Antimalarial drug efficacy and resistance in malaria‐endemic countries in HANMAT‐PIAM_net countries of the Eastern Mediterranean Region 2016–2020: Clinical and genetic studies

The World Health Organization recommends regular monitoring of the efficacy of nationally recommended antimalarial drugs. We present the results of studies on the efficacy of recommended antimalarials and molecular markers of artemisinin and partner resistance in Afghanistan, Pakistan, Somalia, Sudan and Yemen.


INTRODUCTION
Effective case management (i.e., rapid diagnosis and prompt treatment) is a critical component of recommended malaria interventions to control malaria disease [1].Unfortunately, antimalarial drug resistance continues to pose a serious threat to effective antimalarial treatment.World Health Organization (WHO) recommends artemisinin-based combination therapies (ACTs) for the treatment of uncomplicated Plasmodium falciparum [1].The combinations combine potent and fast-acting artemisinin derivatives that rapidly reduce parasite biomass with a partner compound with a long half-life that eliminates remaining parasites.The ACTs currently recommended for the treatment of uncomplicated falciparum malaria are artemether-lumefantrine (AL), artesunate-amodiaquine (ASMQ), artesunate-mefloquine (ASAQ), dihydroartemisinin-piperaquine (DP), artesunate-sulfadoxine/pyrimethamine (ASSP) and artesunate-pyronaridine (ASPY).As recommended by WHO, the therapeutic efficacy of a first-line treatment ACT below 90% should be replaced by an effective ACT with a cure rate of 95% or more [2].
In the WHO Eastern Mediterranean Region (EMR), malaria is endemic in Afghanistan, Djibouti, Pakistan, Somalia, Sudan and Yemen, with an estimated 5.7 million cases in 2020, a 33% increase from 4.3 million in 2015 [22].Estimated malaria deaths also increased (48.2%) from 8300 in 2015 to 12,300 in 2020 in EMR.Sudan accounted for most cases (56%) in the region in 2020, followed by Somalia (15%), Yemen (14%), Pakistan (10%), Afghanistan (4%) and Djibouti (1.3%).In Somalia, Sudan and Yemen, P. falciparum is the predominant species, accounting for more than 90% of infections.Malaria-endemic countries in the EMR recommended ASSP as first-line treatment for uncomplicated falciparum malaria in 2003-2005.Somalia and Sudan replaced the combination in 2016 and 2017, respectively, with AL as treatment failures with ASSP exceeded 10% [23], above the 10% threshold for a change in treatment policy [2].Afghanistan, Pakistan and Yemen also replaced ASSP with AL in 2017 [22] due to difficulties in procuring prequalified ASSP.Somalia (2016), Pakistan and Sudan (2017) recommended dihydroartemisininpiperaquine as second-line therapy, while oral quinine remains the second-line therapy in Afghanistan and Yemen.
EMRO has established two subregional networks to support member countries in monitoring antimalarial drug efficacy and resistance.These are the Horn of Africa Network for Monitoring Antimalarial Treatment (HANMAT) and the Pakistan-Iran-Afghanistan Malaria Network (PIAM-Net), established in 2004 and 2008, respectively.Countries represented in HANMAT include Djibouti, Saudi Arabia, Somalia, Sudan and Yemen in the EMR and Eritrea, Ethiopia and South Sudan in the WHO Africa region.Countries in the PIAM-Net are Pakistan, Iran and Afghanistan.Through these networks, WHO regularly assists member countries in monitoring the effectiveness and resistance of nationally recommended first-and second-line malaria treatments and in generating evidence to inform national treatment guidelines.This paper presents findings on antimalarial drug efficacy and resistance in EMR HANMAT and PIAM-net countries (Afghanistan, Pakistan, Somalia, Sudan and Yemen) between 2016 and 2020.

Study design
A single-arm prospective design was used to assess the clinical and parasitological responses to directly observed ACT treatment of uncomplicated malaria infection.Eligible patients were recruited, treated with study drug on-site, and followed up for 28 (ASSP and AL) or 42 (DP) days.At sites where two different ACTs were being tested, treatments were studied sequentially so that recruitment of patients for testing of one ACT was completed before patients were recruited for testing of a second ACT.

Study sites
Studies were conducted during the transmission seasons between 2016 and 2020 at selected sentinel sites in Afghanistan, Pakistan, Sudan, Somalia and Yemen (Figure 1).Patients were recruited from health facilities in Mehtarlam district (Laghman province) and Kama and Jalalabad districts (Nangarhar province) in Afghanistan, Zhob district (Balochistan province) and Khyber district (Federally Administered Tribal Areas) in Pakistan, Bosaso district (Bari region) in Somalia, Al-Geneina (West Darfur state), Al-Gadaref (Gadaref state), Nyala North (South Darfur state), Kas (South Darfur state), Damazin (Blue Nile state), New Halfa (Kassala state) and Sennar (Sennar) districts in Sudan.In Yemen, the study took place in a community in Bajil district in the Hodeidah governorate.

Patient enrolment
Patients with suspected malaria were enroled if they met the following inclusion criteria: age of 6 months and older, fever (axillary temperature ≥37.5 C) and/or history of fever The figure shows the study countries and their respective study sites.

TROPICAL MEDICINE & INTERNATIONAL HEALTH
during the previous 24 h, microscopically confirmed P. falciparum monoinfection (for the falciparum studies) or P. vivax monoinfection (for the vivax studies) with an asexual parasitaemia of 500-200,000/μL blood, ability and willingness to attend follow-up visits.Written informed consent was obtained from patients or parents/guardians of children before enrolment.In addition to parental/guardian consent, informed assent was obtained from minor participants aged 12 years to the age of maturity.Exclusion criteria included general danger signs (in children) or symptoms of severe falciparum malaria as defined by WHO [24], microscopically proven mixed or mono infections with non-falciparum (for the falciparum studies) or non-vivax species (for the vivax studies), severe malnutrition in children 6-60 months of age with a mid-upper arm circumference <115 mm), comorbid infections, severe anaemia and history of drug reactions.For the studies of ACT, women aged 18 years or older with a positive pregnancy test or who were breastfeeding and unable or unwilling to take a pregnancy test or use contraception were not enrolled.In addition, female minors 12-17 years of age and unmarried women 18 years of age or older were excluded because local customs and cultures do not allow them to take a pregnancy test.The above exclusion criteria do not apply to the chloroquine CQ study of P. vivax.Infected patients who did not meet the inclusion criteria were treated according to the National Malaria Control Programme (NMCP) guideline.

Treatment and follow-up
Patients treated with ASSP received a daily dose of artesunate 4 mg/kg body weight for 3 days plus a single dose of 25/1.25 mg/kg body weight SP on the first day.The ACT was manufactured by Guilin Pharmaceutical Co., Ltd.For the AL studies, patients received a dose of the combination twice daily for 3 days according to the recommended weight bands: one tablet for 5-14 kg body weight, two tablets for 15-24 kg body weight, three tablets for 25-34 kg body weight and four tablets for ≥35 kg body weight.The drug was administered with fat-containing food.It was manufactured by Novartis, for the 2016 (Afghanistan) and 2017 (Sudan) studies and by Cipla for the 2019 studies in Afghanistan, Pakistan, Sudan and Yemen.For the DP trials, patients received a daily dose of 4 mg/kg body weight dihydroartemisinin + 18 mg/kg body weight piperaquine for 3 days.The ACT was produced by Holley-Cotec Pharmaceutical Co., Ltd.(Sudan 2016) and Guilin Pharmaceutical Co., Ltd.(Sudan 2019).For the P. vivax studies, CQ (25 mg/kg body weight) divided over 3 days (Day 1-10 mg/kg, Day 2-10 mg/kg and Day 3-5 mg/kg) to patients in Afghanistan (2016) and artemether-lumefantrine (same dosing schedule as above) was administered to patients in Somalia (2018).The CQ (150 mg base tablet) was manufactured by Medopharma.All study drugs were obtained from WHO headquarters.Patients were treated under direct observation by an assigned study team member and observed for 30 minutes after the start of each dose for adverse reactions or vomiting.Any patient who vomited during this observation period was treated again with the same dose of the drug and observed for an additional 30 min.If the patient vomited again, she/he was withdrawn from the study and offered rescue therapy.Patients or their parents/guardians were asked to return for clinical and parasitological assessment on Days 1, 2, 3, 7, 14, 21, 28 (ASSP, AL and CQ treated groups), 35 and 42 (DP group) or on any unscheduled day if symptoms recurred.

Malaria microscopy examination
Thick and thin blood slides were collected from each patient during scheduled Days 1 (when the patient presents with danger signs and severe symptoms [24], 2, 3, 7, 14, 21, 28 (ASSP, AL, CQ) and 35 and 42 (DP) and unscheduled follow-up visits.Blood slides were stained with Giemsa, examined microscopically to detect malaria parasites and determine parasite density (geometric mean), and quality controlled according to WHO protocol 2009 [24].

Polymerase chain reaction correction
Dried blood spots were collected from P. falciparum patients before treatment (Day 0) and on the day of parasite recurrence beginning on Day 7 to distinguish recrudescence from new infection.DNA was extracted using DNeasy was then extracted using a DNA blood kit (Qiagen) according to the manufacturer's instructions.Paired DNA from patients with recurrent parasites (Day 0 and day of recurrence) were genotyped using nested polymerase chain reaction (PCR) targeting the highly polymorphic genes msp1, msp2 and glurp.The current WHO-recommended algorithm [25] and the two out of three (2/3) algorithm [26] were used.WHO algorithm classifies a recurrent parasite as recrudescence if at least one allele at all three markers (3/3) was common to both paired samples, while the 2/3 algorithm classifies a recurrent parasite as recrudescence, if at least one allele at two of the three markers (2/3) was common to both paired samples.Based on the current tools, we cannot differentiate between relapse, recrudescence and new infection in P. vivax.

Assessed molecular markers for antimalarial drug resistance
Parasite DNA was extracted using DNeasy kits (Qiagen).Sequencing of Pfk13 portion coding for propeller domain was performed using the previously described deoxynucleotide method [6].Amplification of the Pfpm2 and Pfmdr1 genes was assessed by RT-PCR (sybr-green dye) as previously described [9].

Treatment outcomes
Treatment response, both for P. falciparum and P. vivax studies, was classified as early treatment failure, late clinical treatment failure, late parasitological treatment failure and adequate clinical and parasitological response (ACPR) according to the protocol of WHO [2].The primary study end point was PCR-adjusted ACPR at 28 (ASSP and AL) and 42 (DP) days of follow-up.Secondary study end points included PCR-unadjusted ACPR, parasite positivity rate at Day 3, polymorphism in the Pfk13 gene and amplification of the Pfpm2 and Pfmdr-1 genes.In P. vivax studies with AL or chloroquine, patients were followed for 28 days, and the primary study end point was PCR-unadjusted ACPR because there is no method to distinguish between recrudescence, reinfection and relapsing parasites.

Data analysis
Data were double entered by two independent data clerks, cleaned and validated using the WHO Standardised data entry for therapeutic efficacy tests (http://www.who.int/malaria/areas/drug_resistance/en/).Kaplan-Meier (KM) survival to estimate the cumulative probability of ACPR up to Day 28 (CQ, ASSP and AL) and Day 42 (DP) following treatment initiation was conducted.A per-protocol analysis was also conducted, excluding patients with new infections during the follow-up period or unknown (unclassified genotyping results) to calculate the proportion of the ACPR in the PCR-corrected data.Chi-square or Fisher's exact tests were used for categorical data analysis, while independent samples t tests were employed to analyse continuous data.Statistical significance was set at p = 0.05.

Sample size
Treatment failure rate to study drugs was assumed to be 5% in the study areas.At a confidence level of 95% and 5% precision level, a minimum of 73 patients was targeted.With an additional 20%, to allow withdrawals and loss to followup during the follow-up period, a sample of 88 patients per study/site/drug was targeted in Pakistan, Somalia, Sudan and Yemen.For Afghanistan studies, at a confidence level of 95% and 10% precision level, a minimum of 50 patients was estimated due to lower rate of falciparum infection.With a 20% increase to allow loss to follow-up and withdrawals during the follow-up period, 60 patients per site and per drug were targeted.

Ethical approval
The studies were approved by the respective national ethics committees: Afghanistan National Public Health Institute, Ministry of Public Health; National Bioethics Committee, Ministry of Health Pakistan; National Ethical Review Committee, Directorate of Research, Federal Ministry of Health, Sudan; General Doctorate for Research and Studies, Ministry of Health, Yemen.For studies financially supported by WHO (Pakistan, Somalia, Sudan and Yemen), ethical approval was also obtained from WHO's Research Ethics Review Committee.Written informed consent was obtained from each patient or parent/guardian of the child before participation in the study.If the patient or parent/caregiver was illiterate, the signature of a witness and thumbprint of the patient or parent/caregiver were obtained.In addition to parental/caregiver's consent, assent was obtained from minors aged 12-17 years.For the studies of ACT, written informed consent was also required for pregnancy testing and the need for contraception in married female participants.The studies were registered with the Australian New Zealand Clinical Trials Registry.

Baseline characteristics
The baseline profile of patients in the P. falciparum studies by study year, country and site is summarised in Table 1.A total of 1805 (260 for ASSP, 1153 for AL and 392 for DP) patients with uncomplicated P. falciparum infections were recruited across 22 study arms in four countries (Afghanistan, Pakistan, Sudan and Yemen) between 2016 and 2020.For the P. vivax study in Afghanistan, 60 patients were recruited in one site to evaluate the efficacy of CQ, with a mean age of 14.4 (SD: 10.8) years, mean axillary temperature ( C) of 37.2 (SD: 0.7) and geometric parasite density (per L) of 3079 (range 232-108,768).In Somalia, 85 P. vivax patients with a mean age of 20.8 (SD: 13.8) years, axillary temperature (C) of 38 (SD: 0.4) and geometric parasite density (per L) of 7333 (range: 841-41,432) of P. vivax were recruited for evaluation of AL efficacy.

Treatment outcomes
Unadjusted PCR results at Day 28 for ASSP and AL and at Day 42 for DP are summarised in Table 2. Of the 260 patients recruited for the ASSP studies, 11 (4.2%) were lost to follow-up or withdrawn, and 249 (95.8%) met the study end points.Per-protocol PCR unadjusted ACPR of 100% was observed at Day 28 in Afghanistan (2016 and 2019) and Pakistan (2017).Of the 1153 patients enroled in the AL studies, 74 (6.4%) were either lost to follow-up or withdrawn during the follow-up period, and 1079 (93.6%) met study end points.Before PCR adjustment, an ACPR rate of 100% by Day 28 was observed for AL in Afghanistan, Pakistan, Yemen and three sites in Sudan: South Darfur and West Darfur in 2016-2017 and Sennar in 2019 (Table 2).At the remaining sites in Sudan, ACPR rates of 97.6% in 2016-2017 in Gadaref, 95.8%, 96.3% and 89.7% in 2017-2018 in Blue Nile, Kassala and Sennar, respectively, were observed.Of the 392 patients treated with DP, 352 (89.8%) met study end points and all achieved ACPR at Day 42 at all five study sites in Sudan during both study periods (Table 2).Similar ACPR rates were also observed in the KM analysis.
PCR-adjusted treatment responses (Table 3) showed an ACPR rate of 100% in patients treated with AL in all study sites except Gadaref (98.8% in 2016-2017), Kassala (97.3% in 2017-2018) and Sennar (92.1% in 2017-2018).Similar cure rates were observed in the KM analysis.Treatment outcomes of patients with recurrent parasites (n = 11) using 3/3 and 2/3 analysis are shown in Table 4.Of the 11 cases with parasite recurrence, six and five were classified as recrudescence and new infection, respectively, using the 3/3 method.The results of the 2/3 algorithm were consistent with those of the 3/3 method, except for one case in Blue Nile (2016-2017), which was classified as new infection in the 3/3 algorithm and reclassified as recrudescence in the 2/3 analysis.All recruited patients in the 22 study arms were parasite-free at Day 3, except for one case in the AL-treated group in Kassala 2017, Sudan.
Of the 60 P. vivax patients recruited to CQ, 4 were withdrawn due to species misclassification on Day 0 detected during blood slide quality control, one case had parasite recurrence on Day 21, and 55 (98.2%; 95% confidence interval: 90.4-100) patients showed ACPR on Day 28.Two of the 85 P. vivax patients treated with AL in Bosaso, Somalia, were lost to follow-up, and 83 were parasite-free at Day 28, representing an ACPR rate of 100%.

Molecular markers of antimalarial drug resistance
Regarding PfK13 polymorphism, all Day 0 samples with interpretable results from Afghanistan and Yemen carried wild type (Table 5).Non-synonymous mutations (G533D and N523I) were detected at low rate (0.5%, n = 2/371) in Pakistan in 2017, but none were found in the 2019 samples.In Sudan, the prevalence of non-synonymous Pfk13

DISCUSSION
The results of the current studies on the efficacy of ASSP in Afghanistan and Pakistan in 2016 and 2017, respectively, showed a high cure rate (100%) that persisted after more than a decade of its use.Similar high cure rates (>96%) were observed in previous studies in Afghanistan [27,28] and in Yemen [29,30].The combination was abandoned as first- line treatment in Somalia in 2016 and in Sudan in 2017 due to decline in its efficacy to below the 90% threshold recommended for a change in treatment policy [23,31,32].In Saudi Arabia, TES could not be conducted due to limited number of local falciparum infections.However, the increased prevalence of pfdhfr and pfdhps mutations with a quintuple mutation of 23.8% in samples from the Jazan region between 2018 and 2019 [33] calls for an immediate evaluation of ASSP efficacy in the country using integrated drug efficacy surveillance recommended by WHO [2].Due to low demand, there have been recent problems in procuring prequalified ASSP from Guilin, forcing Afghanistan, Pakistan and Yemen to replace ASSP with AL as the firstline treatment for uncomplicated malaria.
The current studies confirmed that AL was highly effective with PCR-corrected cure rates above 97% in Afghanistan, Pakistan, Sudan and Yemen, except for Sennar 2017-2018 study in Sudan with a lower cure rate (92.1%) but still above the 90% threshold for treatment policy change [2].The 2017-2018 Sennar study was able to enrol only 43 eligible patients, which was below the target of 88 patients.This was likely due to conducting the study towards the end (December) of the transmission season.A subsequent study in 2019-2020 at the same site showed a 100% cure rate.Other studies conducted during the same time period as the current studies in Pakistan, Somalia, Sudan and Yemen also reported high AL efficacy above 96% [28,29,31,34].Currently, AL is the first-line drug for the treatment of uncomplicated falciparum malaria in falciparum-endemic countries in the EMR.Such widespread use of this ACT may exert drug pressure on P. falciparum populations over time and could be a driving force for the development of parasites that are resistant to the drug.
Results of the current studies showed high efficacy of DP in Sudan with a cure rate of 100%.Similar high cure rates with the combination at 98.2% in Sudan [32] and 100% in Pakistan [28] and Somalia [34] were reported.Both the 3/3 [25] and the 2/3 [26] algorithms provided similar results in EMRO countries with no impact of final results of the TES, supporting the comparability of the algorithms in low to moderate transmission areas [35].The data from the current P. vivax studies showed that chloroquine and AL were highly effective in Afghanistan and Somalia, respectively, supporting their recommendations in respective countries.
Regarding molecular markers for artemisinin resistance, current studies confirm the absence of Pfk13 mutations, known to be associated with artemisinin resistance, in samples from Afghanistan, Pakistan and Yemen.Other studies conducted in Pakistan between 2015 and 2019 also did not detect Pfk13 mutations [36][37][38].In Sudan, the validated Pfk13 R622I mutant was repeatedly detected at relatively high rates at the Gadaref site in both 2016-2017 and 2019-2020 samples and observed at low frequency at the Sennar site in 2019-2020.However, neither delayed parasite clearance (parasitemia at Day 3) nor treatment failure was associated.Low frequency of the R622I mutation was detected in northwestern Ethiopia bordering the Gadaref region [39], Somalia [34], and in Chinese travellers returning from Mozambique and Somalia [40], suggesting that the mutation circulates in these countries.In addition to the R622I mutation, one case with R561H and one case with C469Y have been detected in the Kassala and Sennar sites, respectively, of Sudan.Both mutations have recently been reported to be associated with delayed parasite clearance in Rwanda and Uganda, respectively [7,8].No validated or associated Pfk13 mutations were detected in Sudan in other studies conducted between 2015 and 2017, the same time period as the current studies [41][42][43].However, none of these latter studies were conducted at the Gadaref and Sennar sites where Pfk13 R622I mutation was detected.The findings from this study highlight the need to closely monitor the efficacy of recommended ACTs, parasite clearance rates, and Pfk13 mutations in Sudan and other EMRs, as recommended by WHO [2].
Amplification of pfpm2, a marker for piperaquine resistance, was either absent (Afghanistan, Yemen) or very rare in Sudan (0.3%) in 2016/2017 and in Pakistan (0.6%) in T A B L E 4 Cure rate of the patients with parasite recurrence using 3/3 and 2/3 algorithms.

Parameters
Study sites with recurrent parasitaemia 2017.Similarly, the absence of pfpm2 amplification has been previously reported from Pakistan and Somalia [28,34].A high frequency (>30%) of pfpm2 amplification was detected in samples from Burkina Faso and Uganda [44] and Burundi [2] without treatment failure.In contrast, high rates of treatment failure following DP treatment associated with piperaquine resistance have been reported in Southeast Asian countries where artemisinin partial resistance is established [12,[15][16][17].Given that DP is used as a second-line treatment in some countries in the region, monitoring the efficacy of DP and piperaquine resistance is warranted.In Yemen, a relatively high prevalence (9.3%) of parasites with pfmdr1 amplification, a marker for mefloquine resistance, was detected, possibly related to the increased use of mefloquine for prophylaxis (Dr Methaq A, Director, National Malaria Control Programme, personal communication, May 15, 2022).Similarly, 2.7% of parasites with Pfmdr1 gene amplification were detected in Pakistan in 2017.
The subregional networks (HANMAT and PIAM-net) nurtured the antimalarial drug efficacy and resistance surveillance system in countries and have played a critical role, including timely detection of treatment failures, assisting countries with treatment policy changes, and the recent detection of the emergence of the validated Pfk13 R622I mutant in the region.Given the success of the HANMAT and PAIM-net networks, countries have requested that network activities be expanded to include other biological threats such as P. falciparum histidine-rich protein 2 and 3 gene deletions of P. falciparum (Pfhrp2/3 gene deletion), invasive vector surveillance, and insecticide resistance monitoring.Civil wars and political unrest have severely hampered TES implementation in the region, resulting in TES either being postponed (Afghanistan, Somalia, Sudan and Yemen) or abandoned (Somalia).

CONCLUSION
The studies confirmed that first-(AL) and second-line (DP) treatment of uncomplicated falciparum malaria and CQ or AL for uncomplicated P. vivax infection remain highly effective and support their recommendations in their respective countries.In Sudan, relatively high rates of the validated Pfk13 R622I mutation have been repeatedly observed without delayed parasite clearance or treatment failure.With the support of the sub-regional networks (HANMAT and PIAMnet), the efficacy of recommended ACTs, parasite clearance and drug-resistance-related markers for both artemisinin and partner drugs need to be closely monitored in Sudan and other malaria-endemic EMR countries.
HANMAT and PIAM-net should continue to support countries in the region to monitor the efficacy of recommended ACTs and resistance of artemisinins and partner drugs.
of Health, Al Mahrah, Yemen 13 Ministry of Health, Puntland, Somalia 14 Global Fund Programme, United Nations Development Programme, Kabul, Afghanistan 15 Directorate of Malaria Control, Common Management Unit Global Fund grant for Malaria Control, Ministry of T A B L E 1 Baseline characteristics of the study patients with uncomplicated falciparum infection.
Polymorphism in PfK13 gene in samples collected between 2016 and 2020.
T A B L E 7 Copy number variations in PfPM2 and Pfmdr1 genes.: DBS, dried blood spot; NI, not interpretable.TROPICAL MEDICINE & INTERNATIONAL HEALTHNational Health Services Regulations and Coordination, Islamabad, Pakistan16Department of Medical Parasitology, Faculty of Medicine and Health Sciences, Sana'a University, Sana'a, Yemen17World Health Organization, Mogadishu, Somalia18Freelance National Consult, Khartoum, Sudan19World Health Organization, Sana'a, Yemen20Global Malaria Programme, World Health Organization, Geneva, Switzerland21World Health Organization, Cairo, Egypt22School of Public Health and Community Medicine, University of Gothenburg, Gothenburg, Sweden Abbreviations