Chloroquine Resistant Plasmodium falciparum in Nigeria: Relationship between pfcrt and pfmdr1 Polymorphisms, In-Vitro Resistance and Treatment Outcome

Abstract

This study was designed to evaluate the association between polymorphisms in pfcrt and pfmdr1 genes and in-vitro chloroquine (CQ) sensitivity in fresh isolates of P. falciparum and patients’ treatment outcome. The modified schizont inhibition assay was used to determine in-vitro sensitivity of P. falciparum. Polymorphisms in pfcrt and pfmdr1 genes were detected using nested PCR and RFLP techniques in 84 P. falciparum isolates obtained from patients with acute uncomplicated malaria.

Eighty five percent (71/84) and 15% (13/84) of the parasites were resistant and sensitive in-vitro to CQ respectively. Molecular analysis showed presence of mutant pfcrtT76, pfmdr1Y86 and pfmdr1F184 alleles in 60%, 33% and 14% of the isolates respectively. There was a significant association between in-vitro and in-vivo CQ resistance (p=0.029) and also between the presence of mutant pfcrtT76+pfmdr1 Y86-Y184 haplotype and in-vitro (p=0.013) or in-vivo CQ resistance (p=0.024).

Overall results from this study demonstrates that the presence of pfcrtT76+ pfmdr1 Y86-Y184 haplotype in Nigerian isolates of Plasmodium falciparum is predictive of in-vitro and in-vivo CQ resistance and therefore may be useful for monitoring resistance to this drug.

INTRODUCTION

The emergence and spread of drug resistant malaria parasites in endemic regions has posed a great threat to usefulness of chloroquine (CQ) and sulphadoxine-pyrimethamine (SP), the cheapest and widely used antimalarial drugs. This widespread prevalence of antimalarial drug resistant parasites has led to strong calls for the introduction of artemisinin based combination therapies (ACTs) for treatment of malaria in most endemic areas, as it is expected to represent an effective approach in curbing the development of resistance by Plasmodium falciparum to currently available compounds [1, 2]. Presently, most malaria endemic countries in Africa including Nigeria have changed their first line antimalarial treatment from CQ or SP to amodiaquine combined with artesunate or the combination of artemether and lumefantrine. ACTs used in most malaria endemic countries have demonstrated high efficacy, protection against the development of resistance to each component and reduction of malaria transmission [3–5]. However, the relatively high of costs, dosing complexity and the limited experience of their use in sub-Saharan Africa may hamper the widespread deployment of these drug combinations [6]. In addition there are reports of the emergence of P. falciparum with reduced susceptibility to some of the ACTs [7].

This situation therefore points to the fact that other more affordable, effective and readily available combinations are still required. An earlier report [8] has demonstrated the return of CQ susceptible parasites in Malawi soon after the country switched from CQ to SP as the first line therapy for malaria. Furthermore, a recent clinical trial in Malawi also confirmed that CQ had excellent clinical efficacy, 12 years after it was removed from use [9, 10]. It is therefore necessary to understand the molecular basis of resistance to already available antimalarial drugs such as CQ, and explore the potentials of the information generated in improving the potency and rational for developing effective drug combinations.

The genetic and biochemical basis of CQ resistance in P. falciparum has been the subject of several research and has not been fully elucidated in field epidemiologic studies. P. falciparum resistance to CQ has been associated with lower drug accumulation in infected erythrocytes [11] or mainly to a mutation at position 76 (K76T) in the Plasmodium falciparum chloroquine resistance transporter (pfcrt) gene [12, 13]. Several other studies have suggested the implication of multidrug resistance-1 (mdr1) gene family in resistance to quinoline containing compounds in P. falciparum [14–16]. Five point mutations on codons 86, 184, 1034,1042 and 1246 have been identified to influence the response of P. falciparum to CQ [14, 17]. Several reports have also shown that there is an association between polymorphisms in these two genes and resistance of P. falciparum to CQ in-vitro [12–14, 16, 18], or with CQ therapy failures in field studies [13, 18–26]. While most laboratory reports points to the involvement of these two genes in mediating resistance to CQ and perhaps other drugs, it is yet to demonstrate under field conditions how the allelic differences in pfcrt and pfmdr1 genes correlate with the resistant phenotype in different malaria endemic areas of the world, or whether there are differences in resistance mechanisms by parasites in various regions of the world where malaria is endemic.

In this study, the involvement of polymorphisms in pfcrt and pfmdr1 genes in in-vitro susceptibility of fresh isolates of P. falciparum as well as CQ treatment outcome in children with acute uncomplicated P. falciparum malaria in Southwest Nigeria was investigated. Point mutations in pfcrt and pfmdr1 genes were found to be associated with both in-vitro and in-vivo CQ resistance. The potential implications of these findings in monitoring CQ resistance in areas where ACTs have been introduced for treatment of P. falciparum malaria are discussed.

STUDY DESIGN

Study Site

The study was carried out at the Malaria Research Laboratory, College of Medicine, University of Ibadan, Nigeria. Malaria in Ibadan is hyperendemic, transmission occurs year round but is more intense from April to October, during the rainy season.

Patients Selection, Treatment and Follow-Up

Children aged 6months to 12 years with acute symptoms of P. falciparum malaria infections were enrolled after clinical examination and microscopic confirmation of infections in a large clinical efficacy study. Informed consent for participation in the study was obtained from parents/guardians of children under the age of 10 years, while assent was obtained from each patient between the ages of 10–12 years. The Joint UI/UCH Institutional Review Committee (IRC) approved the study protocol. Each child was treated with standard dose of CQ (25mg/kg body weight over three days) and followed up for a period of 28 days according to World Health Organization (WHO) protocol [27]. Infection in each child was considered cured if no parasites appeared in the peripheral blood samples during the 28 days follow-up period after treatment. Children in whom parasites reappeared in their blood within the 28 days of follow up were classified as treatment failures and were re-treated with SP to cure the infection.

Determination of In-Vitro Susceptibility of Patient Isolates to CQ

In-vitro susceptibility of each patient isolate of P. falciparum to CQ was determined using a modification of the schizont inhibition assay [28]. Briefly, a template containing three-fold serial dilutions of the working CQ solution (9690nM) was prepared in a 96-well microtiter plate. Wells in row H served as controls without drug. Test plates were derived from each template by transferring 25 µl of the drug dilutions to each plate. Two hundred microliters (200 µl) of 1ml parasitized blood diluted in 19ml of culture medium (RPMI 1640+ HEPES and sodium bicarbonate) was transferred into each well of the plate. Plates containing parasites suspension with CQ in each well were incubated at 37°C for 24 to 36 hours in a plexiglass chambers containing a gas mixture (5% O₂, 5% CO₂, 90% N₂). The final concentration of CQ in test plate ranged between 969nM and 1.3nM.

The assay was terminated when at least 60% of parasites in the control wells (Row H) were schizonts. Each well in a column of 96 well plate was harvested onto glass slides as thick smears, air dried and stained with Giemsa. Parasites development to schizont was determined by counting the number of schizonts against 200 white blood cells in each smear using x100 oil immersion objective of a light microscope. Concentration-response data were analyzed by a nonlinear regression analysis. The 50% inhibitory concentrations (IC-50) for CQ were calculated using GraphPad Prism version 4.0 for windows software (GraphPad software, San Diego, LA, USA, www.graphpad.com).

According to the WHO criteria [29] of in-vitro CQ sensitivity of P. falciparum, a complete inhibition of parasite growth in wells with CQ concentration <107nM (55.5ng/ml) was considered as sensitive, parasite growth in wells with CQ concentration >107nM is defined as resistant and parasite growth in wells of 107nM CQ concentration but not in wells of higher concentration is defined as borderline sensitivity.

Parasite DNA Extraction

Parasite genomic DNA was extracted from blood samples collected on filter paper using the chelex extraction method according to the method of Plowe et al. [30].

Analysis of P. falciparum Point Mutations in pfcrt and pfmdr1 Genes in Isolates Obtained from Patients

The lysine (K) to threonine (T) mutation at codon 76 of the pfcrt gene was detected by nested PCR followed by RFLP as previously described by Happi et al. [24]. DNA from two laboratory adapted P. falciparum clones, 3D7 (Wild type) and Dd2 (mutant) were used as negative and positve controls respectively.

The nested PCR and RFLP methods were also used to evaluate the pfmdr1 N86Y, Y184F S1034C and N1042D point mutations as described previously [26, 31]. The K1 and 7G8 laboratory strains of P. falciparum presenting with different genotypes at the different codons analyzed were used as controls.

All amplifications were performed in a final volume of 25ul in a PCT- 200 Peltier thermal cycler (MJR Research Inc, MA, USA)

Determination of Plasmodium falciparum Clonal Profile in Infections

Previous studies [24–26, 32] on molecular analysis of Plasmodium falciparum malaria in Nigerian children following treatment with CQ, have demonstrated that the merozoite surface protein-2 (msp-2) was the most informative genetic marker to evaluate parasites diversity and the complexity of P. falciparum infections in Ibadan, Nigeria. In this study, isolates from each P. falciparum infection were characterized on the basis of fragment sizes of alleles of msp-2 after amplification by PCR. Infections were defined as polyclonal if samples from patients showed more than one allele of FC27 or IC1/3D7 families of msp-2. If an isolate had one allele at each of the families, the clone number was taken to be one. The complexity of infection was calculated as the mean number of distinct fragments of FC27 and IC1/3D7 per PCR-positive sample. Polymorphism in msp-2 was also used to distinguish between recrudescence i.e. resistant infection and re-infection. A CQ resistant infection was defined as the occurrence of the same or a subset of the alleles at each of the families (FC27 or IC1/3D7) of msp-2 in the pre and post-treatment samples. A lack of allelic identity in the two families of msp-2 in matched pre and post-treatment samples indicated a newly acquired infection.

Data Analysis

Clinical response to treatment with CQ was expressed as cured (cleared) or resistant (failed). In-vitro profile of P. falciparum to CQ was defined as sensitive (S) or resistant (R). Mean values were determined as mean ± standard deviation (SD). The student t-test was used to compare mean values. The Fisher’s exact test was used to assess a statistical association between in-vitro susceptibility profiles of patients’ isolates of P. falciparum to CQ and point mutations in parasites pfcrt, pfmdr1 genes and treatment outcome. P value < 0.05 was considered significant.

RESULTS

One hundred and twenty children with P. falciparum infection were recruited into the study. The mean age of the children was 5.41±2.94years (Range 6months-13years). Plasmodium falciparum blood sample was obtained from 84 children by venipuncture for evaluation of the association between in-vitro CQ susceptibility, markers of CQ resistance, and patients treatment outcome. The geometric mean of parasite density in the children at enrollment was 37,451 parasites/µl of blood (range: 7775 to 150,000 parasites/ µl). There was no significant difference between the age (P=0.82), and geometric mean parasites density (P=0.60) of the 84 children whose blood sample were obtained for in-vitro CQ susceptibility and 120 children that were enrolled in the main study.

Response of Infection to Treatment with Standard Regimen of CQ

Of the 84 children whose blood samples were obtained for the evaluation of the association between in-vitro susceptibility profile of parasites to CQ, molecular markers of CQ resistance and patients parasitological response to treatment, 62% (52 of 84) and 38% were classified as cured and treatment failures respectively after PCR correction by msp-2 (Table 1). Mean parasite clearance time (PCT) in the children cured with CQ was 3.03 ± 0.41 days, while the mean fever clearance time (FCT) was 2.12 ± 0.5 days after initiation of treatment. Parasites in the group of patients that failed treatment with CQ initially cleared but reappeared between 7 and 28 days (mean recrudescent time=15.56±6.03 days) after commencement of treatment. Infections in 26 out of 32 patients (81%) who failed treatment with CQ were classified as RI. Parasites in these patients initially cleared but reappeared between 14 and 28 days after commencement of treatment. RII level of resistance was observed only in six (19%) of the 32 patients who failed CQ therapy. Parasites in these patients cleared as well but reappeared by day 7 after initiation of treatment.

Table 1.

Demographic Characteristics of Patients at Enrollment and Patient Treatment Outcome to Chloroquine

	Treatment with CQ
No of patients/isolates	84
Sex
Male (%)	43 (49%)
Female (%)	41 (51%)
Age (years)
Mean±SD	5.41±2.94
Range	6mths-13years
Parasite count (ul−1)
Geometric mean	37,451
Range	7775–150,000
In-vivo Treatment outcome
Cured (%)	52 (62%)
Resistant (%)	32 (38%)

SD= Standard deviation.

Parasites Population Structure and Complexity of Infections in Pre- and Post-Treatment Isolates

Matched sample pairs collected before and after treatment from all 32 patients who failed CQ treatment were successfully analyzed at the msp-2 locus. Alleles were classified according to the size of PCR fragments.

Genotyping of these samples confirmed our previous report [30] of the presence of different allelic families of msp-2 in parasite DNA derived from a single patient, indicating a polyclonal infection. All pre-treatment isolates were positive for the msp-2 IC1 and/or FC27 alleles and produced up to eight different fragment sizes (IC1/3D7: 390–1090 bp, FC27: 140–1100 bp). The estimated average number of genetically distinct parasite population as determined with msp-2 in pre- and post-treatment isolates from these 32 patients was 5.3 and 3.8 respectively. There was a significant reduction (p=0.03) in the number of msp-2 alleles in post-treatment isolates compared to pre-treatment isolates when infections recrudesced.

Detailed analysis of paired pre- and post-treatment isolates from these patients who failed treatment with CQ showed two categories of infection. The first group of infections consisting of 72% (23 of 32) of patient isolates that failed treatment with CQ, had identical paired PCR fragments at both FC27 and IC1 families of msp-2, indicating genuine recrudescent infections after treatment with CQ. The second group of infections in the remaining 9 patients showed parasites similar to pre-treatment isolates and the presence of new parasite populations with different genotypes. However, infections in these patients were also considered as treatment failures.

Determination of In-Vitro CQ Susceptibility of Patient Isolates

Based on the WHO criteria [29] of in-vitro susceptibility of P. falciparum to CQ in 84 isolates with successful tests, 15% (13/84) had minimum inhibitory concentration (MIC) of CQ below 107nM therefore classified sensitive isolates. MIC of CQ >107nM was observed in 54% (45/84) of the isolates and classified as resistant isolates. Borderline isolates defined as MIC of 107nM were observed in 31% (26/84) of isolates. The mean CQ fifty percent inhibitory concentration (IC-50) for sensitive, borderline and resistant isolates was 6.28nM, 13.70nM and 64.25nM respectively.

Pfcrt and pfmdr1 Polymorphisms Among Patient Isolates

Seventy-eight (78), 81 and 70 of the 84 samples were successfully amplified by PCR at locus 76 of pfcrt, loci 86 and 184 of pfmdr1 respectively, and were considered for analysis (Table 2). Sixty percent (60%) of the patients isolates successfully analyzed harbored the mutant pfcrtT76 allele that has been associated with CQ resistance. Ten percent (10%) showed a mixed genotype, while 30% harbored the wild-type pfcrtK76 allele. Analysis of post-treatment samples obtained from the patients who failed CQ treatment showed that isolates obtained from 58% carried the mutant pfcrtT76 allele. None of the post treatment isolates harbored the wild-type pfcrtK76 allele. Comparison of pre- and post-treatment samples obtained from patients who failed treatment with CQ, showed no increase in the prevalence of mutant pfcrtT76 allele while mixed allele increased from 13% in pretreatment isolates to 42% in the post treatment isolates (Table 2).

Table 2.

Frequency and Prevalence of pfcrt and pfmdr1 Alleles (At Enrollment) in Peripheral Blood Obtained from Children with Acute Uncomplicated Malaria

Genes and Alleles	Frequency	Prevalence (%)
K76T Pfcrt (n=78)
K76a	23	30
T76b	47	60
K76+T76	8	10
N86Y pfmdr1 (n=81)
N86a	26	32
Y86b	27	33
N86+Y86	28	35
Y184F pfmdr1 (n=70)
Y184a	57	82
F184b	10	14
Y184+F184	3	4

Pfcrt: P. falciparum CQ resistance transporter and pfmdr1: P. falciparum multiple drug resistance 1 gene

^awild type/CQ Sensitive allele

^bMutant/CQ Resistant allele.

RFLP analysis of pfmdr1 PCR products using AflIII, Dra I, Dde I and Ase I restriction enzymes were similarly performed [33] to detect pfmdr1 N86Y, Y184F, S1034C and N1042D mutations respectively in isolates obtained from children with P. falciparum infection. Mutant pfmdr1Y86 and pfmdr1F184 alleles were present in 33% (27 of 81) and 14% (10 of 70) in the pre treatment isolates respectively. All isolates analyzed for S1034C and N1042D mutations harbored the wild type S1034 and N1042 alleles respectively, while 32% and 82% of the isolates harbored wild type pfmdr1N86 and pfmdr1Y184 alleles respectively (Table 2). Mixed N86Y and Y184F allele were observed in 35% and 4% of the pre treatment isolates respectively. Analysis of post treatment isolates obtained from children who failed CQ treatment showed mutant pfmdr1Y86 and pfmdr1F184 alleles in 41% and 57% of the isolates respectively. The wild type pfmdr1N86 and pfmdr1Y184 alleles were present in 28% and 43% of the post-treatment isolates respectively. None of the post-treatment isolates harbored the mixed allele at codon 184 of pfmdr1 gene. However 31% harbored mixed allele on codon 86 (Table 2).

Polymorphisms in pfmdr1 gene were clustered into 4 specific haplotypes (haplotypes I- IV). Haplotype I consisted of isolates with both mutant pfmdr1Y86 and pfmdr1F184 alleles (Y86-F184-S1034-N1042). Haplotype II consisted of isolates with wild type pfmdr1N86 and pfmdr1Y184 (N86-Y184-S1034-N1042), while haplotypes III and IV consisted of isolates with mutant pfmdr1F184 (N86-F184-S1034-N1042) and pfmdr1Y86 (Y86-Y184-S1034-N1042) respectively (Table 3). In pre-treatment isolates, the pfmdr1 haplotype I and II were observed in 7% and 17% of the isolates respectively. Haplotypes III and IV were observed in 12% and 64% of infections respectively (Table 3). Comparative analysis of pre- and post-treatment isolates showed a significant and strong selection (p=0.007; χ²=7.13) of the pfmdr1 haplotype I in post treatment samples obtained from patients who failed CQ treatment (Table 3). Haplotypes II, III and IV were found in 4%, 19% and 46% of the post treatment isolates respectively (Table 3).

Table 3.

Mutations Patterns and Prevalence in pfmdr1 Gene in Isolates of P. falciparum Obtained from Children with Uncomplicated Malaria at Enrollment and after Treatment with CQ

PFMDR1CODONS
Haplotypes	86	184	1034	1042	Pre-Treatment (%) [n=70]	Post-Treatment (%) [n=26]	P Values (χ2)
I	Tyr (Y)	Phe (F)	Ser (S)	Asn (N)	5 (7%)	8 (31%)	0.007* (7.13)
II	Asn	Tyr	Ser	Asn	12 (17%)	1 (4%)	0.17
III	Asn	Phe	Ser	Asn	8 (12%)	5 (19%)	0.5
IV	Tyr	Tyr	Ser	Asn	45 (64%)	12 (46%)	0.1

Mutant allele in boldface; Tyr (Y)= Tyrosine; Asn (N)= Asparagine; Phe (F)= Phenylalanine; Ser (S)= Serine

^*P value statistically significant.

Assessment of the combination of pfcrt alleles at codon 76 with pfmdr1 haplotypes in pre and post-treatment samples obtained from patients showed that the wild type pfcrtK76 allele+any pfmdr1 haplotypes was only present in 27% (22/82) of the pre treatment isolates. Mutant pfcrtT76+pfmdr1 haplotype I (Y86-F184-S1034-N1042) was present in 2% (2/82) and 31% (8/26) of pre and post-treatment isolates respectively. Interestingly, the mutant pfcrtT76 allele combined with pfmdr1 haplotype IV (Y86-Y184-S1034-N1042), was observed in 50% and 46% of all pre- and post-treatment isolates respectively. None of the post treatment isolates obtained from patients that failed CQ treatment harboured the wild type pfcrtK76 allele+any pfmdr1 haplotype (Table 4).

Table 4.

Prevalence of pfcrt and pfmdr1 Polymorphisms in Pre-Treatment and Post-Treatment Isolates Obtained from Children Infected with P. falciparum

Alleles of Genes and Haplotype	Pre Treatment (%) n=82	Post Treatment (%) n=26
Pfcrt K76+any pfmdr1 haplotype	22 (27)	(0)
PfcrtT76+pfmdr1 haplotype I	2 (2)	8 (31)
PfcrtT76+pfmdr1 haplotype II	12 (15)	1 (4)
PfcrtT76+pfmdr1 haplotype III	5 (6)	5 (19)
PfcrtT76+pfmdr1 haplotype IV	41 (50)	12 (46)

n= Number of isolates.

Association between In-Vitro Chloroquine Susceptibility and Gene Polymorphisms

The in-vitro sensitivity profile according to WHO criteria showed that 45 out of 84 isolates were resistant to CQ. Correlation analysis between CQ in-vitro susceptibility and gene polymorphisms showed that there was a significant association between in-vitro CQ resistance and mutant pfcrtT76 (p=0.013; OR=3.553; 95%CI=1.278–9.875), wild-type pfmdr-1Y184 (p=0.029; OR=0.203; 95%CI=0.05–0.818), pfmdr1-Y86-Y184-S1034-N1042 haplotype (p=0.004; OR=4.133; 95%CI=1.555–10.99), mutant pfcrtT76+pfmdr1Y86 (p=0.028; OR=2.706; 95% CI=1.103–6.635), mutant pfcrtT76+ pfmdr1-Y86-Y184-S1034-N1042 (p=0.046; OR=2.464; 95%CI=1.009–6.002). There was a strong association (p=0.011; OR=0.27; 95%CI=0.096–0.764) between the presence of the pfcrtK76 allele and CQ sensitivity despite the presence of any pfmdr1 haplotype (Table 5).

Table 5.

Association between pfcrt/pfmdr1 Mutations in P. falciparum Isolates Obtained from Children at Enrollment and CQ In-Vitro Susceptibility Profile

In-Vitro Susceptibility Profile
Alleles of Genes/Haplotypes	Sensitive	Resistant	OR (95% CI)	P Value
Pfcrt
K76	15	8	3.553 (1.278–9.875)	0.013*
T76	19	36
Pfmdr1
N86	16	10	2.590 (0.993–6.76)	0.05
Y86	21	34
Y184	23	34	0.203 (0.050–0.818)	0.029*
F184	10	3
Pfmdr1Y86-Y184-S1034-N1042 haplotype
Present	15	31	4.133 (1.555–10.99)	0.004*
Absent	20	10
Mutants pfcrtT76 + pfmdr1 Y86
Present	14	28	2.706 (1.103–6.635)	0.028*
Absent	23	17
PfcrtT76+pfmdr1Y86-Y184-S1034-N1042 haplotype
Present	14	27	2.464 (1.009–6.002)	0.046*
Absent	23	18
PfcrtK76+ Any pfmdr1 haplotype
Present	15	7	0.27 (0.096–0.764)	0.011*
Absent	22	38

^*P value statistically significant; Alleles of pfcrt and pfmdr1 genes are the same as described in Table 2.

Association between Gene Polymorphisms and Patient Treatment Outcome

Univariate analyses showed that there was a significant association between CQ treatment failure and the presence of pfmdr1Y86-Y184 haplotype (p=0.022; OR=3.36; 95%CI=1.157–9.759) or mutant pfcrtT76+pfmdr1Y86-Y184 haplotype (p=0.024; OR=2.864; 95%CI=1.138–7.209) (Table 6).

Table 6.

Association between pfcrt/pfmdr1 Mutations in P. falciparum Isolates Collected from Children at Enrollment and CQ Treatment Outcome

Alleles of Genes/Haplotypes	In-Vitro CQ Outcome		OR (95%CI)	P Value
	Cured	Failed
Pfcrt
K76	16	7	1.77 (0.628–4.983)	0.277
T76	31	24
Pfmdr1
N86	20	6	2.778 (0.967–7.981)	0.053
Y86	30	25
Y184	33	24	0.41 (0.102–1.661)	0.344
F184	10	3
Pfmdr1 Y86-Y184-S1034-N1042
Present	25	21	3.36 (1.157–9.759)	0.022*
Absent	24	6
pfcrtT76+pfmdr1 Y86
Present	22	20	2.121 (0.856–5.258	0.102
Absent	28	12
PfcrtT76+pfmdr1 Y86-Y184-S1034-N1042
Present	20	21	2.864 (1.138–7.209)	0.024*
Absent	30	12
Pfcrt K76+any pfmdr1 haplotype
Present	16	6	0.49 (0.169–1.427)	0.187
Absent	34	26

^*P value statistically significant.

Alleles of pfcrt and pfmdr1 genes are the same as described in Table 2.

Association between In-Vitro CQ Susceptibility and In-Vivo Treatment Outcome

The correlation between in-vitro susceptibility profiles of P. falciparum isolates obtained from patients and responses of infection to CQ showed that among the 32 patients who failed CQ treatment, 22 were resistant to CQ in-vitro, while 10 (31%) had parasites that were sensitive to the drug. P. falciparum isolates obtained from 23 (44%) of the 52 patients who were cured with CQ, showed a resistant profile to CQ in-vitro. A significant association (p=0.029; RR=2.774; 95%CI= 1.098–7.005) was observed between in-vivo treatment outcome and in-vitro sensitivity of P. falciparum isolates to CQ.

DISCUSSION

This study has demonstrated high level P.falciparum resistance to CQ in-vitro and in-vivo in Ibadan, Southwest Nigeria. In-vitro and in-vivo CQ resistances were 54% and 38% respectively. The current level of in-vivo resistance (38%) is apparently lower compared to 51% [24–26] reported previously from the same study site. This sharp reduction in the level of in vivo CQ resistance may be explained by the 2005 antimalarial treatment policy change from CQ to artemether-lumefantrine or artesunate-amodiaquine as first lines treatments for acute uncomplicated malaria in Nigeria, although, chloroquine has not been completely withdrawn, despite the change in drug policy. It has been argued that the withdrawal of CQ or SP as first line treatment for uncomplicated malaria in other disease endemic settings of Africa has led to the reemergence of drug sensitive Plasmodium falciparum [8–10]. The fact that in-vitro CQ resistance (54%) is higher than in-vivo resistance (38%), may be attributed to immunopotentiation of antimalarial drugs by patients immunity, which would have helped older patients (>5years of age) clear drug CQ resistant parasites.

Previous reports from malaria endemic areas have shown that children over the age of 5 years who have acquired some level of immunity to Plasmodium falciparum can clear drug resistant parasites [26, 32–36]. The value of in-vitro susceptibility testing of P. falciparum for elucidating the epidemiology of drug resistant malaria cannot be overemphasized. This study showed a strong association (p=0.029) between resistance to CQ in-vitro and in-vivo, confirming the fact that despite the technical challenges posed by the in-vitro technique in disease endemic countries of Africa, it is still a very useful tool for monitoring drug resistance. To our knowledge, this study is one of the very few studies in West Africa that shows a correlation between Plasmodium falciparum resistance to CQ in-vitro and in-vivo resistance to the drug.

This study showed similarities in the parasite msp-2 genotype from pre and post treatment infections in 72% (23 of 32) of the children who failed CQ treatment. The rest of the isolates (9) obtained from children that failed CQ treatment showed the presence of new infections in the post treatment isolates in addition to similarities in the pre and post treatment infections. These new infections observed in the post treatment isolates are confirming the polyclonality of infections observed in this study and in previous study in the same area [24, 26, 32]. It is possible that these new genotypes represent minor parasite populations in pre-treatment isolates amplified by PCR below the threshold of ethidium bromide detection. Polyclonality of infections may also have implications in epidemiology of antimalarial drug resistance. Ten patients who failed CQ treatment in spite of the parasite sensitivity to CQ in-vitro is one of the discrepancies between in-vitro and clinical outcome. One possible reason to this may be the complexity of infection as shown by msp-2 genotyping. It is possible that pre-treatment isolates consisted of parasites with different drug susceptibilities. In this case, a major population sensitive to CQ would have been cleared by the drug leaving a minor population which is resistant to CQ.

This may be supported by the fact that 6 of these isolates have borderline in-vitro sensitivity to CQ and, from the msp-2 genotyping, a significant (p= 0.03) reduction was observed in the average number of clones (3.8) in the post treatment isolates compared to pre treatment isolates (5.3).

Although point mutations in pfcrt and pfmdr1 genes as molecular markers of CQ resistance represent valuable tools for surveillance and monitoring changes in CQ efficacy, the interplay between polymorphisms in these two genes is not fully understood, as they may account for the differences observed in different malaria epidemiological settings [37]. This study analyzed the association of mutations at codon 76 of pfcrt gene and codons 86, 184, 1034 and 1042 of pfmdr1 gene with in-vitro and in-vivo CQ resistant or sensitive phenotypes. The pfmdr1Y184F mutation was identified in 14% (10 of 70) of isolates analyzed in this study. This point mutation has been reported to be associated with higher in-vitro resistance in laboratory strains [14] and field isolates from South America [38, 39]. No isolate analyzed in this study showed mutants Pfmdr1C1034 or pfmdr1D1042. This is in contrast with previous observations made in South America where these two alleles have been associated with high grade CQ resistance in isolates of P. falciparum [39, 40].

Correlation analysis between mutations in pfcrt and pfmdr1 genes and in-vitro CQ susceptibility in 82 isolates for which both in-vitro and molecular data were available, showed a very interesting feature. Some isolates of P. falciparum harboring the mutant pfcrtT76 and pfmdr1 Y86 alleles were sensitive to CQ in-vitro. These finding are similar to earlier reports from Senegal [41], the Philippines [42] and South East Asia [24, 43].

The occurrence of molecular markers of CQ resistance (mutant pfcrtT76 or pfmdr1 Y86 alleles) in some isolates of P. falciparum that are CQ sensitive in-vitro can be attributed to either the involvement of compensating mutations in either these two genes or other P. falciparum genes not previously associated with modification of drug response.

In addition, a recent report from Johnson and colleagues [43] showed that the presence of a mutation at codon 163 of the pfcrt (S163R) gene, where a serine (S) is replaced by an arginine (R) is associated with CQ sensitivity in some isolates of Plasmodium falciparum, despite the presence of the pfcrtT76 mutation. However, the role of parasites population dynamics on the detection of either the mutant or wild-type alleles of both pfcrt and pfmdr1 depends on the predominant parasites populations in patients’ isolates and cannot be ignored in an area of intense transmission like Ibadan, Nigeria. Further studies are needed for a better understanding of the association between in-vitro susceptibility of parasites to CQ, markers of CQ resistance and clinical outcome in this area of high malaria transmission.

Although there was no association between in-vitro CQ resistance and mutant pfmdr1Y86 alone, association between the double mutant pfcrtT76+pfmdr1Y86 or pfmdr1Y86-Y184 haplotype confirms the involvement of pfmdr1 in CQ resistance in-vitro. This is in agreement with previous observations in South East Asia where pfmdr1 Y86-Y184-S1034-N1042 haplotype is associated with CQ resistance in cultured isolate [44]. Previous reports [21, 24, 26, 45] have shown the primary role of the mutant pfcrtT76 as the major determinant of CQ resistance while mutation in pfmdr1 gene plays a modulatory role in the mechanism of resistance. It is also possible that the presence of wild type pfmdrY184 in isolates harboring mutant pfmdr1Y86 also confer more fitness on the parasite against CQ.

The association observed between mutant pfcrtT76 allele+pfmdr1 Y86-Y184 haplotype and in-vitro CQ resistance in our study further confirms the role of polymorphisms in these two genes in the mechanism of CQ resistance in P. falciparum. This is further strengthened by one interesting observation in our study, as we show an association (p=0.011) between in-vitro CQ sensitivity and the presence of wild type pfcrtK76 allele+ any pfmdr1 haplotype (Table 5). This observation demonstrates and confirms the primary and important role of pfcrt polymorphisms, irrespective of pfmdr1 polymorphisms and perhaps other genes in P. falciparum CQ resistance.

It has been difficult to establish an association between the Plasmodium falciparum mutations and patients’ treatment outcomes in Ibadan, Nigeria [24] and in many other malaria endemic countries of Africa and Madagascar [37]. The high prevalence (60%) of the mutant pfcrtT76 allele in pre-treatment isolates of all patients treated with CQ in this study makes it difficult for this allele to be predictive of treatment failures. This high prevalence of the pfcrtT76 allele is consistent with rates ranging from 60% to 100% reported in other malaria endemic regions [13, 22, 23, 33, 41, 46–48].

Unlike the observations made in-vitro, a significant association was observed between CQ treatment failure and pfmdr1 Y86-Y184 haplotype (p= 0.022) or mutant pfcrtT76+pfmdr1 Y86-Y184 haplotype (p=0.024) (Table 6). The significant association between in-vivo CQ failure and pfmdr1 Y86-Y184 haplotype or mutant pfcrtT76+pfmdr1Y86-Y184 confirms that polymorphisms on these two genes are involved in in-vivo CQ resistance. The mutant pfmdr1 Y86-F184 haplotype was observed to significantly (p=0.007) increase in post treatment isolates compared to pre treatment samples indicating a selection of the haplotype. The reasons for this selection remain unclear.

Some patients treated with standard doses of CQ cleared infections containing mutant allele of pfcrt and/or pfmdr1 that have been associated with both in-vitro and in-vivo CQ resistance. The ability of such patients to clear parasites with mutant genotype may be due to acquired immunity, as previously shown in other studies [26, 32, 35].

CONCLUSIONS

Overall, this study showed that the mutant pfcrtT76+pfmdr1Y86-Y184 haplotype is associated with in-vitro and in-vivo CQ resistance and can be used to identify Plasmodium falciprum resistant phenotypes in isolates from Nigeria. To our knowledge, this is one of the few studies in West Africa that clearly demonstrates the role of parasites mutations in phenotypic CQ resistance in-vitro and in-vivo. Further studies in other malaria endemic countries are needed in order to validate these findings, especially, in areas where ACTs have been introduced, and there is a need to monitor the return of CQ sensitivity.