapy for the treatment of uncomplicated severe acute malnutrition (SAM). Because children with uncomplicated SAM may have asymptomatic infection due to immune suppression, presumptive treatment with a broad-spectrum antibiotic may be beneficial by clearing any existing infection and improving outcomes. Two randomized placebo-controlled randomized trials have evaluated amoxicillin for uncomplicated SAM and have found conflicting results. These results may indicate either that antibiotics are not helpful for the management of uncomplicated SAM, or that a better antibiotic is needed. Recently, we demonstrated that biannual mass azithromycin distribution as a single oral dose reduces all-cause child mortality in sub-Saharan Africa. Children with uncomplicated SAM, who have an elevated risk of mortality relative to their well-nourished peers, may particularly benefit from presumptive azithromycin treatment. Our pilot data demonstrated feasibility in rapid enrollment of children with uncomplicated SAM in our study area, and showed no significant difference between azithromycin and amoxicillin, demonstrating equipoise for a full-scale trial. Here, we propose an individually randomized trial in which children will be randomized to a) azithromycin, b) amoxicillin, or c) placebo, and evaluated for differences in weight gain, nutritional recovery, and the gut microbiome. The results of this study will strengthen the evidence base for policy related to the use of antibiotics as part of the management of uncomplicated SAM, including additional evidence of amoxicillin versus placebo as well as evaluation of an antibiotic class that has not been considered for uncomplicated SAM, which may lead to changes in guidelines for treatment.

General study design. We propose a 1:1:1 individually randomized placebo-controlled trial in which children aged 6-59 months with SAM (based on weight-for-height Z-scores (WHZ) and/or mid-upper arm circumference (MUAC), defined in Eligibility) are randomized to one of three study arms: 1) a single oral 20 mg/kg dose of azithromycin followed by 13 doses of placebo; 2) a 7-day twice-daily course of oral amoxicillin (14 total doses); or 3) 7 days of twice-daily placebo (14 total doses). Children will be followed weekly until nutritional recovery, and then at 8 weeks (primary outcome), and 3, 6, 9, and 12 months to assess relapse and vital status. The primary outcome will be weight gain in g/kg/day at 8 weeks from enrollment. Children in all groups will receive ready-to-use therapeutic food per standard of care guidelines (described below). Study area and study team. This study will be conducted in Boromo District, Burkina Faso in West Africa. Boromo is in central Burkina Faso and experiences a large burden of SAM annually. As in much of the Sahel, food insecurity and malnutrition are highly seasonal, with the malnutrition season aligning with the rainy season from approximately July through October, prior to the annual harvest in November to December.37-39 The central Sahel, which includes Burkina Faso, is a particularly vulnerable region for childhood malnutrition due to seasonal food insecurity, ongoing political instability, and climate change which may alter or shorten growing seasons.37,40 The COVID-19 pandemic has increased risk of poor nutritional outcomes among children, particularly in already vulnerable settings.2 The Sahel, and Burkina Faso in particular, is a critical region for nutritional interventions due to the continued high prevalence of underweight and high mortality rates, and lack of progress in reducing underweight in children.41 Previous evidence has suggested that the etiology of SAM varies across sub-Saharan Africa, and that SAM in the Sahel may be more likely due to calorie insufficiency (marasmus) compared to other regions that have a higher prevalence of protein malnutrition (kwashiorkor). Amoxicillin has been hypothesized to have greater efficacy in children with kwashiorkor vs marasmus, which may partially explain discrepant results in amoxicillin trials from Malawi and Niger.7-9 Given the large burden of malnutrition in the Sahel, evidence tailored specifically to this setting is critical to inform policy. The trial will be conducted jointly by the University of California, San Francisco (PI: Dr. Catherine Oldenburg) and the Centre de Recherche en Santé de Nouna (PI: Dr. Ali Sié). Our team has collaborated on multiple randomized controlled trials for child health since 2016.42-45 In addition to expertise in design, conduct, and analysis of antibiotic trials, our team has extensive expertise in pediatric microbiome and resistome outcomes in antibiotic trials (led by Dr. Thuy Doan).18,19,45,46 Enrollment facilities. Our pilot study was conducted in 6 health facilities in Boromo over a single malnutrition season. For the full trial, we propose to expand to 18 primary healthcare facilities and enroll over a 3-year period (covering 3 malnutrition seasons). These facilities represent the first tier of the country's government-run healthcare system and provide basic preventative and curative care and are often nurse-led. Healthcare for children under 5 years of age attending public facilities is free of charge. Primary care facilities typically provide outpatient treatment of common childhood illnesses, vaccination clinics, and antenatal and maternity care. Each facility hosts a nutrition clinic 1-2 days/week during which children are screened and receive care for uncomplicated SAM on an outpatient basis. Children with a clinical complication requiring inpatient treatment will be referred to the local district hospital for treatment and will not be enrolled in the trial. Children in the outpatient nutritional program receive weekly follow-up care, although rates of default outside of trial settings are high. These facilities are typically under-resourced and experience frequent stock-outs of key components of the outpatient SAM treatment package (e.g., RUTF). All enrolled children will receive all components of outpatient SAM package through the trial. We propose a 1:1:1 randomized double masked placebo-controlled trial to determine whether a single oral dose of azithromycin is superior to 1) amoxicillin or 2) placebo for weight gain in children with SAM. Children aged 6-59 months with SAM per Burkinabé national guidelines will be randomized to one of three study arms and followed for 12 months, with the primary outcome being weight gain (g/kg/day) at 8 weeks after enrollment in the study. Children will be followed weekly until recovery, at 8 weeks, and then every 3 months to assess for relapse and mortality.

Children enrolled in the trial will be randomized to either the azithromycin, amoxicillin, or placebo arm. Children randomized to the azithromycin arm will receive all standard severe acute malnutrition (SAM) outpatient treatment per Burkinabe national guidelines, except that the standard amoxicillin treatment will be changed to azithromycin. Children will receive a directly observed dose of azithromycin (20 mg/kg, single directly observed dose, oral suspension), followed by a 7-day course of placebo (administered at 80 mg/kg, split into 2 daily doses for 7 days, oral suspension).

Children enrolled in the trial will be randomized to either the azithromycin, amoxicillin, or placebo arm. Children randomized to the amoxicillin arm will receive all standard severe acute malnutrition (SAM) outpatient treatment per Burkinabe national guidelines, including a 7-day course of amoxicillin (administered at 80 mg/kg, split into 2 daily doses for 7 days, oral suspension).

Children enrolled in the trial will be randomized to either the azithromycin, amoxicillin, or placebo arm. Children randomized to the placebo arm will receive all standard severe acute malnutrition (SAM) outpatient treatment per Burkinabe national guidelines, except that the the standard amoxicillin treatment will be changed to placebo (administered at 80 mg/kg, split into 2 daily doses for 7 days, oral suspension).

Standard of Care. A 7-day, twice-daily course of amoxicillin will be offered as a syrup to children (80 mg/kg divided into twice-daily doses, estimated with weight-based dosing).

A 7-day, twice-daily course of placebo will be offered as a syrup to children (80 mg/kg divided into twice-daily doses, estimated with weight-based dosing).

The primary outcome will be α-diversity using inverse Simpson's diversity index at the 8-week study visit.

Nutritional recovery will be defined as per Burkinabé national guidelines: weight-for-height z score (WHZ) ≥ -2 on two consecutive visits and no acute complication or edema for the past 7 days OR mid-upper arm circumfrence(MUAC)≥ 125mm on two consecutive visits and no acute complication or edema for the past 7 days. The criterion chosen to define recovery is the same as the one used to admit the child to the program.

Children will be transferred to inpatient care if they develop medical complications, or their condition deteriorates.

Vital status will be assessed at all scheduled follow-up time points, and the child's vital status (alive, died, defaulted, unknown) will be recorded in the study's mobile application. A serious adverse event form will be filled out for any child who dies during the course of the study, and the death will be reported to the medical monitor within 24 hours of the report. This report will include additional information related to the circumstances surrounding the child's death and any diagnostics and/or care sought for the child. 12 months mortality status will be compared by arm and reported

MUAC will be measured at all study visits. We will assess the change in MUAC over time across study arms using MUAC as a continuous variable. MUAC will be measured with a standard MUAC tape and measurements will be taken in triplicate. The median measurement will be used for analysis.

MUAC will be measured at all study visits. We will assess the change in MUAC over time across study arms using MUAC as a continuous variable. MUAC will be measured with a standard MUAC tape and measurements will be taken in triplicate. The median measurement will be used for analysis.

MUAC will be measured at all study visits. We will assess the change in MUAC over time across study arms using MUAC as a continuous variable. MUAC will be measured with a standard MUAC tape and measurements will be taken in triplicate. The median measurement will be used for analysis.

MUAC will be measured at all study visits. We will assess the change in MUAC over time across study arms using MUAC as a continuous variable. MUAC will be measured with a standard MUAC tape and measurements will be taken in triplicate. The median measurement will be used for analysis.

Weight-for-height (WHZ) Z-scores will be calculated for each study visit. Height will be measured with a ShorrBoard height/length measuring board with measurements taken in triplicate.

Weight-for-height (WHZ) Z-scores will be calculated for each study visit. Height will be measured with a ShorrBoard height/length measuring board with measurements taken in triplicate.

Weight-for-height (WHZ) Z-scores will be calculated for each study visit. Height will be measured with a ShorrBoard height/length measuring board with measurements taken in triplicate.

Weight-for-height (WHZ) Z-scores will be calculated for each study visit. Height will be measured with a ShorrBoard height/length measuring board with measurements taken in triplicate.

Weight-for-age (WHZ) Z-scores will be calculated for each study visit. Weight will be measured with a SECA 874 scale to the nearest 0.01kg.

Weight-for-age (WHZ) Z-scores will be calculated for each study visit. Weight will be measured with a SECA 874 scale to the nearest 0.01kg.

Weight-for-age (WHZ) Z-scores will be calculated for each study visit. Weight will be measured with a SECA 874 scale to the nearest 0.01kg.

Weight-for-age (WHZ) Z-scores will be calculated for each study visit. Weight will be measured with a SECA 874 scale to the nearest 0.01kg.

Height-for-age (HAZ) Z-scores will be calculated for each study visit. Height will be measured with a ShorrBoard height/length measuring board with measurements taken in triplicate.

At each study visit, caregivers will report if they sought medical care for their child for any reason since their last study visit, and if so, how many times they sought care, if the child was hospitalized, and the reason for seeking care (e.g., malaria, pneumonia, etc). different in number and type of clinic visits will be compared by arm and reported at 12 months

Azithromycin has been shown to have some activity against Plasmodium falciparum in vitro by targeting the plasmodial apicoplast. Some studies of azithromycin for trachoma and child mortality have shown a benefit of mass azithromycin distribution on malaria parasitemia. At baseline and 8 weeks, we will measure malaria positivity via rapid diagnostic test, tympanic temperature, and hemoglobin.

Hemoglobin will be measured in all children at baseline, 8 weeks, and 3 months using a portable Hemocue 301 system in the field. Anemia will be defined as hemoglobin <11.0 g/dL.

Hemoglobin will be measured in all children at baseline, 8 weeks, and 3 months using a portable Hemocue 301 system in the field. Anemia will be defined as hemoglobin <11.0 g/dL.

Hemoglobin will be measured in all children at baseline, 8 weeks, and 3 months using a portable Hemocue 301 system in the field. Anemia will be defined as hemoglobin <11.0 g/dL.

Nonhosts read pairs will be aligned to the MEGARes reference antimicrobial database using Burrows-Wheeler alignment with default settings. To decrease false-positive ARD identification, only ARDs with a gene fraction of >80% will be identified as present in the sample and included in analyses. We will classify each identified ARD at the class and gene level. We will evaluate the resistome at each time point by arm to evaluate the persistence of alterations to the resistome following oral antibiotics.

Long-term relapse outcomes will also be assessed at 12 months by study arm by classifying children as having SAM or recovered, regardless of whether they were considered recovered previously.

Children with uncomplicated SAM per Burkina Faso's national guidelines who present to an eligible enrollment site during the study period and meet all of the eligibility criteria below will be considered for enrollment: Inclusion criteria: Age 6-59 months WHZ<-3 SD or MUAC<115 mm Primary residence within a catchment area of an enrollment site Available for full 8-week study (primary endpoint) Not admitted to a nutritional program for SAM treatment in the previous 2 weeks No edema No antibiotic use in the past 7 days No clinical complications requiring antibiotic or inpatient treatment** No congenital abnormality or chronic debilitating illness that would lead to predictable growth faltering or reduce likelihood of SAM treatment benefit (such as cerebral palsy, Down syndrome, congenital heart disease, cleft lip/palate, etc) No known allergies to macrolides/azalides or amoxicillin/penicillin Sufficient appetite according to a feeding test with RUTF Written informed consent from at least one parent or guardian Exclusion criteria: Age less than 6 month or more than 59 months WHZ>-3 SD or MUAC>115 mm Primary residence is not within a catchment area of an enrollment site Not Available for full 8-week study (primary endpoint) Admitted to a nutritional program for SAM treatment in the previous 2 weeks Edema Antibiotic use in the past 7 days Clinical complications requiring antibiotic or inpatient treatment** Congenital abnormality or chronic debilitating illness that would lead to predictable growth faltering or reduce likelihood of SAM treatment benefit (such as cerebral palsy, Down syndrome, congenital heart disease, cleft lip/palate, etc) Known allergies to macrolides/azalides or amoxicillin/penicillin No Sufficient appetite according to a feeding test with RUTF No Written informed consent from at least one parent or guardian Per Burkinabé guidelines, children any of the following conditions will not be eligible for the trial and will be referred to an inpatient facility: MUAC <115 mm with complications; MUAC <115 mm plus edema; bipedal pitting edema; anorexia or no appetite for RUTF; diarrhea and dehydration; unable to ingest anything without vomiting; severe pneumonia; open cutaneous lesions; hypothermia (35*C); fever (38.5*C); paleness suggesting severe anemia; hypoglycemia; very weak, lethargic, or unconscious; convulsions; signs of vitamin A deficiency; or a condition requiring IV infusion or an NG tube.

The informed consent documents will contain language permitting secondary use with broad data sharing under controlled access with general use restrictions in GitHub. Any data shared will be de-identified. Participants will not be contacted or re-consented for future sharing or accessing data through repositories. Privacy and confidentiality protections will be consistent with applicable local laws in Burkina Faso. Data will be de-identified by removing all 18 HIPAA identifiers prior to sharing. All data sharing plans will be reviewed and approved by the respective institutional review boards at all participating institutions and will be reviewed during the informed consent process with caregivers. Caregivers may opt out of data sharing, for study data overall or specifically for biospecimen and/or genomic data.