Epidemiology of malaria and chloroquine resistance in Mizoram, northeastern India, a malaria-endemic region bordering Myanmar

Background Mizoram, a northeastern state in India, shares international borders with Myanmar and Bangladesh and is considered to be one of the key routes through which drug-resistant parasites of Southeast Asia enter mainland India. Despite its strategic location and importance, malaria epidemiology and molecular status of chloroquine resistance had not been well documented, and since chloroquine (CQ), as the first-line treatment in Plasmodium falciparum infection was discontinued since 2008, it was expected that CQ-sensitive haplotype would be more abundant. Methods Malaria epidemiology data for the period 2010 to 2018 was collected from the office of State Vector Disease Control Programme. Plasmodium falciparum-positive blood samples were collected from government district hospitals, community health centres, primary health centres, sub-centres, and diagnostic centres from six malaria-prone districts. The samples were processed and analysed using genes–P. falciparum chloroquine-resistant transporter (pfcrt) and P. falciparum multidrug resistance 1 (pfmdr1) via sequencing of PCR amplicon from 2015 to 2017. Results Malaria occurred throughout the year and P. falciparum accounted for > 89% of total malaria cases. During 2010–2018, the highest number of malaria incidence was recorded in Lawngtlai (36% of total malaria cases; average API2010–2018 of 34.8) while Champhai remained consistently low (0.4%; average API2010–2018 of 0.04). Males of ≥ 15 years old contributed maximum (35.7%) among gender and age malarial distribution recorded during 2014–2018. Death due to malaria gradually decreased over the years. A higher abundance of mutated pfcrt (58.5% of the total sample analysed) and a lower prevalence of mutated pfmdr1 (48.7%) were observed. All mutations identified for pfcrt belong to the Southeast Asian CVIET haplotype. Only a single point mutation was observed at 86 (N → Y) position in pfmdr1 (48.7%). The key N86Y mutation in pfmdr1 that had been shown to modulate CQR was found in 67.1% of the samples positive for the CVIET haplotype. Conclusions This is the first report that details malaria epidemiology and also the molecular status of CQ-resistance in P. falciparum population of the region. The efforts of the State Vector Borne Disease Control Programme have proved to be quite effective in controlling the malaria burden in the state. Despite the discontinuation of CQ for a decade, local P. falciparum is observed with decreased CQ-sensitive haplotype. It is believed that the present findings will form a basis for further studies on genetic diversity in P. falciparum, which could confer better understanding of the complexity of the disease in Southeast Asia.


Background
Malaria transmission in India occurs predominantly either as Plasmodium falciparum or Plasmodium vivax infections. India and 15 sub-Saharan African countries are estimated to carry 80% of malaria burden across the world in 2018 [1]. In India, 0.84 million malaria cases were reported in 2017, of which North Eastern (NE), Eastern and Central Indian regions contributed 80% of the total cases [2]. Mizoram, along with Assam, Arunachal Pradesh, Tripura, Manipur, Mizoram, Sikkim and Meghalaya form the eight NE States of India. NE India accounts for 4% of the population but contributed 6.6% of malaria cases and 25% of malaria mortality in India in 2018 [3].
Mizoram is a landlocked state in NE India with an estimated population of 1 million in the 2018 Census. Mizoram extends between latitude 21°58′N to 24°35′N and longitude 92°15′E to 93°29′E, and altitude ranges from 500 to 2157 m. Mizoram is under the direct influence of southwest monsoon characterized by wet summers (18-33 °C) and dry winters (11-24 °C) [4]. The state lies within the Indo-Burma region, one of the four biodiversity hotspots of India. The eastern and western regions of Mizoram share international borders with Myanmar and Bangladesh, respectively and shares domestic borders with Manipur, Assam and Tripura. The porous border with Myanmar could be one of the major entry routes of drug-resistant parasites to mainland India.
In 1973, the first cases of P. falciparum chloroquine resistance (CQR) were reported from Karbi Anglong district in Assam [5]. In the early 1980s, CQR was reported in Mizoram [6]. Most likely, CQR has spread from Southeast Asia (SEA) to the Indian mainland through NE India [7]. Treatment failures to sulfadoxine/pyrimethamine (SP) were first reported in NE India in 2005 [8] and molecular evidence to its resistance in P. falciparum was reported in 2013 [9,10]. This ushered in a change in the national drug policy of India; artesunate plus sulfadoxine-pyrimethamine (SP) has been replaced by a coformulated tablet of artemether-lumefantrine (AL) in NE India since 2013 [11].
Within the 72 to 76 amino acid loci of P. falciparum chloroquine resistant transporter gene (pfcrt), the chloroquine-sensitive (CQS) strains have been marked with CVMNK allele [12], while polymorphism within this locus conferring CQR [13] is characterized by CVIET and CVIDT in parts of SEA and Indochina, respectively [14], SVMNT in Africa [15], and CVMNT in South America [16,17]. In addition, 32-point mutations in pfcrt have been identified to date and of these 11 are associated with CQR. Worldwide, K76T is considered to be the major molecular marker and a key determinant of CQR [18]. CQR, mediated by pfcrt mutations, has been shown to be modulated by mutations in P. falciparum multidrug resistance gene-1 (pfmdr1). Among the five (N86Y, Y184F, S1034C, N1042D, D1246Y) mdr1 mutations, N86Y and Y184F have been linked with CQR [19][20][21].
Because the molecular status of CQR in P. falciparum in Mizoram remained poorly documented, the present study aimed at screening for polymorphisms within the genes linked to CQR (pfcrt and pfmdr1) across Mizoram. Additionally, the study aimed to present malaria epidemiology from 2010 to 2018. Due to the discontinuation of CQ as the first-line P. falciparum treatment in 2008, it was expected that the prevalence of CQS strain would be more abundant than CQR haplotype.

Epidemiology data
Mizoram is divided into 8 districts: Aizawl, Lunglei, Kolasib, Mamit, Lawngtlai, Champhai, Saiha, and Serchhip (Fig. 1). Available monthly and annual records of malaria (P. falciparum and P. vivax), including the number of deaths due to malaria (P. falciparum and P. vivax combined) and anti-malarial drug treatment for P. falciparum were collected from the office of the State Vector Borne Disease Control Programme (SVBDCP) for the 8 districts of Mizoram from 2010 to 2018 (Fig. 1). Malaria case distribution was further categorized into gender (proportion of males and female), age (0-4 years, 5-14 years and 15 + years; as recorded and available in SVBDCP records), passive case detection (PCD) i.e., febrile patients suspecting malaria seeking medical confirmation), and pregnancy infections.

Sample collection
From 2015 to 2017, confirmed P. falciparum-positive blood samples were collected from government district hospitals, community health centres, primary health centres, sub-centres, and diagnostic centres of Aizawl, Lunglei, Saiha, Kolasib, Mamit, Lawngtlai, and Serchhip districts (Champhai district was excluded for sample collection due to its consistently low annual parasite incidence (API) (Fig. 1). Samples were collected after obtaining consent from patients and guardians. For patients under 10 years of age, consent was obtained from their guardians (i.e., either one or both parents present at the time of blood sampling). Details regarding new or Keywords: Mizoram, Plasmodium falciparum, Chloroquine resistance, pfcrt, pfmdr1 relapse infection, age and gender of the patient, date of blood examination and collection of samples; onset and duration of anti-malarial drug treatment administered were recorded for each blood sample collected. Blood samples were spotted on filter paper or were collected in EDTA vials, and preserved at − 20 °C for further analysis.

DNA extraction and species-specific identification of malaria parasite
QIAamp Mini Kit (QIAGEN) was used to extract genomic DNA from positive P. falciparum samples according to manufacturer protocol. The extracted DNA was stored at − 20 °C prior to use. Confirmatory PCR to detect P. falciparum and P. vivax was performed by targeting the 18SrRNA region as previously described [22].

PCR amplification and sequencing
The details used for pfcrt and pfmdr are given in Additional file 1: Table S1. For pfcrt, three nested PCRs were performed using gene-specific primers to detect polymorphism at exons 1-2, 2-8 and 9-12. The primary PCR was carried out in 25 µl volume using EmeraldAmp GT PCR mix (Takara Bio Inc., Japan), 0.

Nucleotide and amino acid data
The pfcrt and pfmdr1 nucleotide and amino acid sequences were aligned using MAFFT [23] and MEGA 7 [24]. The pfcrt sequences were aligned with 3D7 CQS strain (

Statistical analysis
Epidemiological and molecular data were entered, analysed in Excel (Microsoft) and graphical presentations were also performed in Excel. Association between pfcrt and pfmdr1 mutation was analysed using Chi square test in Statistical Package for Social Science (IBM SPSS Statistics for Windows, Version 24.0. IBM Corp., Armonk, NY, USA).

Malaria epidemiology analysis
Malaria epidemiology recorded during 2010-2018 showed that malaria was persistent throughout the year, and across all districts in the state. The surveillance coverage accounted for ≥ 95% in all sites of study ( Fig. 2). Majority of malarial infection and malaria-related death was due to P. falciparum infection and accounted for 89.3% of total malaria cases ( Fig. 2 and Additional file 3: Table S3). The observed highest malaria incidence and malaria-related deaths was in 2014. Reason for the high mortality rate in 2014 was because the majority of infection had evolved to hyper-reactive malarial splenomegaly. The factors were: (i) lack of timely detection of the parasite due to lack of transportation (mainly roadblock due to landslides); and, (ii) less availability of P. falciparum treatment (ACT-AL) in almost all rural medical facilities; patients were devoid of immediate medical assistance. However, a decline in the number of malariarelated deaths was observed in subsequent years (Fig. 2). The total number of death due to malaria was 172; Mamit and Lawngtlai recorded deaths annually throughout the study period (Additional file 3: Table S3).
Overall, most malaria incidence was contributed by Lawngtlai (average API 2010-2018 = 34.8; total number of malaria cases = 41,516; number of P. falciparum = 37,892; number of P. vivax = 3664); followed by Mamit (average API 2010-2018 = 34.4; total malaria cases = 23,391; P. falciparum = 21,497; P. vivax = 1894). The total sum of malaria cases (total malaria, total P. falciparum, P. vivax infections and death) throughout 9 years of study was calculated. This was an attempt to observe the trend of infection and to find any discrepancy in its abundance at monthly resolution (Fig. 3). A consistent and cyclic pattern of gradual increase in malaria cases during the rainy summer/monsoon season (March to September) was observed. The cases peaked in June to July and then decreased during the dry winter season (October to January) (Fig. 3). A similar pattern was also observed for the number of malaria-related deaths in all study areas, excluding Mamit and Lawngtlai. Maximum deaths were reported from Mamit (n = 13) during August.
Distribution of age and gender-specific malarial infection, passive case detection and pregnancy infections recorded for 2014 to 2018 were analysed (Table 1). Overall, 15 + years of age had the highest malaria incidence (39, Table S3). The total number of pregnancy infections recorded from 2014 to 2018 was 364 (0.53% total malaria cases), and deaths during pregnancy were not reported. A total of 61,501 individuals were recorded for passive case detection (PCD).
The result of Chi square analysis in 265 samples tested for both pfcrt and pfmdr1 polymorphisms showed a significant association existed between the mutations (χ 2 = 121.48, df = 1, p = 0.001, φ c = 0.68). The key 86Y mutation in pfmdr1 that had been shown to modulate CQR was found in 67.1% of R-population of pfcrt gene (n = 155) analysed, meanwhile its mutation was found to be nil in WT-populations of pfcrt (n = 110) (Fig. 4).

Discussion
Historically the spread of CQR has been linked from SEA to Africa via the Indian subcontinent [25,26], and NE states could have contributed to CQR spread in mainland India through host movements [7]. The eastern region of Mizoram shares its international border with Myanmar. By land, one can enter Tedim in the Chin state of Myanmar through the Zokhawthar-Rih border of Champhai district. Both regions share a historical similarity in ethnic, religion, culture, and language [27]. Trade relations between the locals in the border areas have led to continuous movement of people in both directions. Despite the strategic location of Mizoram, and its importance as a conduit for the spread of drug-resistant P. falciparum, malaria epidemiology and drug resistance status have been poorly documented. This is the first study to give a detailed account of malaria epidemiology in Mizoram in this decade. Plasmodium falciparum contributes to the majority of the    [28,29]. Additionally, recent studies on vector distribution identified Anopheles campestris, Anopheles jeyporiensis, Anopheles maculatus, and Anopheles nivipes as the most abundant species in malaria endemic parts of Mizoram [30,31]. It has also been documented that the rise in Anopheles populations is subsequently followed by a peak in malaria cases [31]. In all districts of Mizoram, malaria cases peak during the major monsoon season: May to July (Fig. 3). Malaria transmission is perennial and cases are reported throughout the year; the burden is highest in the districts of Lunglei, Lawngtlai and Mamit. Champhai, which links with Myanmar through an international highway, consistently reported the least API score (Additional file 3: Table S3). The low parasite prevalence could be attributed to its high elevation (1678 m above sea level), in parallel with low abundance of anopheline species in the region. Even though Mizoram is a high malaria-endemic region, a gradual decrease in the number of P. falciparum cases and mortality due to malaria has been observed from 2010 onwards. One important factor could be attributed to the current P. falciparum treatment (ACT-AL), and to the strategies implemented by the SVDCP that includes: visual and verbal awareness campaigns about the disease, and about potential breeding habitats, outdoor residual spraying of DDT twice a year, and quinquennial distribution of > 500,000 deltamethrin-treated bed nets. This is evidenced by the substantial number of PCD seeking medical care (Table 1). Furthermore, efficient and easy access to healthcare facilities could contribute to effective management and timely treatment of pregnancy infections as not a single pregnancy death due to malaria has been reported during the study. This is the first report conducted to determine the molecular status of CQR in Mizoram; the study targeted the established molecular markers of pfcrt and pfmdr1. Artesunate + SP is currently the treatment for uncomplicated P. falciparum in all parts of India, except NE India. It has been a decade since the use of CQ for P. falciparum cases; the change from CQ to ACT-AL as first-line treatment was expected to result in wide spread distribution of CQS allele within the P. falciparum population, as documented in many African countries [32][33][34][35][36]. Unexpectedly, only 41.5% (110/265) of the samples were identified as CQS type in the study sites (Fig. 4). Perhaps this is because CQ (the current the treatment for P. vivax infection) could be employed for treating unrecognized mixed infections, thereby making P. falciparum vulnerable to the drug [37]. In addition, the present study on pfcrt polymorphisms indicated that CQS (41.5%) and CQR (58.5%) co-existed, albeit when CQ treatment ceased, in a manner similar to other parts of Asia [38,39] (Fig. 4). In India, the major pfcrt haplotype identified is the SVMNT; yet, varying frequencies of CVIET, CVIDT and CVMNT have also been reported [40][41][42][43][44]. Interestingly, in Mizoram was observed with 58.5% (155/265) of the samples were positive for CVIET allele, and is the only haplotype available in the region; CVIET is the predominant CQR-allele found mainly in SEA. The P. falciparum Fig. 4 Distribution of chloroquine resistance transporter (pfcrt) and multidrug-resistant (pfmdr1) genotypes across Mizoram. WT wild type, R CQR genotype, R1 CQR type-1, R2 CQR type-2; 86Y: N → Y substitution at 86 amino acid of pfmdr1; All calculations are in percentages population, having the CVIET allele was further divided into two groups: R1 (CVIET-AQ-S-R) and R2 (CVIET-SE-S-I); R2 contains amino acid substitutions similar to Old World-resistant phenotype [45,46] (Table 2). Subsequent in vitro studies on R1 and R2 of local P. falciparum populations may give rise to further prospective study on CQ-sensitivity in the future.
Globally, five-point mutations (N86Y, Y184F, S1034C, N1042D and D1046) have been described for pfmdr1 and, importantly, 86Y polymorphism is linked with decreased sensitivity to CQ [47] and amodiaquine (chemically similar to CQ) [48,49]. It has been recorded that the N terminal, N86Y and Y184F mutations are more prevalent in African and Asian parasites, while the C terminal mutations: S1034C, N1042D and D1046Y are more common in South American parasites [15]. This study identified a single 86Y substitution at the N terminal region with 48.9% (174/356) while the remaining 51.1% (182/356) displayed uniformity throughout the amino acid sequence. Substitutions at 184 and the later three codon positions discussed above were not observed. Analysis indicates that 67.1% of samples with CVIET haplotype also carried N86Y mutation. A similar pattern had been observed in parts of India [37], Asia [38,[50][51][52] and Africa [35,53,54].

Conclusion
Malaria is perennial throughout Mizoram and P. falciparum transmission accounts for more malaria cases than P. vivax. However, the efforts of SVDCP have proved to be quite effective in controlling the malaria burden in the state, reducing malaria transmission and malaria deaths over the years. The only available CQR allele found across the state is the South Asian CVIET haplotype. Despite the discontinuance of CQ for a decade, polymorphisms at pfcrt (76T) and pfmdr1 (86Y) continued to exist at an unexpected rate. In addition, a single polymorphism at site 86 of pfmdr1 proved to be a unique characteristic feature of local resistant P. falciparum population. It is speculated that less abundance of mutated pfmdr1 86 (N → Y) substitution and absence of polymorphism at codon 184, 1034, 1042 and 1046 loci may still contribute to high efficacy in artemisinin-lumefantrine treatment currently employed within the region. Further studies on genetic diversity of local P. falciparum in greater detail would confer better understanding about the complexity of the disease.