First TES in Djibouti: A Critical Milestone Amid Rising Regional Threats Authors: Sinknesh Wolde Behaksra, Samatar Kayad Guelleh, Solomon Sisay, Yisihak Abreham Weldehana, Daniel Mussa Weyesa, Tirhas Endale, Assia Adlao Hamad, Natnael Lemessa Hundera, Dawit Hailu Alemayehu, Elias B Tafa, William Louie, Andres Aranda-Diaz, Meklit Shiferaw, Matiwos Kidane Ayele, Fikregabrail Aberra Kassa, Dagim Hailegebrel Lakew, Amanuel Shimelash, Betelhem Akililu Kebede, Melat Abdo, Migbaru Keffale Bezabih, Yahya Osman Moussa, Tesfaye Sisay Tessema, Bryan Greenhouse, Abdoulilah Ahmed Abdi, Bouh Abdi Khaireh, Fitsum G Tadesse Published: 2026 ABSTRACT AbstractBackgroundMalaria control in the Horn of Africa faces growing challenges from emerging antimalarial resistance, diagnostic failures due to gene deletions, and the invasive Anopheles stephensi. In Djibouti, where artemether-lumefantrine (AL) has been the first-line treatment since 2014, no prior therapeutic efficacy studies (TES) had assessed in-country drug performance.MethodsA 28-day, single-arm TES was conducted at Arhiba Health Center in Djibouti City from 16 April 2024 to 5 May 2024, enrolling patients aged ≥6 months with uncomplicated Plasmodium falciparum malaria. Artemether-lumefantrine was administered under direct observation, with follow-up on days 1–3, 7, 14, 21, and 28 for clinical assessment, microscopy, rapid diagnostic tests (RDT), and qPCR. Molecular analyses included detection of resistance markers, gene deletions, parasite diversity, and identity-by-descent.ResultsOf 88 enrolled patients, 70 completed follow-ups. The PCR-adjusted adequate clinical and parasitological response (ACPR) was 98.6% (95% CI, 96.1–100.0), with one recrudescence. Gametocytes were observed in 11.4% at baseline and persisted through day 21. HRP2-based RDT detected only 37.5% of infections compared to 100% using pLDH-based RDT, consistent with pfhrp2/3 deletions (95.6%, 44/46). Resistance markers included Pfk13 R622I (one case), PfCRT CVIET 90.9% (70/77) and SVMNT 2.6% (2/77), PfDHFR AIRNI 40.2% (31/77) and PfDHPS ISGEAA 72.2% (52/72), and PfMDR1 NFSND 96.1% (74/77). Most infections were monoclonal (mean complexity of infection = 1.62) and exhibited high genetic relatedness.ConclusionsArtemether-lumefantrine remains effective in Djibouti; however, widespread pfhrp2/3 deletions, persistent gametocytemia, and resistance markers underscore the need for integrated molecular surveillance and the adoption of transmission-blocking interventions such as single-dose primaquine.Clinical Trials RegistrationPan African Clinical Trial Registry (PACTR202508856741773).Keywords: artemether-lumefantrine, drug resistance, horn of Africa, r622I, therapeutic efficacy BACKGROUND Malaria remains a formidable public health challenge in sub-Saharan Africa, responsible for an estimated 597 000 deaths and 263 million cases globally in 2023 with 95% of fatalities occurring in Africa [1]. While Djibouti historically maintained low malaria transmission and made commendable progress toward elimination between 2006 and 2012, the landscape has changed dramatically in the past decade [2]. A troubling resurgence, particularly in urban centers, now threatens these gains. This resurgence coincides with the detection and subsequent establishment of Anopheles stephensi, an invasive urban-adapted mosquito species, first reported in Djibouti in 2012 [3, 4]. Unlike traditional African vectors, An. stephensi thrives in urban environments and has rapidly expanded across the Horn of Africa, complicating conventional vector control efforts [5, 6]. Simultaneously, the country faces a high prevalence of Plasmodium falciparum parasites with deletions in the histidine rich protein 2 and 3 (pfhrp2 and pfhrp3) genes, rendering HRP2-based rapid diagnostic tests (RDT) ineffective. Djibouti reported unusually high prevalence of gene deletions, up to 86.5% of infections lacking the pfhrp2/3 genes [7], prompting the Djiboutian Ministry of Health to shift to pLDH-based diagnostic tools in 2022 [8]. Further compounding the crisis is the regional emergence of artemisinin resistance (ART-R), posing a direct threat to the efficacy of artemisinin-based combination therapy (ACT). Validated pfk13 mutations associated with delayed parasite clearance, such as R561H, R622I, and A675V, have been confirmed in countries in East Africa including Rwanda, Uganda, Eritrea, Ethiopia, and beyond, signaling the encroachment of ACT resistance toward Djibouti [9–11]. As a strategic hub for trade, travel, and migration, Djibouti is particularly vulnerable to the introduction and spread of resistant strains. Amid the growing entomological, diagnostic, and therapeutic challenges, robust monitoring of first-line antimalarial treatment efficacy has become essential. Djibouti first adopted artesunate-sulfadoxine-pyrimethamine (AS + SP) as the first-line therapy for uncomplicated P. falciparum in 2006, with artemether-lumefantrine (AL) designated as the second-line option [12]. In 2008, the policy shifted to artesunate–amodiaquine (ASAQ) as the first-line regimen, which remained in place until 2014. In 2014, AL was elevated to first-line status, with ASAQ reassigned as the second-line option [13]. Despite these updates, the country has never conducted therapeutic efficacy studies (TES), the gold standard for evaluating antimalarial drug performance under real-world conditions [14]. This absence stands in stark contrast to the World Health Organization (WHO) guidelines, which recommend routine efficacy monitoring as a cornerstone of malaria control and treatment policy [15]. The absence of clinical efficacy data has left the country with a critical gap of evidence to guide its malaria response. This data gap hinders the ability to detect emerging resistance and undermines the responsiveness of national malaria surveillance systems. This study marks Djibouti's first-ever TES and is being conducted at a time when the Horn of Africa is confronting a convergence of malaria biological threats including vector adaptation, diagnostic challenges, and looming ART-R. It provides the first in vivo assessment of the efficacy of AL in the Djiboutian context and will serve as an essential evidence base for national treatment policy. Moreover, it contributes to the regional and global understanding of ACT performance in an area of increasing epidemiological complexity. METHODS Study Design and Setting We conducted a prospective, single-arm, 28-day TES at Arhiba Health Center, located in Djibouti City, the capital of Djibouti, between 16 April 2024 and 5 May 2024 following the WHO protocol for antimalarial efficacy monitoring, aiming to recruit 88 patients [15]. Djibouti, located in the Horn of Africa, spans 23 200 km2 and had a population exceeding one million in 2021 of whom around 76% live in Djibouti city. Djibouti City houses most of the country's health infrastructure, including 10 community health centers, 6 polyclinics, 10 public health facilities, and 2 referral hospitals. Malaria transmission is seasonal, and the annual parasite index (API) in 2024 was 31.4 per 1000 population. Participants and Eligibility Criteria Patients aged ≥6 months with fever (axillary temperature ≥37.5°C) or a history of fever within the preceding 24 hours, confirmed P. falciparum mono-infection by microscopy and parasite densities between 500 and 200 000/μL, and hemoglobin ≥7.0 g/dL were eligible. Additional inclusion criteria included ability to take oral medications, body weight >5 kg, and provision of written informed consent (and assent when applicable). Exclusion criteria included mixed Plasmodium species infections, danger signs or severe malaria, and inability to attend follow-up visits. Treatment and Follow-Up Participants received AL (20/120 mg), dosed by weight and administered twice daily for 3 days under direct observation. Each dose was co-administered with 500 mL of Douda Milk. Patients were observed for 30 minutes postdose for adverse reactions. Vomiting within this window prompted re-dosing and observation for an additional 30 minutes. Participants were followed on days 1, 2, 3, 7, 14, 21, and 28 and were instructed to return at any time if symptoms recurred. Missed visits triggered follow-up phone calls and home visits by community health workers. Participants were excluded from the study if they missed more than one dose of AL treatment and had second vomiting during the first dose of AL administration. Finger prick blood samples collected at baseline; all follow-up and unscheduled visits were used to prepare thick and thin blood films and dried blood spot (DBS) on Whatman 3MM filter papers. At recruitment, samples were also tested using 2 RDTs: first response Bioline Ag P.f./P.v. (HRP2-based) and BIOCREDIT Malaria Ag Pf/Pv (pLDH/pLDH). At each visit, clinical assessments were conducted. Hemoglobin (SD Biosensor) level was measured on days 0, 3, 7, 14, and 28. Parasite density was determined by counting asexual parasites per 200 WBCs and converting to parasites per microliter using a standard WBC count of 8000/μL. Each slide was independently examined by 2 microscopists, and any discrepancies adjudicated by a third reader blinded to previous results. Classification of Treatment Outcomes The primary aim of the study was to measure the clinical and parasitological efficacy of AL by determining the proportion of participants with clinical and parasitological treatment failure, with adequate clinical and parasitological response constituting indicators of efficacy. Therapeutic responses were classified per WHO guidelines [15]. Safety Monitoring Adverse events (AEs) and serious adverse events (SAEs) were recorded at each visit, with grading based on clinical severity. An adverse event was defined and recorded as any unfavorable, unforeseen sign, symptom, syndrome, or illness that appeared or worsened whether or not related to that drug. Events were reported regardless of treatment attribution. Laboratory Procedures Genomic DNA extracted from DBS using the MagMAX™ DNA Multi-Sample Kit (Thermo Fisher Scientific) was used to detect and quantify P. falciparum in an 18S rRNA gene-based quantitative PCR (qPCR) [16]. Parasite density was determined using standard curves (107 to 102 parasites/mL) generated from NF54 ring-stage cultures. Next-generation sequencing (NGS) libraries were prepared by targeting 243 loci in the P. falciparum genome using the multiplex amplicon sequencing panel MAD4HatTeR [17] and Clean Plex library preparation reagents (Paragon Genomics, USA). DNA was amplified in a multiplex PCR with primer pools R1.2, R2.1 and D1.1 for 15 cycles. The amplified loci included genes implicated in drug resistance, diagnostic targets, and genetic diversity markers. Sequencing was performed on the Illumina NextSeq550 platform using a high-output kit, generating 151-cycle paired-end reads. Genotyping by capillary electrophoresis, targeting the msp1 and msp2 genes of P. falciparum, was used to classify recurrent infections as described before [18, 19]. Data Management, Statistical and Bioinformatic Analysis Study data, initially recorded on paper-based forms, were double-entered into REDCap (mobile application version 5.20.11). Discrepancies were resolved by cross-referencing with the original forms and reaching consensus among data managers. Data cleaning and validation were performed according to WHO TES protocols [15]. Baseline characteristics were summarized using proportions, geometric mean (GM) with 95% confidence intervals (CIs), and medians with interquartile ranges (IQRs). Treatment efficacy was assessed using per-protocol and Kaplan–Meier (KM) survival analyses, conducted in Stata/MP 17.0 and R version 4.4.3. Participants lost to follow-up were censored at their last visit in the KM estimates. AL efficacy was reported as proportions with 95% CIs, and only those with complete 28-day follow-up or confirmed treatment failure were included in the per-protocol analysis. Raw FASTQ files were processed using the MAD4HatTeR bioinformatic pipeline (v0.2.1) [17] to filter low-quality reads and infer sequence variants. Quality control thresholds included a minimum total read count per sample equivalent to 50 times the number of targeted loci (243), with at least 70% of the targeted loci covered, and a requirement that at least 50% of the loci in a sample had 100 or more reads. Only samples meeting these criteria were retained for downstream analysis. Complexity of infection (COI) and effective COI (eCOI) were estimated using the moire R package, utilizing data from diversity marker loci [20]. Pairwise infection-level genetic relatedness (r) was estimated using dcifer, an R package. Significance testing was performed under the null hypothesis that r < 0.125. Multiple hypothesis correction was done with the Benjamini–Hochberg method, with a false discovery rate threshold of 0.01. Relatedness networks were visualized using the igraph package in R [21]. To identify deletions of the pfhrp2 and pfhrp3 genes, targets in and around these genes were used to estimate copy numbers as previously described [17]. RESULTS Patient Recruitment and Baseline Characteristics Between 16 April 2024 and 5 May 2024, of a total of 1118 patients screened for malaria at Arhiba Health Center, 292 (26.1%) tested positive for P. falciparum (n = 148), P. vivax (n = 129), and mixed species infections (n = 15). Eighty eight patients who met the inclusion criteria were enrolled in the study, while the rest were excluded based on predefined exclusion criteria (Figure 1). Patients who completed 28-day follow-up (n = 70) were included in the final per-protocol analysis. Eighteen participants were excluded from analysis due to loss to follow-up (n = 14), protocol violation (n = 2), voluntary withdrawal (n = 1), or detection of P. vivax co-infection during follow-up (n = 1). Patient recruitment and follow-up chart. Patients screened for eligibility to join the study and those excluded are indicated together with the total number of eligible patients in Djibouti. The rows represent follow-up days (Day 01 to Day 03 and weekly from Day 07 to Day 28), with outcomes recorded at each visit day. Study participants had a median age of 22 years (IQR 18–29), with 68.2% (60/88) being male. The median body weight was 52 kg (IQR 45–59), and hemoglobin concentration was 14.4 g/dL (IQR 12.5–17.6). While nearly all participants (97.7%) reported recent fever, only 27.3% were febrile (axillary temperature ≥37.5°C) at enrollment. The GM baseline asexual parasite density was 5172 parasites/µL (95% CI 4036–6628) (Table 1). All patients were positive using pLDH-based RDT, while only 37.5% (33/88) were positive using HRP2-based RDT. Table 1.Baseline Characteristics of Study ParticipantsCharacteristicsArhiba Polyclinic (n = 88)Male sex, % (n/N)68 (60/88)Age years, median (IQR)22 (18–29)Median weight, kg (IQR)52 (45–59)Median hemoglobin, g/dL (IQR)14 (12.5–17.6)Median temperature, °C (IQR)37 (36.2–37.5)Fever (temperature ≥37.5°C), % (n/N)27 (24/88)History of fever, % (n/N)98 (86/88)pLDH-based RDT positivity, % (n/N)100 (88/88)HRP2-based RDT positivity, % (n/N)37.5 (33/88)Parasitemia (geometric mean), parasites/µL (95% CI)5172 (4036–6628)Gametocyte prevalence-microscopy, % (n/N)11.4 (10/88)Gametocytemia (geometric mean), gametocytes/µL (95% CI)182 (61–549) Baseline Characteristics of Study Participants Therapeutic Efficacy and Parasite Clearance Among 70 participants included in the per-protocol analysis, only 1 individual experienced confirmed recrudescent infection on Day 21 (Supplementary Figure 1). Overall, there was a 98.6% (95% CI: 90.2–99.8) per protocol treatment success rate. Similarly, the intention to treat KM treatment success rate was 98.6% (95% CI: 96.1–100) (Figure 2A). Rapid parasite clearance was observed, with a steep decline in positivity using both microscopy and qPCR (Figures 2B–2E). Although low-level parasitemia persisted by qPCR in some individuals, microscopy-detectable parasites were cleared in all but 2 patients by Day 3 (2.5%, 2/79). A strong correlation was obtained between qPCR-estimated total parasite load and microscopy-based asexual parasite density (R = 0.88, P < .001, Figure 2B). Treatment outcome and parasite clearance. Treatment outcomes over a 28-day follow-up period for Plasmodium falciparum is presented using a KM plot indicating probability of treatment success among eligible patients (A). Correlations in parasite load estimates using asexual-stage microscopy and 18S-based qPCR (detecting both asexual parasite and gametocytes) are presented in (B), supporting the reliability of both diagnostic methods. Parasite prevalence trends based on microscopy and qPCR (C and D) are presented together with parasite density distributions (E) for microscopy-positive cases, alongside data from patients positive by 18S qPCR. In all panels (except B) microscopy is presented in red and qPCR in blue. Numbers at risk and censored are displayed below the KM plot (A) and (C). The bar figure below (C) indicates censored study participants due to withdrawal, lost to follow-up or acquisition of non-P. falciparum infections. Fever Clearance and Hemoglobin Recovery Fever resolution was rapid, with all but one participant afebrile by Day 3. Hemoglobin recovery lagged, with 88.9% of anemic participants reaching normal levels by Day 28 (Supplementary Figure 1). No serious adverse events were reported during the study. The most common adverse event observed was headache (Supplementary Table 1). Gametocyte Dynamics At recruitment, microscopy-detected gametocyte prevalence was 11.4% (10/88), with a GM density of 182 gametocytes/µL (95% CI: 61–549). Gametocyte prevalence declined following treatment initiation, reaching 0% by Day 28 (Figure 3A). The highest prevalence and densities occurred during the early follow-up period (Days 0–7) (Figure 3B). Peak prevalence was observed on Day 3 (17.7%), followed by Day 2 (15.7%) and Day 7 (14.5%). Notably, gametocytes were frequently observed in peripheral blood in the absence of asexual parasites up to Day 21 posttreatment (Figure 3C). Furthermore, incident gametocyte emergence was observed in patients who were gametocyte-negative at baseline, with new gametocytes detected as early as Day 1 posttreatment, and persisting in some cases until Day 14. In the sole patient with recrudescent infection, gametocytes were detected on Day 1 and again on the day of failure (Day 21). Gametocyte prevalence and density over time. The bar plot represents gametocyte prevalence (microscopy) over a 28-day follow-up period (A). The error bars indicate standard errors (SE). The violin plot illustrates the density of sexual parasites (log10 parasites/µL) over time (B). Each point in (C) represents an individual measurement. Molecular Characterization and Resistance Markers Amplicon sequencing was performed on 85 baseline and the recrudescent sample, of which 81 (94.1%) samples passed quality control and included for downstream analysis (Supplementary Figure 3). A total of 218 loci had an average read of ≥100. Within-host diversity was limited, the mean COI was 1.62, and 71 (88.7%) infections were monoclonal, defined as having an eCOI ≤1.1 (Figure 4A). Genomic characteristics of Plasmodium falciparum infections at baseline. The plot in (A) shows the COI and effective COI (eCOI) distribution of samples. The dark bar represents COI and the transparent gray bars represent eCOI. eCOI represents the complexity of infection taking within host relatedness into account; thus, a sample having a high COI value might have a lower eCOI value if the strains are highly related. Most of the samples have eCOI value close to one, with the average COI being 1.62, consistent with a majority of infections being monoclonal. The relatedness between samples using IBD approach is presented in (B). Values range from 0 to 1 with 0 being not related and 1 being identical. P-value significance categories were based on FDR-adjusted values (Benjamini–Hochberg correction). Shown in (C) are the haplotype distribution of multidrug resistance genes. Parasite strains with similar gene deletions are highly related, as shown in (D). In a few instances, parasites with double pfhrp2/3 deletions and those with single pfhrp3 deletions also exhibit high relatedness. In these cases, the samples containing the pfhrp3 deleted strains were found to be polyclonal. The figure only includes samples with r > 0.5. A high prevalence of resistance-associated mutations was detected (Supplementary Table 1 and Figure 4B). The PfCRT 72–76 CVIET haplotype, linked to chloroquine resistance, was observed in 90.9% of samples, while the SVMNT haplotype associated with amodiaquine resistance appeared in 2.6% of cases. The wild-type CVMNK was identified in only 6.4% of cases. Mutations in pfdhfr and pfdhps, which together contribute to SP resistance, were widespread: 40.2% of samples carried mutant PfDHFR haplotypes AIRNI (A16I51R59N108I164), and 72.2% had the PfDHPS 436/437/540/581/613 SGEAA (S436G437E540A581A613) haplotype. These combinations represent quintuple mutant genotypes. The AICNI (A16I51C59N108I164) haplotype in PfDHFR was also present in 46.7% of samples. Sextuple mutant genotypes PfDHFR I164L, or PfDHPS A581G and A613S, were not observed. The PfMDR1 86/184/1034/1042/1246 NFSND mutant haplotype was dominant suggesting the Y184F mutation was nearly fixed (detected in 96.1% of samples). A WHO-validated ART-R marker, Pfk13 R622I, was found in only a single sample. Details of drug resistance markers detected in this study are presented in Supplementary Table 2. Gene deletions affecting diagnosis were widespread. Of 83 samples with valid data, 62.1% (51) had dual pfhrp2/3 deletions, 36.5% (30) had pfhrp3 only deletions, and no pfhrp2-only deletions were observed. Only one sample was wild type. Among HRP2-based RDT negative cases, 95.6% (44/46) carried dual deletions. Genetic relatedness analysis of 3160 pairs of baseline samples revealed that every sample was related to at least one other sample, and 90% of the samples were related to at least one other sample, with relatedness value of >0.5, suggesting localized parasite propagation with limited genetic diversity (Figure 4C). Clustering analysis showed several clusters of varying size, with pfhrp2/3 deletions segregated into different clusters, with the exception of polyclonal samples (Figure 4D). DISCUSSION This first TES of AL in Djibouti confirms its continued clinical efficacy against uncomplicated P. falciparum malaria, with a cure rate of 98.6% at Day 28. The study also highlights major challenges, including high baseline gametocyte prevalence, persistence and emergence posttreatment, together with widespread pfhrp2/3 gene deletions affecting diagnostic reliability, and molecular evidence of old and emerging drug resistance markers. Population genetics indicated a predominantly monoclonal parasite population with local propagation of distinct parasite clones. The observed high cure rate, with all but one participant achieving ACPR, mirrors outcomes from TES in Somalia, Ethiopia, and Sudan [22, 23], where efficacy remains high despite regional resistance concerns [24]. No early treatment failures were detected, only 2 patients had detectable parasites on Day 3 and this aligns with the very low prevalence of the validated Pfk13 R622I variant that is linked with delayed parasite clearance in Eritrea [11]. The presence of R622I in Djibouti, albeit at low frequency, raises concerns due to the country's role as a regional travel hub. This finding warrants intensified and continued surveillance, particularly given Djibouti's proximity to regions where Pfk13 mutations have emerged and expanded [11]. In addition to this, resistance-associated haplotypes were detected in other drug resistance loci. The SVMNT haplotype, which is typically rare or absent in neighboring Ethiopia and in Africa in general [25] and is commonly linked to amodiaquine resistance [26, 27] was detected in this study. This suggests that the haplotype may have evolved locally due to unique drug pressure. Djibouti used ASAQ as a first-line treatment between 2006 and 2014, before endorsing AL [3]. Change in treatment policies might have influenced drug resistance patterns, affecting how malaria parasites evolve [28]. The other marker detected in this study with near fixation (>90%) was the NFSND PfMDR1 haplotype, which is indirectly linked with reduced lumefantrine susceptibility in other settings [29]. The utility of this marker as a proxy for lumefantrine resistance is questionable, especially considering the continued drug pressure from chloroquine in the region related to co-endemic P. vivax [30]. Mutant haplotypes of PfDHFR and PfDHPS, associated with SP resistance, were also found at high frequencies, suggesting possible ongoing transmission of resistant strains or residual selection from past SP use [30, 31]. A key operational challenge identified in this study was the poor performance of HRP2-based RDT, driven by high rates of pfhrp2/3 deletions, 95.6% of RDT negative infections were confirmed to have double deletions. These findings support Djibouti's decision to adopt pLDH-based RDTs. Parasite population structure analyses indicated a very low complexity of infection, with most infections classified monoclonal and closely related to at least one other infection, suggesting focal transmission, and limited within-host genetic diversity. Similar patterns have been observed in Eritrea and Ethiopia underscoring the importance of regional genetic surveillance and data sharing [32, 33]. More concerning was the gametocyte dynamics observed in this study, underscoring a major limitation of AL, its limited gametocytocidal activity [34]. Gametocytes were detected in over 11% of patients at baseline, and these persisted in some individuals up to Day 21, with new gametocyte emergence observed as early as Day 1 posttreatment, emphasizing the independent kinetics of sexual and asexual stage parasites in P. falciparum. In several cases, gametocytes were present without detectable asexual parasites. This raises concerns about transmission potential even in patients classified as cured and highlights potential sustained transmission postsuccessful treatment, including resistant strains. In the recrudescent infection, gametocytes were observed both at recruitment and at the time of failure, potentially contributing to ongoing transmission risk. Given the potential for posttreatment spread of drug-resistant parasites such as R622I or pfcrt/pfmdr1 mutants, transmission-blocking strategies are worth considering. AL alone may lead to delayed gametocyte clearance, allowing malaria transmission to persist [34]. WHO currently recommends a single low dose of primaquine in areas of low transmission or known resistance emergence to break transmission chains [34, 35]. Incorporating this intervention into Djibouti's malaria policy may provide a valuable safeguard as therapeutic pressure and gene deletions increase. Taken together, this study presents the first documented evaluation of AL treatment efficacy conducted by the Ministry of Health of Djibouti. Its strengths lie in the direct administration of treatment to patients with a fatty diet, ensuring accurate absorption of lumefantrine, high follow-up retention, and the use of comprehensive molecular and diagnostic tools that contribute significantly to the reliability of its findings. Its generalizability may be limited as it was conducted at a single urban health center. Future, ex vivo susceptibility assays would further inform on emergence of resistance. In conclusion, AL remains clinically effective for uncomplicated malaria in Djibouti. However, the emergence of resistance-associated molecular markers, persistent gametocytemia, and persisting pfhrp2/3 deletions present serious threats to malaria control. Enhanced efficacy monitoring including at transportation hubs and refugee camps, molecular surveillance, and transmission-blocking interventions should be prioritized to maintain treatment efficacy and curb the spread of resistant parasites. Supplementary Material Notes Author contributions. S.W.B. and S.K.G. led execution of the study, manuscript development, and project coordination. F.G.T. conceived the study and contributed to research leadership, manuscript development and strategic oversight. S.S. and Y.A. led data analysis and interpretation. D.M.W., T.E., A.A.H., N.L.H., D.H.A., E.B.T., W.L., A.A.D., M.S., M.K.A., F.A.K., D.H.L., A.S., B.A.K., and M.A. contributed to laboratory work and field data collection. M.K., Y.O.M., T.S.T, B.G., A.A.A., and B.A.K. provided technical support and supervision. Acknowledgments. The author(s) gratefully acknowledge the support and collaboration of the Armauer Hansson Research Institution, the Horn of Africa Malaria Molecular Surveillance (HAMMS) project, and the London School of Hygiene & Tropical Medicine (LSHTM). We also extend our sincere appreciation to the Ministry of Health of Djibouti, the Ministry of Health of Ethiopia, and the World Health Organization (WHO) for their invaluable contributions to the coordination, facilitation, and implementation of this research. Ethics approval and consent to participate. The research protocol was reviewed and approved by the National Ethics Committee of the Ministry of Health, Republic of Djibouti, on 13 March 2024 (N°248/SG/2024/MS). Written informed consent was obtained from all study participants and their parents or guardians before recruitment. The study adhered to the Declaration of Helsinki principles. Data availability. The data supporting this study and codes used for analyses are available upon justified request from the corresponding author. Funding. This work was supported by the Gates Foundation (Grant Number INV-048214) awarded to the Armauer Hansen Research Institute for the Horn of Africa Malaria Molecular Surveillance (HAMMS) project. Contributor Information Sinknesh Wolde Behaksra, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia; Department of Biotechnology, Addis Ababa University, Addis Ababa, Ethiopia. Samatar Kayad Guelleh, Programme National de Lutte Contre Le Paludisme, Direction des Programmes de Santé Prioritaires, Ministère de La Santé, Djibouti, Republic of Djibouti. Solomon Sisay, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Yisihak Abreham Weldehana, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Daniel Mussa Weyesa, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Tirhas Endale, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Assia Adlao Hamad, Programme National de Lutte Contre Le Paludisme, Direction des Programmes de Santé Prioritaires, Ministère de La Santé, Djibouti, Republic of Djibouti. Natnael Lemessa Hundera, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Dawit Hailu Alemayehu, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Elias B Tafa, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. William Louie, EPPIcenter Program, Division of HIV, ID and Global Medicine, University of California, SanFrancisco, California, USA. Andres Aranda-Diaz, EPPIcenter Program, Division of HIV, ID and Global Medicine, University of California, SanFrancisco, California, USA; ISGlobal, Hospital Clínic – Universitat de Barcelona, Barcelona, Spain. Meklit Shiferaw, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Matiwos Kidane Ayele, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Fikregabrail Aberra Kassa, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Dagim Hailegebrel Lakew, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Amanuel Shimelash, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Betelhem Akililu Kebede, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Melat Abdo, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia. Migbaru Keffale Bezabih, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia; Department of Medical Microbiology, Radboud University Medical Centre, Nijmegen, The Netherlands. Yahya Osman Moussa, Clinical Biology Department, Police Health Service, Djibouti, Republic of Djibouti. Tesfaye Sisay Tessema, Department of Biotechnology, Addis Ababa University, Addis Ababa, Ethiopia. Bryan Greenhouse, EPPIcenter Program, Division of HIV, ID and Global Medicine, University of California, SanFrancisco, California, USA. Abdoulilah Ahmed Abdi, Presidency Health Council, Djibouti Armed Forces Health Services, Djibouti, Republic of Djibouti. Bouh Abdi Khaireh, Programme National de Lutte Contre Le Paludisme, Direction des Programmes de Santé Prioritaires, Ministère de La Santé, Djibouti, Republic of Djibouti; Ministry of Health, Global Fund to Fight AIDS, TB, Malaria, Djibouti, Republic of Djibouti. Fitsum G Tadesse, Malaria and NTD Division, Armauer Hansen Research Institute, Addis Ababa, Ethiopia; Department of Medical Microbiology, Radboud University Medical Centre, Nijmegen, The Netherlands; Department of Infection Biology, London School of Hygiene and Tropical Medicine, London, UK; Malaria Centre, London School of Hygiene & Tropical Medicine, London, UK. Supplementary Data Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author. References Associated Data Supplementary Materials