Efficacy of artemether-lumefantrine combination for the treatment of uncomplicated Plasmodium falciparum malaria after 16 years of its introduction into Ghana’s antimalarial medicines policy

Abstract

Background

Artemether-lumefantrine (AL) was introduced into Ghana’s antimalarial medicines policy in 2008 as an alternative to artesunate-amodiaquine (AS-AQ) for the treatment of uncomplicated malaria. Studies conducted between 2010 and 2021 had shown the efficacy of AL to be above 90%. This study aimed at determining the current efficacy of AL in 10 sentinel sites after 16 years of its formal introduction as one of the first-line medicines for uncomplicated malaria in the country.

Methods

The study was conducted among febrile children aged 6 months to 9 years between August 2024 and January 2025 using the WHO protocol for surveillance of antimalarial drug efficacy and reported using the standardized antimalarial therapeutic efficacy reporting (STARTER) checklist. Children enrolled were followed-up for 28 days. Parasite recrudescence and re-infections were determined using merozoite surface protein 1 (msp1), merozoite surface protein 2 (msp2), and one microsatellite (poly-α) in the recommended 3/3 PCR-genotyping approach.

Results

Out of 1387 suspected uncomplicated malaria cases screened, 747 met the inclusion criteria and were subsequently enrolled. Out of this, 738 completed their follow-up visits. PCR-uncorrected cure rates across the 10 sentinel sites ranged between 85.9% (95% CI 76.6–92.5) and 100% with a national rate of 96.3% (95% CI 94.7–97.6) whilst PCR-corrected cure rates ranged between 91.9% (95% CI 78.1–98.3) and 100% with a national rate of 98.9% (95% CI 97.8–99.5). No early treatment failures were observed at any of the sites. Also, no child enrolled was parasitaemic on Day 3.

Conclusions

AL has remained highly efficacious (> 90%) in the treatment of uncomplicated malaria in Ghana, even after 16 years of its formal introduction into the country’s antimalarial medicines policy. Preserving this high efficacy through strategies such as the deliberate implementation of multiple first-line therapy (MFT) has become critical, particularly as lumefantrine is needed for the development of non-artemisinin combination therapies in response to the spread of artemisinin resistance.

Trial Registration Pan African Clinical Trial Registry—PACTR202502791855505.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12936-026-06003-x.

Background

Artemisinin-based combination therapy (ACT) has remained the mainstay of malaria case management across the globe since 2001, when WHO recommended its use for treating uncomplicated malaria because of the advantages of enhancing efficacy, lowering malaria incidence as well as lowering the rate of emergence and spread of resistance [1]. Currently, WHO recommends the use of six ACTs in the treatment of uncomplicated Plasmodium falciparum malaria in children and adults. These ACTs are artemether-lumefantrine (AL), artesunate-amodiaquine (AS-AQ), artesunate-mefloquine (AS-MQ), dihydroartemisinin-piperaquine (DHAP), artesunate + sulfadoxine-pyrimethamine (AS-SP), and artesunate-pyronaridine (AP) [2]. Of these ACTs, AS-AQ and AL are the most widely adopted for use in sub-Saharan Africa [3, 4]. However, most African countries have over the years depended heavily on AL in the treatment of uncomplicated malaria mainly because it is well tolerated [5–8]. As far back as 2017, it was estimated that AL accounted for almost 75% of all procured quality-assured ACTs in malaria endemic countries [9].

The high dependence on AL increases the risk of treatment failure within the context of the global challenge of artemisinin resistance, which was first reported in Cambodia in 2008 [10, 11]. Artemisinin resistance has spread to all the countries in the Greater Mekong sub-region [10, 12, 13] with confirmed cases in four African countries: Rwanda, Uganda, Eritrea, and the United Republic of Tanzania [14], and suspected cases in other African countries including Namibia and Zambia [14]. Reduced efficacy of the artemisinin component of ACT has the tendency of subjecting the partner drug to increased parasite load, which subsequently leads to partner drug resistance and overall ACT treatment failure [9, 10].

In 2008, AL was introduced in Ghana as an alternative medicine for uncomplicated malaria patients who were unable to tolerate AS-AQ introduced in 2005. A study conducted in 2003 in two health facilities representing the savannah and forest zones of the country showed AL efficacy of 97.6% and AS-AQ efficacy of 100% [15]. Since then, there have been several studies that have shown AL efficacy ranging between 95.4% (95% CI 90.3–98.0) in 2010 to 97.7% (95% CI 95.4–99.1) in 2021 [16, 17]. Currently, AS-AQ, AL, and AP are the first-line medicines for uncomplicated malaria in Ghana whilst DHAP is the second-line medicine [17]. We report the therapeutic efficacy of AL studied between August 2024 and January 2025, 16 years after its official introduction as a first-line ACT in the treatment of uncomplicated malaria in the country.

Methods

Study design

The study was a one-arm, prospective, evaluation of clinical, parasitological and haematological responses among children diagnosed with uncomplicated malaria and treated with AL under direct observation. The study was conducted using the WHO protocol for surveillance of antimalarial drug efficacy [18] and reported using the Standardized Antimalarial Therapeutic Efficacy Reporting (STARTER) checklist [19] (see Additional file 1).

Study sites

The study was conducted in all 10 therapeutic efficacy study (TES) sites across the country—three in the northern belt, which is largely savannah; four in the middle belt, which is largely forest; and three in the southern belt, which is largely coastal [20] (Fig. 1). Sites in the northern belt were Wa Urban Health Centre (WUHC), Navrongo War Memorial Hospital (NWMH), and Yendi Municipal Hospital (YMH). Sites in the middle belt were Sunyani Municipal Hospital (SMH), Bekwai Municipal Hospital (BMH), Begoro Government Hospital (BGH), and Volta Regional Hospital (VRH). Sites in the southern belt were Tarkwa Apinto Government Hospital (TAGH), Ewim Polyclinic (EWP), and Ada East District Hospital (AEDH).

Fig. 1.

Map of Ghana showing study sites. WUHC: Wa Urban Health Centre; NWMH: Navrongo War Memorial Hospital; YMH: Yendi Municipal Hospital; SMH: Sunyani Municipal Hospital; VRH: Volta Regional; BMH: Bekwai Municipal Hospital; BGH: Begoro Government Hospital; TAGH: Tarkwa Apinto Government Hospital; AEDH: Ada East District Hospital; EWP: Ewim Polyclinic

The operational zoning of districts in Ghana, within the context of national malaria elimination, places six of the 10 TES sites (EWP, SMH, BGH, TAGH, YMH, and WUHC) in the stratum of districts with high malaria burden (incidence: ≥ 500 cases/1,000 population, prevalence: ≥ 15%, and under five mortality: ≥ 70); three sites (BMH, VRH, and NWMH) in the stratum of districts with moderate malaria burden (incidence: 200–499 cases/1,000 population, prevalence: 5%—< 15%, and under five mortality: 60—< 70); and one site (AEDH) in the stratum of districts with low malaria burden (incidence: 50—199 cases/1000 population, prevalence: 1%—< 5%, under five mortality: 50—< 60) [21].

Patient enrolment, treatment and follow-up

Children visiting the study sites were included in the study if they were aged 6 months to 9 years; had a history of fever during the past 24 h or axillary temperature ≥ 37.5 °C; mono-infection with Plasmodium falciparum and asexual parasite count of 1000/µL to 250,000/µL; and haemoglobin concentration > 5 g/dl. The exclusion criteria included presence of signs of severe malaria; severe acute malnutrition; presence of other causes of fever such as pneumonia, otitis media, gastroenteritis, measles, urinary tract infection; presence of chronic diseases such as cardiac, renal, and hepatic; and unwillingness of parent or legal guardian to give consent and adhere to study protocol.

Children who met the inclusion criteria were given AL (20/120 mg) – Lumartem DT^®—produced by CIPLA Ltd., India, and supplied by the NMEP. All treatment were by weight and under direct supervision by the study Nurse at times 0, 8, 24, 36, 48, and 60 h. Children weighing 5 kg to < 15 kg were given one tablet of AL per treatment hour; those weighing 15 kg to < 25 kg were given two tablets of AL per treatment hour; those weighing 25 kg to < 35 kg were given three tablets of AL per treatment hour; and those weighing ≥ 35 kg were given four tablets of AL per treatment hour. All doses were administered with milk, and children observed for 30 min to confirm retention of AL. Children who vomited within the 30-min observation time were given a repeated dose of AL and observed for another 30 min. Children who vomited repeatedly were withdrawn from the study and treated as severe malaria per national malaria case management guidelines [22]. Additionally, children failing treatment during the 28-day follow-up period were treated with an alternate ACT (either AS-AQ or DHAP).

Enrolled children were followed-up for 28 days. Each child was clinically assessed on days 0, 1, 2, 3, 7, 14, 21, 28, and on any unscheduled visit day. Drug administration was done on days 0, 1, and 2. Parasitological assessment was done on days 0, 2, 3, 7, 14, 21, 28, and on any unscheduled visit day. Haemoglobin level assessment was done on days 0 and 28.

Laboratory procedures

Malaria microscopy

Two slides with thick and thin smears were obtained during the initial screening of each patient. The first thick and thin smear slide was stained rapidly with 10% Giemsa for 8–10 min whilst the second thick and thin smear slide was retained. If the patient was subsequently enrolled, the second slide was stained with 3% Giemsa for 45–60 min for final quantification of parasites. All follow-up slides were stained with 3% Giemsa for 45–60 min. The thick smear was used to quantify parasites per white blood cells (WBCs) whilst the thin smear was used to identify parasite species. A slide was declared negative when the examination of 1000 WBCs showed no asexual parasites. Mixed infections were excluded after examining 100 fields of the thin film. Presence of gametocytes were noted after each slide examination. All slides examined by field microscopist were independently examined by a WHO-certified (level 1) microscopist. Discordant slides (in terms of presence or absence of asexual parasites or gametocytes, species identification, and Day 0 count meeting the inclusion criterion of 1000–250,000/µL blood) were independently re-examined by a second WHO-certified (level 1) microscopist. The third reading was used to confirm the final results. Parasitaemia was expressed per µL blood assuming WBC of 8000 per µL blood [23].

DNA extraction

Genomic DNA was extracted from dried blood spots (DBS) for all treatment failures: pre-treatment (Day 0) and day of parasite recurrence using the QIAamp^® Blood Mini Kit (QIAGEN, Germany) according to the manufacturer’s instructions. The extracted DNA from the samples were stored at -20 °C until use.

Plasmodium falciparum msp1, msp2 and poly-α genotyping

Genotyping of Plasmodium falciparum parasites from both pre-treatment and post-treatment samples were performed for all treatment failures to differentiate recrudescence from new infections. Polymorphic regions of merozoite surface protein 1 (msp1) and merozoite surface protein 2 (msp2) were amplified for all samples using nested polymerase chain reaction (nPCR). The msp1 allelic families analyzed were K1, MAD20, and RO33, while the msp2 allelic families were FC27 and 3D7. Primary PCR reactions were performed in a total volume of 10 μL, and secondary reactions were carried out in 15 μL volumes containing 2X Go Taq master mix (Promega, USA). Positive controls used were 3D7, 7G8, and DD2 from BEI Resources (USA). PCR products were separated by 1.5% agarose gel electrophoresis, stained with SYBR Safe (Invitrogen, USA), and fragment sizes estimated by comparison with a 100 bp molecular weight marker (Promega, USA). Alleles were scored based on the fragment size for each sample. Fragment sizes for msp1 ranged from 150 to 1000 bp and from 200 to 900 bp for the msp2.

Additionally, the microsatellite locus poly-α was analysed for all samples using semi-nested PCR. Primary PCR reactions were performed in a total volume of 10 μL consisting of 2X Go Taq master mix (Promega, USA), 0.4 μM of Forward and Reverse primers, while secondary reactions were carried out in 15 μL reaction volumes. The resulting PCR products were sent to Macrogen (Europe) for fragment analysis by capillary electrophoresis. Fragment separation was performed using an automated ABI 3730xl genetic analyzer, and allele sizes were determined using Applied Biosystems GeneMapper Software 5. To minimize the inclusion of PCR artefacts, peaks detected during microsatellite fragment analysis were considered valid only if their fluorescence intensity exceeded the analytical detection threshold and their peak height was at least one-third of the dominant peak height within the electropherogram. All electropherograms were visually inspected to confirm allele calls and ensure consistent scoring across paired samples. For poly-α, the sizes were between 122 and 189 bp. Primer sequences and cycling conditions used for msp1, msp2 and poly-α can be found in Table 1 [24, 25].

Table 1.

Oligonucleotide primers and cycling conditions for msp1, msp2 and poly-α genotyping

Marker	Primer name	Primer sequence (5’-3’)	Fluorophore Modification 5’	Cycling conditions
MSP1	M1-OF	CTAGAAGCTTTAGAAGATGCAGTATTG		95 °C for 5 min | 94 °C 1 min, 58 °C 2 min, 72 oC 2 min | 72 oC for 5 min X25 cycles
	M1-OR	CTTAAATAGTATTCTAATTCAAGTGGATCA
	K1-F	AAATGAAGAAGAAATTACTACAAAAGGTGC		95 °C for 5 min | 94 °C 1 min, 61 °C 2 min, 72 oC 2 min | 72 oC for 5 min X30 cycles
	K1-R	GCTTGCATCAGCTGGAGGGCTTGCACCAGA
	MAD20-F	AAATGAAGGAACAAGTGGAACAGCTGTTAC
	MAD20-R	ATCTGAAGGATTTGTACGTCTTGAATTACC
	RO33-F	TAAAGGATGGAGCAAATACTCAAGTTGTTG
	R033-R	CATCTGAAGGATTTGCAGCACCTGGAGATC
MSP2	M2-OF	ATGAAGGTAATTAAAACATTGTCTATTATA		95 °C for 5 min | 94 °C 1 min, 58 °C 2 min, 72 oC 2 min | 72 oC for 5 min X25 cycles
	M2-OR	CTTTGTTACCATCGGTACATTCTT
	FC27-F	GCTTATAATATGAGTATAAGGAGAA		95 °C for 5 min | 94 °C 1 min, 61 °C 2 min, 72 oC 2 min | 72 oC for 5 min X35 cycles
	FC27-R	TTTTATTTGGTGCATTGCCAGAACTTGAAC
	3D7-F	GCTTATAATATGAGTATAAGGAGAA
	3D7-R	CTGAAGAGGTACTGGTAG
Poly-α	PA-F1	AAAATATAGACGAACAGA		94 °C for 2 min | 94 °C 30 s, 42 °C 30 s, 40 oC 30 s | 65 °C for 2 min X35 cycles
	PA-R1	ATCAGATAATTGTTGGTA
	PA-R2	GAAATTATAACTCTACCA	FAM	94 °C for 2 min | 94 °C 20 s, 45 °C 20 s, 65 oC 30 s | 65 °C for 2 min X25 cycles

Classification of recurrent infections

The recommended 3/3 algorithm was used to classify a recurrent infection as either recrudescence or re-infection by comparing allelic profiles between pre-treatment (Day 0) and post-treatment (day of parasite recurrence) samples across all successfully genotyped loci. For each marker, a match was defined as the presence of at least one allele in the recurrent sample with a fragment size consistent with an allele detected in the corresponding Day 0 sample. Cases where all three loci (msp1, msp2, and poly-α) matched between pre-treatment and post-treatment samples were classified as recrudescence. Samples with amplification failure at any locus after repeated attempts were classified as indeterminate and excluded from the analyses (see Additional file 2) [26]. The 2/3 algorithm was also used as a sensitivity analysis. This involved a classification of recrudescence when two out of the three loci matched (see Additional file 2) [26].

Data management and statistical analysis

Assuming PCR-corrected treatment failure rate of 5% at each site, a precision of 5%, and a 95% confidence level, the minimum sample size was estimated to be 73 using the simple proportion formula:

$$\text{n}={\text{Z}}^2\text{p}\left(1-\text{p}\right)/{\text{d}}^2$$

where:

p = estimated PCR-corrected treatment failure rate.

d = precision.

Z = 1.96 for a 95% confidence level.

After adjusting for 20% loss to follow-up the final sample size for each site was 88.

Data was captured electronically using Research Electronic Data Capture (REDCap) application. Data was then exported to SPSS for cleaning using source documents. The cleaned data was used to populate the WHO Excel^® template for TES to generate treatment outcomes [27]. The primary treatment outcome has been described elsewhere [20]. Briefly, this is the proportion of children classified as early treatment failure (ETF), late parasitological failure (LPF), late clinical failure (LCF), and adequate clinical and parasitological response (ACPR) during the 28-day follow-up period. A combination of ETF, LPF, and LCF provides the overall treatment failure rate whilst ACPR provides success or cure rate. The patterns of these treatment outcomes (PCR-uncorrected and PCR-corrected) were described using per protocol (PP) and Kaplan–Meier (KM) survival analyses. Children lost to follow-up, withdrawn, confirmed to have violated the protocol, have non-falciparum infections were excluded from the PP PCR-uncorrected analysis whilst KM analysis censored last day of follow-up for such children. Furthermore, the PCR-corrected PP analysis excluded children lost to follow-up, withdrawn, confirmed to have violated the protocol, have non-falciparum malaria, confirmed to have reinfection, and indeterminate PCR outcomes whilst KM PCR-corrected analysis censored last day of follow-up for such children.

Another primary treatment outcome was prevalence of parasitaemia on Day 3 assessed as the proportion of children with parasites on the day after completion of the 3-day treatment (Day 0 as day of commencement of treatment).

Secondary treatment outcomes assessed included prevalence of fever (reported or temperature ≥ 37.5 °C) and parasitaemia during the first week of follow-up. Other secondary outcomes assessed were prevalence of gametocytaemia during the 28-day follow-up period, mean difference between pre-treatment (Day 0) and post-treatment (Day 28) haemolgobin levels, and prevalence of vomiting as an adverse event. Proportions were compared using Chi-square and fisher’s exact tests whilst means were compared using Student’s t-test (significant at p < 0.05).

Results

Baseline characteristics

A total of 1387 suspected uncomplicated malaria cases were screened in all ten sentinel sites. Out of this, a total of 747 children met the inclusion criteria and were subsequently enrolled. Majority of those not enrolled (78%) were microscopy negative. The total number of children enrolled per site ranged between 42 in AEDH and 88 in NWMH and BMH (see Additional file 3). Generally, there were more males than females (409/338) (Table 2). Median age of the children enrolled ranged between 3.3 years (Q1-Q3: 1.8–5.2 years) in YMH and 7.8 years (Q1–Q3: 4.3–9.0 years) in EWP with majority of the children enrolled (52.1%) aged between 5 and 9 years. Six of the 10 sites (WUHC, NWMH, SMH, VRH, TAGH, and EWP) had a greater proportion of their children aged between 5 and 9 years (Table 2). Median weight of the children enrolled ranged between 13.0 kilogrammes (Q1–Q3: 9.0–17.0 kilogrammes) in YMH and 22.0 kilogrammes (Q1 – Q3: 14.0 – 30.0 kilogrammes) in EWP whilst mean axillary temperature ranged between 37.2 °C (± 0.4) in YMH and 38.7 °C (± 0.8) in EWP (Table 2). Geometric mean parasite density ranged between 13,336/µL blood among the children enrolled in AEDH and 60,479/µL blood among those enrolled in BGH whilst mean haemoglobin levels ranged between 9.5 g/dL (± 1.6) in AEDH and 10.8 g/dL (± 1.1) in YMH (Table 2). History of previous intake of an antimalaria within the past two weeks prior to reporting at the clinic was reported in two out of the 747 children enrolled (0.3%) whilst in the case of nine children, this could not be determined (1.2%) (Table 2). The two children who were reported to have taken an antimalarial had taken AL. The child in NWMH started and ended treatment the day before her visit to the clinic whilst the child in WUHC was reported to have had 3 days of AL treatment 14 days before visiting the clinic.

Table 2.

Background characteristics of children enrolled

Characteristic	Sentinel site										TOTAL (N = 747)
	WUHC (N = 85)	NWMH (N = 88)	YMH (N = 86)	SMH (N = 47)	VRH (N = 85)	BMH (N = 88)	BGH (N = 86)	TAGH (N = 83)	AEDH (N = 42)	EWP (N = 57)
Male/Female	42/43	52/36	52/34	24/23	53/32	44/44	39/47	50/33	26/16	27/30	409/338
Median Age in years (Q1- Q3)	5.2 (3.0 – 7.6)	6.5 (3.8 – 8.3)	3.3 (1.8 – 5.2)	5.8 (2.9 – 7.8)	6.3 (4.1 – 9.2)	5.1 (2.3 – 7.8)	4.4 (2.2 – 6.3)	6.0 (2.8 – 8.7)	5.0 (2.9 – 6.9)	7.8 (4.3 – 9.0)	5.7 (3.0 – 8.3)
Age group (years)
< 5 (%)	39 (45.9)	25 (28.4)	63 (73.3)	21 (44.7)	29 (34.1)	46 (52.3)	55 (64.0)	37 (44.6)	23 (54.8)	20 (35.1)	358 (47.9)
5–9 (%)	46 (54.1)	63 (71.6)	23 (26.7)	26 (55.3)	56 (65.9)	42 (47.7)	31 (36.0)	46 (55.4)	19 (45.2)	37 (64.9)	389 (52.1)
Median weight (Kg) (Q1 – Q3)	16.0 (14.0 – 23.0)	20.0 (15.0 – 24.0)	13.0 (9.0 – 17.0)	18.5 (12.0 – 26.0)	19.0 (14.0 – 26.0)	16.0 (11.0 – 22.5)	14.0 (11.0 – 18.0)	20.0 (13.0 – 26.0)	16.0 (13.0 – 22.0)	22.0 (14.0 – 30.0)	17.0 (13.0 – 24.0)
Axillary temperature (oC)
Mean temperature (SD)	37.5 (1.1)	38.3 (0.9)	37.2 (0.4)	38.2 (1.0)	38.6 (0.6)	37.6 (1.0)	37.8 (0.8)	38.1 (0.6)	38.3 (1.1)	38.7 (0.8)	38.0 (1.0)
Range (min, max)	35.5, 40.0	36.1, 40.2	36.3, 38.0	36.1, 40.7	37.4, 40.0	36.0, 40.1	36.1, 40.0	36.6, 39.4	36.2, 40.0	37.5, 40.3	35.5, 40.7
Parasitaemia/µL
Geometric mean	44232	42,755	46,541	21,988	19,686	24,391	60,479	27,327	13,336	29,611	32,122
Range (min, max)	1095, 246037	1132, 244100	2097, 197,000	1087, 232,660	1040, 231,000	3040, 203,000	6099, 242,865	6240, 220,000	1135, 114,240	1326, 239,760	1040, 246,037
Haemoglobin level (g/dL)
Mean(SD)	10.3 (1.4)	10.5 (1.2)	10.8 (1.1)	10.2 (1.6)	10.0 (1.7)	10.4 (1.2)	10.6 (1.3)	9.8 (1.3)	9.5 (1.6)	9.7 (1.7)	10.2 (1.4)
Range (min, max)	6.0, 13.4	8.0, 12.9	9.0, 13.6	6.0, 14.1	5.5, 13.9	6.6, 13.1	5.3, 14.1	6.5, 13.3	6.9, 13.1	5.4, 13.7	5.3, 14.1
History of previous antimalarial intake
Yes (%)	1 (1.2)	1 (1.1)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	2 (0.3)
No (%)	76 (89.4)	87 (98.9)	86 (100.0)	46 (97.9)	85 (100.0)	88 (100.0)	86 (100.0)	83 (100.0)	42 (100.0)	57 (100.0)	736 (98.5)
Unknown (%)	8 (9.4)	0 (0.0)	0 (0.0)	1 (2.1)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	9 (1.2)

WUHC: Wa Urban Health Centre; NWMH: Navrongo War Memorial Hospital; YMH: Yendi Municipal Hospital; SMH: Sunyani Municipal Hospital; VRH: Volta Regional Hospital; BMH: Bekwai Municipal Hospital; BGH: Begoro Government Hospital; TAGH: Tarkwa Apinto Government Hospital; AEDH: Ada East District Hospital; EWP: Ewim Polyclinic; Q1-Q3: first and third quartiles; SD: standard deviation

Out of the 747 children enrolled in all 10 sites, 738 (98.8%) were assessed on Day 28. Children not assessed on Day 28 were from WUHC (1), NWMH (3), and AEDH (5) (see Additional file 3). The parent of the child in WUHC withdrew her consent on Day 0. In NWMH, one child was withdrawn on Day 21 because of misdiagnosis (child was reported to have tested positive for malaria and was given alternate ACT but quality checks by WHO-certified microscopists indicated a negative test result). Another child in NWMH was withdrawn on Day 0 because she convulsed, and was subsequently managed as a case of severe malaria. The third child not assessed on Day 28 in NWMH was lost to follow-up after Day 21. In AEDH, one of two children withdrawn was reported to have tested positive a day after the scheduled Day 7 visit (i.e. Day 8) and was given an alternate ACT but the test results could not be confirmed because the extra-day slide was unavailable for re-examination. The second withdrawal in AEDH was because microscopy was not done on Day 28. The other three children in AEDH were lost to follow-up after Day 3, Day 7 and Day 14, respectively.

Primary treatment outcomes

Per protocol analyses showed that PCR-uncorrected cure rates on Day 28 ranged between 85.9% (95% CI 76.6–92.5) in NWMH and 100% in WUHC, VRH, BGH, and TAGH yielding an overall rate of 96.3% (95% CI 94.7–97.6) whilst PCR-corrected cure rates, using the 3/3 algorithm ranged between 91.9% (95% CI 78.1–98.3) in AEDH and 100% in WUHC, YMH, VRH, BGH, and TAGH with an overall rate of 98.9% (95% CI 97.8–99.5) (Fig. 2 and Table 3). PCR-corrected cure rates, using the 2/3 algorithm, ranged between 90.1% (95% CI 81.5–95.6) in NWMH and 100% in WUHC, VRH, BGH, and TAGH with an overall rate of 97.4% (95% CI 96.0–98.4)—not significantly different from the estimate using the 3/3 algorithm (98.9%; 95% CI 97.8–99.5) (p = 0.055) (Fig. 2). Majority of the PCR-uncorrected and PCR-corrected treatment failures observed were late parasitological failure (LPF): 25/27 (92.6%; 95% CI 75.7–99.1) and 7/8 (87.5%; 95% CI 46.7–99.3), respectively (see Additional file 4). There were no cases of early treatment failure (ETF) at any of the 10 sites (see Additional file 4). Also, none of the sites reported parasitaemia on Day 3.

Fig. 2.

Per protocol PCR-uncorrected and PCR-corrected cure rates following treatment with artemether-lumefantrine (AL). Blue bars represent PCR-uncorrected rates, orange bars represent PCR-corrected cure rates using the 3/3 algorithm, and purple bars represent PCR-corrected cure rates using the 2/3 algorithm. WUHC: Wa Urban Health Centre; NWMH: Navrongo War Memorial Hospital; YMH: Yendi Municipal Hospital; SMH: Sunyani Municipal Hospital; VRH: Volta Regional; BMH: Bekwai Municipal Hospital; BGH: Begoro Government Hospital; TAGH: Tarkwa Apinto Government Hospital; AEDH: Ada East District Hospital; EWP: Ewim Polyclinic

Table 3.

Site-specific Day-28 PCR-uncorrected and PCR-corrected cure rates (per protocol and Kaplan–Meier using 3/3 and 2/3 approaches)

Site	PCR-uncorrected				PCR-corrected
					3/3 Approach				2/3 Approach
	PP		KM		PP		KM		PP		KM
	N	% (95% CI)	N	% (95% CI)	N	% (95% CI)	N	% (95% CI)	N	% (95% CI)	N	% (95% CI)
WUHC	84	100 (95.7–100)	84	100 (n/a)	84	100 (95.7–100)	84	100 (n/a)	84	100 (95.7–100)	84	100 (n/a)
NWMH	85	85.9 (76.6–92.5)	85	85.9 (76.6–91.8)	75	97.3 (90.7–99.7)	75	97.5 (90.5–98.4)	81	90.1 (81.5–95.6)	81	90.4 (81.7–95.1)
YMH	86	97.7 (91.9–99.7)	86	97.7 (91.0–99.4)	84	100 (95.7–100)	84	100 (n/a)	86	97.7 (91.9–99.7)	86	97.7 (91.0–99.4)
SMH	47	95.7 (85.5–99.5)	47	95.7 (84.0–98.9)	46	97.8 (88.5–99.9)	46	97.9 (85.8–99.7)	47	95.7 (85.5–99.5)	47	95.7 (84.0–98.9)
VRH	85	100 (95.8–100)	85	100 (n/a)	85	100 (95.8–100)	85	100 (n/a)	85	100 (95.8–100)	85	100 (n/a)
BMH	88	94.3 (87.2–98.1)	88	94.3 (86.9–97.6)	84	98.8 (93.5–100)	84	98.9 (92.2–99.8)	86	96.5 (90.1–99.3)	86	96.6 (89.7–98.9)
BGH	86	100 (95.8–100)	86	100 (n/a)	86	100 (95.8–100)	86	100 (n/a)	86	100 (95.8–100)	86	100 (n/a)
TAGH	83	100 (95.7–100)	86	100 (n/a)	83	100 (95.7–100)	83	100 (n/a)	83	100 (95.7–100)	83	100 (n/a)
AED	37	91.9 (78.1–98.3)	37	92.1 (77.4–97.4)	37	91.9 (78.1–98.3)	37	92.1 (77.4–97.4)	37	91.9 (78.1–98.3)	37	92.1 (77.4–97.4)
EWP	57	94.7 (85.4–98.9)	57	94.7 (84.6–98.3)	55	98.2 (90.3–100)	55	98.2 (87.8–99.7)	55	98.2 (90.3–100)	55	98.2 (87.8–99.7)

PP Per protocol, KM Kaplan–Meier, WUHC Wa Urban Health Centre, NWMH Navrongo War Memorial Hospital, YMH Yendi Municipal Hospital, SMH Sunyani Municipal Hospital, VRH Volta Regional Hospital, BMH Bekwai Municipal Hospital, BGH Begoro Government Hospital, TAGH Tarkwa Apinto Government Hospital, AEDH Ada East District Hospital, EWP Ewim Polyclinic, n/a not applicable

Kaplan–Meier survival analyses showed a 100% cumulative incidence of treatment success (PCR-uncorrected) in all sites until Day 14, when three sites (BMH, AEDH, and EWP) showed an incidence of 98.9% (95% CI 93.8–100), 97.4% (95% CI 86.5–99.9), and 98.2%, (95% CI 90.5–100), respectively. By Day 21, two additional sites (NWMH and YMH) had moved from 100% treatment success to 97.7% (95% CI 91.8–99.7) in both cases. Incidence of treatment success further declined in NWMH to 96.5% (95% CI 90.0–99.3) on Day 27 whilst the other sites maintained the Day-21 incidence. Four of the sites (WUHC, VRH, BGH, and TAGH) maintained their 100% treatment success throughout the 28-day follow-up period. Only NWMH had < 90% PCR-uncorrected treatment success on Day 28 (85.9%; 95% CI 76.6–91.8) (Table 3 and Additional file 5). PCR-correction showed similar pattern with one site (AEDH) having an incidence < 95%: 92.1% (95% CI 77.4–97.4) (Table 3 and Additional file 6).

Secondary treatment outcomes

During the first week of follow-up the overall prevalence of fever (reported fever or axillary temperature ≥ 37.5 °C) declined from 100% on Day 0 (pre-treatment) to 12.8% (95% CI 10.4–15.4) after the first day of AL treatment (Day 1). For Day 2, five of the 10 sites had no child with fever or temperature ≥ 37.5 °C. These were AEDH, EWP, VRH, TAGH, and YMH. Prevalence of fever on Day 2 ranged between 1.2% (95% CI 0.0–3.4) and 4.5% (95% CI 0.2–8.9) in the other five sites. Only one site (NWMH) had a child with fever on Day 3 whereas BMH was the only site with a febrile child on Day 7 (Fig. 3). This child was aparasitaemic and remained so throughout the follow-up period.

Fig. 3.

Proportion of children with fever (reported or axillary temperature ≥ 37.5 °C) during the first week of follow-up. Dark blue line with diamond spots represents WUHC: Wa Urban Health Centre; dark red line with square spots represents NWMH: Navrongo War Memorial Hospital; olive green line with delta spots represents YMH: Yendi Municipal Hospital; purple line with (x) spots represents SMH: Sunyani Municipal Hospital; aqua line with (⁎) spots represents VRH: Volta Regional; orange line with round spots represents BMH: Bekwai Municipal Hospital; light blue-gray line with (|) spots represents BGH: Begoro Government Hospital; light red line without spots represents TAGH: Tarkwa Apinto Government Hospital; light green line without spots represents AEDH: Ada East District Hospital; light purple line with square spots represents EWP: Ewim Polyclinic; and aqua line with square spots represents the overall prevalence of fever

During the first week of follow-up, the overall prevalence of parasitaemia declined from 100% on Day 0 (pre-treatment) to 0.9% (95% CI 0.4–1.9) after second day of AL treatment (Day 2). For the same day, no child was parasitaemic in five sites: BGH, BMH, TAGH, WUHC, and YMH. Furthermore, no child was parasitaemic in all sites on Day 3 and Day 7.

Gametocytaemia on Day 0 was present among children in three of the 10 sites: AEDH (4.8%; 95% CI 0.6–16.2), EWP (1.8%; 95% CI 0.0–9.4) and BMH (1.1%; 95% CI 0.0–6.3). By Day 7, it was only one child in BMH who was gametocytaemic (1.1% 95% CI 0.0–6.2). By Day 14 no child in any of the sites was gametocytaemic. This remained same on Day 21 and Day 28 (Fig. 4).

Fig. 4.

Proportion of children with gametocytes during the 28-day follow-up period. Dark blue line with diamond spots represents WUHC: Wa Urban Health Centre; dark red line with square spots represents NWMH: Navrongo War Memorial Hospital; olive green line with delta spots represents YMH: Yendi Municipal Hospital; dark purple line with (x) spots represents SMH: Sunyani Municipal Hospital; aqua line with (⁎) spots represents VRH: Volta Regional; orange line with round spots represents BMH: Bekwai Municipal Hospital; light blue-gray line with (|) spots represents BGH: Begoro Government Hospital; light red line without spots represents TAGH: Tarkwa Apinto Government Hospital; and light green line with (-) spots represents AEDH: Ada East District Hospital; and light purple line with diamond spots represents EWP: Ewim Polyclinic

Generally, mean Hb levels significantly increased from 10.2 g/dl (± 1.4) at baseline (pre-treatment) to 10.7 g/dl (± 1.2) on Day 28 (last day of follow-up after treatment with AL) (p < 0.001). This significant Hb increase occurred in six sites: YMH, SMH, BMH, BGH, TAGH, and EWP. There was a significant decrease in one site (WUHC), where the mean Hb level declined from 10.3 g/dl (± 1.4) to 9.9 g/dl (± 1.2) (p = 0.048). The increases observed in the remaining two sites (NWMH and VRH) were not statistically significant (Table 4).

Table 4.

Changes in mean hemoglobin levels following treatment with AL

Site	Mean Hb (SD)		P-value
	Day 0	Day 28
WUHC	10.3 (1.4) N = 85	9.9 (1.2) N = 84	0.048
NWMH	10.5 (1.2) N = 88	10.8 (1.1) N = 79	0.095
YMH	10.8 (1.1) N = 86	11.4 (0.8) N = 77	< 0.001
SMH	10.2 (1.6) N = 47	10.8 (1.2) N = 47	0.043
VRH	10.0 (1.7) N = 85	10.4 (1.2) N = 85	0.078
BMH	10.4 (1.2) N = 88	11.5 (0.9) N = 86	< 0.001
BGH	10.6 (1.3) N = 86	11.0 (0.7) N = 86	0.013
TAGH	9.8 (1.3) N = 83	10.4 (1.1) N = 83	0.002
AEDH	9.5 (1.6) N = 42	10.1 (1.0) N = 35	0.058
EWP	9.7 (1.7) N = 57	10.6 (1.2) N = 57	0.002
Overall	10.2 (1.4) N = 747	10.7 (1.2) N = 719	< 0.001

WUHC Wa Urban Health Centre, NWMH Navrongo War Memorial Hospital, YMH Yendi Municipal Hospital, SMH Sunyani Municipal Hospital, VRH Volta Regional Hospital, BMH Bekwai Municipal Hospital, BGH Begoro Government Hospital, TAGH Tarkwa Apinto Government Hospital, AEDH Ada East District Hospital, EWP Ewim Polyclinic

A total of 49 episodes of vomiting were recorded during the 3-day treatment with AL. Prevalence of vomiting during treatment was significantly greatest after administration of the first dose (30/745 or 4.0%; 95% CI 2.7–5.7) compared with the subsequent doses: 3/745 or 0.4% (95% CI 0.0–1.2) for second dose, 9/741 or 1.2% (95% CI 0.6–2.3) for third dose, 4/741 or 0.5% (95% CI 0.2–1.4) for fourth dose, 2/739 or 0.3% (95% CI 0.0–1.0) for fifth dose, and 1/739 or 0.1% (95% CI 0.0–0.8) for sixth dose. Although prevalence of vomiting after first dose of AL was higher among children less than 5 years compared with those aged 5–9 years, the difference was not statistically significant (5.6%; 95% CI 3.5–8.5 vs. 2.6%; 95% CI 1.2–4.7; p = 0.058). Similarly, although prevalence of vomiting after the third dose of AL was higher among children aged 5–9 years compared with those aged less than 5 years, the difference was not statistically significant (1.8%; 95% CI 0.7–3.7 vs. 0.6%; 95% CI 0.0–2.0; p = 0.250) (Fig. 5).

Fig. 5.

Proportion of children vomiting during treatment with AL. Blue line with diamond spots represents children < 5 years; dark red line with square spots represents children 5–9 years; and olive-green line with delta spots represents all participants

Discussions

Following the introduction of AL as one of the first-line ACTs for treating uncomplicated malaria in Ghana in 2008, there have been five efficacy studies conducted in TES sentinel sites between 2010 and 2021 [16, 17]. The 2024 study was conducted to provide an update on the therapeutic efficacy of AL after 16 years of its introduction into the antimalarial medicines policy of the country.

The study has shown that AL remains highly efficacious in all TES sites across Ghana ranging between 91.9% (95% CI 78.1–98.3) and 100%, with a national PCR-corrected cure rate of 98.9% (95% CI 97.8–99.5). This finding compares well with findings from previous studies conducted in 2021 in Ghana (96.1%—100% cure rate) [17] and studies conducted between 2021 and 2024 in other parts of Africa including Nigeria (100% cure rate) [28, 29], Togo (97.5%—98.6% cure rate) [30], Liberia (95.9–100% cure rate) [31], Niger (90.0%—100% cure rate) [31–34], Ethiopia (95.7–100% cure rate) [35–38], and Mozambique (97.6%—100% cure rate) [39]. This notwithstanding, there has been a report of cure rate < 90% (WHO threshold for policy change) in Burkina Faso following treatment with AL [40]. There is therefore the need to continue with TES activities without interruption particularly in northern Ghana, where one of the TES sites (NWMH) is located close to the border with Burkina Faso. This reinforces the need for regional/subregional collaboration in responding to the threat of antimalarial drug resistance [3].

The study also showed that none of the ten TES sites had children with parasites on day 3. This observation suggests that there is no evidence of delayed parasite clearance in the treatment of uncomplicated malaria in Ghana. This finding compares well with findings from recent studies on the efficacy of DHAP and AP in nine sentinel sites across the country [41]. Efficacy studies from some West African countries that studied AL between 2021 and 2024 have also reported Day-3 parasite positivity rates less than the WHO threshold of 10% for suspected partial artemisinin resistance [42]: Nigeria (0%) [29], Burkina Faso (1.1–1.2%) [40], Liberia (3%) [31], Togo (4.3–8.5%) [30]. This notwithstanding, the risk of spread of artemisinin and partner drug resistance in the sub-region calls for the strategic implementation of interventions such as multiple first-line therapy, where two or more effective ACTs are used together or in a rotation to treat uncomplicated malaria, and consequently reducing drug pressure for any of the ACTs being used [43].

Treatment with AL showed a good effect on fever and parasite clearance in all 10 sites. Although all children had fever on day of enrollment, only one child in one site was febrile on Day 3 following treatment. Similarly, no child was parasitaemic on Day 3 or Day 7 at any of the study sites. This observation suggests that AL remains effective in fever and parasite clearance in Ghana as in previous studies in the country and other African countries [16, 29, 39, 44–47]. Though gametocyte carriage was less than 5% in the three (3) sites with gametocytes on Day 0 (pre-treatment), no child at any of the 10 sites had gametocytes between Day 7 and Day 28, demonstrating the gametocidal effects of AL and its importance in the journey towards malaria elimination even when used alone [48, 49].

Generally, mean haemoglobin levels significantly increased from 10.2 g/dl (± 1.4) to 10.7 g/dl (± 1.2) (p < 0.001) during the 28-day follow-up period. This finding compares well with previous studies [16, 44], suggesting that AL remains effective in preventing malaria-related anaemia resulting from haemolysis of parasitized and unparasitized erythrocytes as well as bone marrow dyserythropoiesis [50].

Furthermore, AL was well tolerated as evidenced by the significant decline of vomiting after administration of subsequent treatment doses in both age-groups: from 5.6% to 0.3% among children less than 5 years and from 2.6% to 0.0% among children aged 5–9 years. The tolerability of AL has been reported in several studies [33, 44, 51].

There are four main limitations to this study. First limitation is the unavailability of pharmacokinetic data to help distinguish between clinical treatment failures due to inadequate lumefantrine concentrations and failures due to drug resistant parasites. Even though some individual studies have established statistically non-significant associations between treatment failures and sub-therapeutic Day-7 blood lumefantrine levels [40, 52, 53], evidence from a systematic review and meta-analysis suggests otherwise [54]. It is therefore worthwhile conducting pharmacokinetic analyses to help interpret treatment outcomes following treatment of uncomplicated malaria with AL. Pharmacokinetic analysis of samples obtained in this study was not done because of interrupted funding. Second limitation is the lack of molecular resistance marker data to help determine associations between markers of drug resistance and treatment failure as well as the detection of emergence and spread of resistance to better inform treatment policy. Again, interrupted funding did not allow for molecular analysis of markers of resistance. Third limitation is the less discriminatory power of the recommended markers used for molecular correction (msp1, msp2, and poly-α) [55] compared with targeted amplicon deep sequencing (TADS), which is the most robust and reliable genotyping method in high transmission settings, including Ghana [26, 56]. Failure to detect minority parasite clones in multiclonal infections using msp1, msp2 and poly-α may lead to underestimation of within-host-diversity and multiplicity of infection leading to recrudescence infections being classified as reinfections when previously undetected clones re-emerge during the follow-up period. Additionally, using poly-α may also lead to an overestimation of recrudescence infections because microsatellites have lower genetic diversity compared with other markers [56]. Fourth limitation is the small sample sizes obtained at three sites (SMH: N = 47, AEDH: N = 42, and EWP: N = 57), which limits the precision and reliability of efficacy estimates at those sites.

Conclusions

We conclude that AL remains highly efficacious (> 90%) as one of Ghana’s first-line ACTs for the treatment of uncomplicated malaria even after 16 years of its formal introduction into the country’s antimalarials medicines policy. Preserving this high efficacy through measures such as the deliberate implementation of MFT has become critical, particularly as lumefantrine is needed for the development of non-artemisinin combination therapies in response to the spread of artemisinin resistance [55].

Abbreviations

ACPR

Adequate Clinical and Parasitological response

ACT

Artemisinin-based Combination Therapy

AEDH

Ada East District Hospital

AL

Artemether-lumefantrine

AS-AQ

Artesunate-amodiaquine

AS-MQ

Artesunate-mefloquine

AS-SP

Artesunate + sulfadoxine-pyrimethamine

AP

Artesunate-pyronaridine

BGH

Begoro Government Hospital

BMH

Bekwai Municipal Hospital

COR

Corrected

DHAP

Dihydroartemisinin piperaquine

ETF

Early Treatment Failure

EWP

Ewim Polyclinic

ERC

Ethics Review Committee

GFATM

Global Fund to fight AIDS, Tuberculosis and Malaria

GHS

Ghana Health Service

Hb

Haemoglobin

IRB

Institutional Review Board

LCF

Late Clinical Failure

LPF

Late Parasitological Failure

KM

Kaplan–Meier

NMEP

National Malaria Elimination Programme

NMIMR

Noguchi Memorial Institute for Medical Research

NWMH

Navrongo War Memorial Hospital

PACTR

Pan African Clinical Trial Registry

PCR

Polymerase Chain Reaction

PP

Per Protocol

REDCap

Research Electronic Data Capture

SMH

Sunyani Municipal Hospital

STARTER

Standardized Antimalarial Therapeutic Efficacy Reporting

TADS

Targeted Amplicon Deep Sequencing

TAGH

Tarkwa Apinto Government Hospital

TES

Therapeutic Efficacy Study

UNCOR

Uncorrected

USAID

United States Agency for International Development

VRH

Volta Regional Hospital

WBC

White Blood Cell

WHO

World Health Organization

WUHC

Wa Urban Health Centre

YMH

Yendi Municipal Hospital

Author contributions

BA, PB, NYP, AA, FOA, KAK and KLM conceived and designed the study. NODN, SAM, EB, SB, KT and NQ carried out PCR-genotyping to distinguish between recrudescence and reinfection. BA, PB, AA, JIA, EOA, EG, NAY, BL, PHA, and KLM coordinated the study. BA, NODN and BL participated in the data analysis. BA drafted the manuscript. All authors read and approved the final manuscript.

Funding

The study received initial funding from the United States Agency for International Development (USAID–72064123CA0007) and final funding from the Global Fund to fights AIDS, Tuberculosis and Malaria (GFATM) through the Ghana National Malaria Elimination Programme (NMEP).

Data availability

All data analyzed during the study are included in the manuscript and supplementary information files.

Declarations

Ethics approval and consent to participate

The study was approved by the Institutional Review Board of the Noguchi Memorial Institute for Medical Research, University of Ghana (NMIMR-IRB CPN 087/23-24), and the Ethics Review Committee of the Ghana Health Service (GHS-ERC: 025/05/24). The study was retrospectively registered under the Pan African Clinical Trial Registry with an identification number PACTR202502791855505. Prior to enrollment, parents/guardians were informed of the objectives, methods, anticipated benefits and potential risks of the study to their children. They were given the opportunity to seek clarifications before signing the written informed consent form.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.