Clinical Ecacy and Safety of Artemisinin-based Combination Therapies in the Treatment of Uncomplicated Plasmodium Falciparum Malaria in Cameroon: A Systematic Review and Meta-Analysis from Individual Patient Data (2004-2020)

Background: Cameroon remains a high malaria endemic country. The rapid emergence and spread of Plasmodium falciparum resistant parasites compelled the World Health Organisation (WHO) to recommend the change from monotherapies to artemisinin-based combination therapies (ACTs). This study aimed to assess the clinical ecacy and safety of artemisinin-based combination therapies in the treatment of uncomplicated Plasmodium falciparum malaria in Cameroon from 2004 to 2020. Methods: The preferred reporting items for systematic review and meta-analysis (PRISMA) statement were adopted for the selection of studies. The heterogeneity of the included studies was determined using Cochrane Q and the I 2 . The random effects model was used as standard to combine studies showing heterogeneity of Cochrane Q with P < 0.10 and I 2 > 50. Results: Out of the 4,920 articles and unpublished datasets screened, 16 records with a sample size of 3,737 participants on 8 generic ACTs fullled the inclusion criteria. The per protocol (PP) analysis pooled ecacy of the ACTs was 97.9 % (95 % CI, 97.2-98.7, P<0.01). Sub-group analyses were performed for ASAQ, AL and DHAP. The aggregated ecacies of ASAQ, AL and DHAP were 97.5 % (95 % CI, 96.3-98.8, P<0.01), 99.4 % (95 % CI, 98.6-100.0, P=0.39), 98.0 % (95 % CI, 96.3-99.7, P=0.28) respectively. The pooled ecacies were above the WHO minimum benchmark of 90.0%. The ACTs are well tolerated and common adverse events reported were asthenia, diarrhoea, abdominal pain, anorexia, nausea and vomiting, headache and dizziness. Conclusion: This study reported a high pooled ecacy for all ACTs. AL and DHAP were found to have higher cumulative ecacies than ASAQ. The ACTs are still ecacious and well tolerated for the treatment of malaria in Cameroon. However, there is need for continuous monitoring of ecacy of ACTs despite the high cure rates as resistance


Background
Malaria is a major public health disease in Cameroon with Plasmodium falciparum responsible for most of the cases [1]. In 2018, malaria accounted for 228 million cases and 405,000 related deaths worldwide (1). Early diagnosis and treatment of clinical cases remain the main tools for the control of malaria in several regions of Africa [2]. In Cameroon, until 2002, the rst-line recommended therapy for uncomplicated falciparum malaria was chloroquine (CQ) and later amodiaquine (AQ) monotherapy between 2002 and 2004 [3]. However, this policy was threatened by the emergence and spread of Plasmodium falciparum resistance to chloroquine (CQ), amodiaquine (AQ) and sulphadoxinepyrimethamine (SP) in most malaria endemic countries. This resulted in major challenges to malaria control in sub-Saharan Africa [2,4]. Parasite resistance to monotherapies compelled WHO to recommend combination of dual or triple therapy, which combines molecules with independent modes of action or distinct target enzymes [5]. Therefore, WHO recommended the adaptation and implementation of artemisinin-based combination therapies (ACTs) as the rst-line treatment for malaria since early 2000 in most countries with endemic P. falciparum malaria (6). Five ACTs are currently recommended by WHO for the treatment of uncomplicated Plasmodium falciparum infection: artemether-lumefantrine (AL), artesunate-amodiaquine (ASAQ), artesunate-me oquine (ASMQ), artesunate-sulphadoxinepyrimethamine (ASSP), dihydroartemisinin-piperaquine (DHAP) [7]. The basis for the use of ACT relies on the rapid reduction of the parasite biomass, reduction of transmission (reducing gametocytes), protection of partner drug against resistance, and rapid fever reduction [8]. In January 2004, Cameroon o cially aligned with the recommendations of WHO and adopted artesunate-amodiaquine (ASAQ-75%) and later included artemether-lumefantrine (AL-25%) in 2006 as rst-line treatment of uncomplicated malaria [9]. These drugs are distributed by those proportions in public facilities, while AL is relatively predominant within the private health facilities and vendors [10]. A recent network meta-analysis (NMA) study on ACTs in Cameroon revealed that AL was more e cacious than ASAQ [11]. The advantage of the NMA approach is that it provides estimates of the effect of each intervention relative to each other [12]. However, the method adopted for the NMA study is not without limitations namely: non-inclusion of observational studies, use of intention-to-treat (ITT) approach and the non-adoption of the individual patient data (IPD) in quantitative syntheses [11]. The adoption of multiple rst line ACTs has the potential to delay the emergence of parasites resistant to the anti-malarials [13]. This study was done to provide additional information on the pooled per protocol (PP) e cacy of ACTs using IPD. Hence, this study aimed to assess the clinical e cacy and safety of Artemisinin-based combination therapies in Cameroon from January 2004 to June 2020 with emphasis on ACTs adopted for rst-line treatment of uncomplicated falciparum malaria.

Methods/design
Searching strategies Studies included in this review were selected using the preferred reporting items for systematic review and meta-analysis (PRISMA) statement [14]. A computerised systematic strategy based on key words was used to search articles from PubMed/Medline, Google Scholar, and Science Direct databases. Both interventional and observational studies were retrieved to be included in the review using the following MeSH search terms: 'Cameroon AND malaria AND artemether-lumefantrine', 'Cameroon AND malaria AND artesunate-amodiaquine', 'Cameroon AND malaria AND artesunate-me oquine', 'Cameroon AND malaria AND dihydroartemisinin-piperaquine', 'Cameroon AND malaria AND artesunate-sulphadoxinepyrimethamine', 'Cameroon AND malaria AND artesunate-atovaquone-proguanil', 'Cameroon AND malaria AND artesunate-sulphamethoxypyrazine-pyrimethamine', and 'Cameroon AND malaria AND artesunatechloguanil-dapsone'. Additional information on Clinical Trials was also obtained from the libraries and from researchers at the Universities of Yaounde I, Buea, Douala, Dschang and Bamenda. The library of the Catholic University of Central Africa based in Yaounde Cameroon was also consulted. Moreover, information was obtained from the OCEAC Bulletin and from the National Malaria Control Programme (NMCP), the Ministry of Public Health annual reports. Furthermore, studies conducted in any of the four sentinel sites in Cameroon were sought from the NMCP or from individual researchers who provided this information voluntarily. In addition to published studies, unpublished thesis reports were accessed for inclusion in the study.

Inclusion criteria
The following studies were included in this systematic review and meta-analysis: original articles of studies that investigated at least one ACT in the treatment of uncomplicated falciparum malaria; studies published that included study periods from January, 2004 to June, 2020; studies written in English or French; all multi-centric studies in which Cameroon was one of the sites were included in this report. The population intervention comparator outcome (PICO) format was used to select and include studies (Additional le 1). The primary objective of this review was to assess the e cacy of ACT measured as treatment success at days 14, 28, 42 or 63 for uncomplicated malaria caused by Plasmodium falciparum, while the frequency of adverse events (AEs) was the secondary objective. AEs were de ned as 'signs and symptoms that rst occurred or became more severe post-treatment' or 'as a sign, symptom, or abnormal laboratory value not present on day 0, but which occurred during follow-up, or was present on day 0 but became worse during follow up'. Serious adverse events were de ned according to International Conference on Harmonisation (ICH) guidelines. Studies included in this review are shown in Additional le 1.

Non-inclusion criteria
The following papers were excluded from this systematic review and meta-analysis: studies that used artemisinin monotherapies or non-artemisinin monotherapies; non-artemisinin combination therapies; studies that assessed malaria treatment outcomes at times less than 14 days and studies with PCR unadjusted cure rates. Studies that were excluded from this review are shown in Additional le 1.

Review process
All of the research articles identi ed from searches of the electronic databases were screened for eligibility based on title and abstract. Ineligible articles and duplicates were eventually removed using Zotero standalone software version 5.0.56. Full-length articles of the selected studies were read to con rm for ful lling of the inclusion criteria before data extraction began. Two reviewers (PTNN and CMM) independently screened the titles and abstracts to identify potentially eligible studies and data extracted from full-length articles that ful lled the inclusion criteria ( Figure 1). Discrepancies were resolved by mutual consent after discussion and independent review from the third researcher (AMN). WFM reviewed the whole process.
Data extraction procedure Data on the types of study design (observational versus interventional), year the studies were conducted, duration of study, and geographic location of the study area was rst extracted. Participants' age ranges were then extracted. Finally, data regarding the types of anti-malarial treatments, treatment outcome measures (including treatment success rates, treatment failure rates), treatment duration, and adverse events (AEs) were extracted to be included in the systematic review and meta-analysis.

Methodological quality assessment and sensitivity analysis
The quality of the reviewed studies was assessed through sensitivity analysis, which classi ed the included studies into high quality and low quality according to modi ed Jadad scale for randomised controlled trials (RCTs) [15] and the strengthening the reporting of observational studies in epidemiology (STROBE) statement for observational studies [16]. Modi ed Jadad scale assesses the quality of a trial with the range from 0 to 8 (randomisation and its appropriate use, blinding and its appropriate use, withdrawals and dropouts, description of inclusion and exclusion criteria, assessment of adverse effects, and description of statistical analysis). The score for the modi ed Jadad scale range of 0-3 represents low or poor quality and score ranges of 4-8 represents good to excellent quality. The observational studies were categorised as low quality with a score under 75% of the STROBE checklist and high quality with a score over 75% of the STROBE checklist. The reviewers independently assessed the quality of the methodology of included studies.

Assessment of treatment outcomes
Treatment outcome was assessed as treatment failure and treatment success. The outcomes of all the studies included in this review were assessed and analysed on the 14 th , 28 th , 42 nd and 63 rd day of treatment. Treatment failure included: early treatment failure (ETF), late parasitological failure (LPF) and late clinical failure (LCF). The indicator for treatment success was adequate clinical and parasitological response (ACPR). ACPR was de ned as absence of parasitaemia by the end of treatment (days 14, 28, 42, 63) irrespective of axillary temperature without previously meeting any of the criteria for early treatment failure or late clinical failure or late parasitological failure (17)(18)(19)(20). The treatment success was de ned based on PCR genotyping according to current World Health Organisation (WHO) recommendation.

Publication bias
Publication bias was assessed using funnel plot with the standard error of each study plotted against its effect size (Additional le 2). The Egger test was also used to assess publication bias.

Data analysis and heterogeneity assessment
Page 7/24 The traditional meta-analysis that estimates a common effect of the same intervention A, B and C by pooling individual patient data (IPD) from various studies was adopted. The R software package version 3.5.2 was used to carry out all the meta-analyses of malaria treatment e cacy. The heterogeneity of the included studies was investigated using Cochrane Q and the I 2 . The random effects model was used as standard to combine studies showing heterogeneity of Cochrane Q with P < 0.10 and I 2 > 50 [21]. Heterogeneity using was classi ed as low (0-49%), moderate (50-74%) and high (75-100%).

Ethical considerations
The PRISMA guideline recommendations were used and strictly followed to carry out this systematic review and meta-analysis. Ethical approval is not recommended and was not needed since it is a systematic review and meta-analysis.

Study identi cation and selection process
A computerised systematic strategy was used to identify and screen articles from PubMed, Google Scholar, and Science Direct databases for eligibility ( Figure 1).

Qualitative synthesis
The search of studies published from 2004 to 2020 identi ed 13 articles important to the topic under review [22][23][24][25][26][27][28][29][30][31][32][33][34], out of which 10 were RCTs [22-30, 33, 34] and 3 non-comparative clinical trials without randomisation [31, 32] (Additional le 1). Data from 3 unpublished studies that were conducted by the Cameroon National Malaria Control Programme and independent researchers were also included in this review. A total of 39 studies from 16 records with the same or different ACTs ful lled the inclusion criteria and were included in this systematic review and meta-analysis with a total sample size of 3,747 participants that ranged from a minimum of 48 patients [22]

Fever and parasite clearance by ACTs
Fever temperature measurements were mostly axillary and on average were above 38 °C . Fever was rapidly cleared by the ACTs with low proportions with fever on D1 of 17 % and 5 % by D3 (Additional le 5). Some authors did not measure fever clearance on subsequent days post drug administration and only choose D3 for this clinical measurement.
Starting parasitaemia with geometric mean densities varied between 3800 to 42000 parasites per ml. Parasite clearance was very rapid sometimes dropping to 7-8 % on D1 but on average stayed at 30-50 %. Parasite clearance times were similar across study arms in most studies. Nji et al., observed that there appeared to be a study site effect on parasite clearance between Garoua and Mutengene [30]. Although not signi cant most participants in Mutengene site (75 %) (Figure 8), did not clear their parasites as measured by microscopy by the end of D1 post-treatment compared to participants in Garoua (32 %). Comparing the parasite clearance time across the different treatment arms and site did not show any signi cant difference by D3 post-treatment (p>0.05). Overall, the proportion of patients with parasites on D2 was much lower than on D1 and even lower on D3 (Figure 8). Apinjoh et al., found a high occurrence of parasitaemia of 17% and 35% respectively in ASSP and ASAQ arms in Buea-Tole on D28 [33] and from which study an e cacy of ASAQ of about 92 % was recorded. The average parasite clearance did not take into account the two studies with outlier gures (Figure 8 and Figure 9). Only two of these studies were interested to have measured parasite levels up till D42 or D63 [30,32]. It should be noted that parasitaemia tended to re-occur between D14-D63.
A total of 12 (76.5 %) articles and unpublished data reported adverse events [23,24,26,[28][29][30][31]33]. The ACTs are well tolerated with few adverse events, the most reported across all studies being those of asthenia, diarrhoea, abdominal pain, anorexia, nausea, vomiting, headache and dizziness. Out of 12 studies, 5 reported severe adverse events which included: 1 jaundice, 1 haemoglobulinuria, 3 anaemia, 1 convulsion, 1 severe fatigue, 3 severe malaria and 3 deaths (Additional le 5). A three-arm randomised non-inferiority controlled trial to assess the e cacy and safety of ASAQ, DHAP and AL was conducted in Mutengene and Garoua from 2009 to 2013. The authors demonstrated that the frequency of adverse events such as vomiting, cough, rashes, and anorexia was slightly higher in the groups of participants on ASAQ and DHAP. The drugs did not differ with respect to the type of AEs (all p values <0.05). Although there was no signi cant statistical difference (P=0.09) in the occurrence of all AEs when comparing the trial drugs, ASAQ (35.5 %) and DHAP (37.9 %) had higher number of AEs than AL (27.5 %). One serious AE occurred involving a child who experienced severe fatigue after AL ingestion [30].
In another study in Buea, it was observed that while administering ASAQ and SPAS, at least one adverse event (AE) was reported in 69.2 % (117/169) of patients in both treatment groups during the posttreatment period that was not present on admission. These were probably related to the study drug and mainly mild or moderate in intensity. The most frequent AEs were cough, dizziness, fatigue, catarrh, and gastrointestinal disorders (nausea, abdominal pain, and diarrhoea). A total of 36 (43.4 %) and 81 (94.2 %) patients in the AS/SP and AS/AQ group, respectively, experienced AEs by day three. By day seven, the number of patients with AEs had reduced to two (4.5 %) and eight (17.7 %) in the AS/SP and AS/AQ groups, respectively [33].

Quality assessment and sensitivity analysis
The articles included in this review were of high qualities according to modi ed Jadad scale for randomised controlled trials (RCTs) with values ranging from 5 to 8 (Additional le 3) while and the strengthening the reporting of observational studies in epidemiology (STROBE) statement for observational studies with a range of 86 % to 91 % (Additional le 4). There was no need for sub-group analysis because of the high quality of studies included.

Discussion
The systematic review and meta-analysis aimed to assess the pooled clinical e cacy and Safety of artemisinin-based combination therapies from individual participant data for over 16 years after adoption and use for the treatment of uncomplicated falciparum malaria in Cameroon. We demonstrated that the cure rates of most ACTs were above the WHO minimum limit of 90% with a cumulative value of 97.9%. This observation is in agreement with cure rate of 98.0% reported in meta-analysis on ACTs used in Sudan [35], but higher than the e cacy of 92.9% recorded in Ethiopia [36]. However, 2 studies on Artesunate-chloguanil-dapsone (AS-CD) and Artesunate-sulphadoxine-pyrimethamine (AS-SP) failed to meet the WHO, Day 28, PCR-adjusted cut-off of > 90% e cacy [27,33]. Despite the high success rates of ACTs, resistance to anti-malarial drugs poses a major threat globally and if the parasites develop resistance to these anti-malarial regimens, it is inevitable that treatment would be more di cult, unsuccessful and high rates of relapse could be due to multidrug-resistant malaria. Therefore, monitoring the anti-malarial drug e cacy is important to enable early detection of emergence of drug resistance before it spreads to most of the parasite population, similarly to what happened with chloroquine, sulphadoxine-pyrimethamine, and amodiaquine monotherapies in Cameroon [2,37,38]. More so, there is no guarantee how long the currently used anti-malarial drugs will remain effective based on evidence that previous monotherapies were associated with higher rates of treatment failures in P. falciparum infected patients [2,39]. This calls for concerted action for search of new alternative anti-malarial drugs to treat malaria in the near future.
The e cacy of ASAQ the rst line treatment adopted in 2004 for the treatment of uncomplicated falciparum malaria in Cameroon was 97.5%. Contrary, a slightly lower value of 93.9% was recorded in a multi-centric study on pooled e cacy of ASAQ in sub-Saharan Africa [40]. Moreover, the e cacy of AL, the parallel rst-line drug for the treatment of uncomplicated falciparum malaria in Cameroon was 99.4%. This is in concordance with the 97.3% (95.9-98.3%, 95% CI) treatment success in children with uncomplicated malaria recorded in the pooled analysis of data from seven studies supported by Novartis [41]. This suggests that the results of treatment success with AL in uncomplicated malaria patients in Cameroon are evenly distributed with other high malaria endemic countries. The aggravated e cacy of DHAP was 98.0%. AL and DHAP were found to have higher pooled e cacies than ASAQ. This nding is in agreement with those of different studies on the e cacy of ACTs assessed using the network metaanalysis approach [11,12,42]. DHAP has also been shown to be highly e cacious in the clearance of malaria parasites among Human immunode ciency virus (HIV) patients in Malawi and Mozambique [43]. The standard uncomplicated malaria treatment guideline for Cameroon recommends a three-days administration of ASAQ or AL depending on the weight and age of the patient [44]. Based on the present evidence, it will be advisable to consider DHAP in the treatment of malaria especially among HIV patients where ASAQ and AL may be contraindicated concurrently taking efavirenz-or nevirapine-based antiretroviral therapy. It was observed that the e cacy of ASAQ declined an original value of 100% to a current value of 91.8% while that of AL declined from an original value of 100% to a current value of 96.7% over time. It was also shown that the e cacy of ASAQ declined faster in some ecological zones when compared to others. The greatest deep in ASAQ was noticed in the Littoral ecological zone. This change could be due to the decline in the e cacy of ASAQ in Buea [33]. However, it is important to note that the study in Buea was among the rst studies to evaluate the e cacy and safety of ACTs in real time.
Most studies included in the present review achieved a rapid reduction of fevers and parasitaemia between D0 and D3 of assessment. Majority of these studies treated patients with ASAQ, AL and DHAP. A previous aggregate study on the clinical predictors of early parasitological response to ACTs (ASAQ, AL and DHAP) in African patients with uncomplicated falciparum malaria con rmed the rapid decrease of parasite positivity rate (PPR) from 59.7% (95% CI: 54.5-64.9) on day 1 to 6.7% (95% CI: 4.8-8.7) on day 2 and 0.9% (95% CI: 0.5--1.2) on day 3 [45].
Some studies showed a delayed clearance on D1 with a proportion of 75% persistence of parasite in Mutengene/Garoua [30] and 35% parasitaemia on D28 Buea-Tole [33]. It should be recalled that Buea is in a zone which borders with Limbe and Mutengene which are sites for high chloroquine-resistance [2,46].
The ACTs, ASAQ, AL and DHAP are well tolerated in spite of a few adverse events such as asthenia, diarrhoea, abdominal pain, anorexia, nausea, vomiting, headache and dizziness reported during different studies. These ADRs were not serious enough to discontinue anti-malarial treatments except for individuals in studies that reported serious adverse events such as jaundice, haemoglobulinuria, anaemia, convulsion, severe fatigue, severe malaria and deaths. In the current review, 3 study participant died during treatment [26]. This rate of mortality was not related to the study drugs on patients included in the study with uncomplicated malaria. Similarly, systematic reviews and meta-analyses conducted in Ethiopia and Sudan reported similar adverse events when patients with uncomplicated falciparum malaria were administered ACTs [35,36,47].

Strengths And Limitations Of The Study
The current study has several strengths. A total of 39 studies with the same or different ACTs derived from 13 published articles and 3 unpublished studies were included that gave a total of 3,747 study participants. The study assessed e cacy of commonly used anti-malarial drugs: ASAQ and AL. Study outcomes were measured both clinically and parasitologically. Most studies evaluated the comparative e cacy of different anti-malarial medications.
However, the current study is not without limitations. The study included only mono-infection with P. falciparum with no available data on the other Plasmodium species. Moreover, not all studies reported AEs to anti-malarial drugs. Furthermore, there were fewer studies carried out in the Northern Regions of the country compared to the Southern Regions.

Conclusion
The present systematic review and meta-analyses reported a high overall e cacy of ACTs (97.9%). The standard regimens, ASAQ, AL and DHAP showed high cure rates of 97.5%, 99.4% and 98.0% respectively.
A number of adverse drug events were encountered such as asthenia, diarrhoea, abdominal pain, anorexia, nausea, vomiting, headache and dizziness after the administration of ACTs. However, the ADRs were not serious enough to discontinue the use of anti-malarial treatment except for a few patients who experienced severe adverse events. ACTs in Cameroon are still e cacious and well tolerated albeit with a slight decline in the e cacies of ASAQ and AL over time. There is need for continuous monitoring of e cacy of ACTs despite the high success cure rates as resistance seems inevitable since cases of antimalarial drug resistance have been reported in some areas of the world. Ethics approval and consent to participate Not applicable because this is a systematic review and meta-analysis.

Consent for publication
Not applicable.

Availability of data and materials
All data and materials used for the analysis of this systematic review and meta-analysis are included in this write-up and the additional documents.

Competing interests
The authors declare that they have no competing interests.