We collected retrospective data from the National Malaria Reference Centre (Paris, France) and Public Health Ontario (Toronto, ON, Canada) during 1988–2010 and 2007–2010, respectively. P. falciparum infection was considered probably acquired in Haiti if biologically confirmed by thin and thick blood smears from persons who had recently traveled to Haiti in the 2 months before infection was diagnosed. Basic demographic and epidemiologic data, clinical and parasitologic information, treatment, and history of travel and P. falciparum infection were collected systematically. Forty of 80 participating hospitals in the sentinel network in France also documented resistance to antimalarial drugs. Pretreatment isolates were collected to determine chloroquine susceptibility by molecular analysis of the Pfcrt76 locus and by comparing the ratio of in vitro chloroquine response of the clinical isolate with a chloroquine-sensitive reference clone.

Figure 1

Figure 1. . . Surveillance during 23 years for antimalarial drug resistance in travelers returning to France and Canada from Haiti after the January 10, 2010, earthquake. A) Imported malaria cases from Haiti...

Seventy-nine imported P. falciparum infections were recorded: 49 before the earthquake (all in France) and 30 after the earthquake (3 in Canada and 27 in France). The number of confirmed malaria cases imported from Haiti doubled during 2009–2010 (Figure 1). Approximately half of the travelers were in Haiti 2–4 weeks before the earthquake and >1 month after the earthquake. The main purpose of travel, visiting friends and relatives, decreased from 59% before to 44% after the earthquake. More than 75% of travelers did not take prophylactic medication. The proportion of severe malaria increased from 3% to 11% after January 2010 (Table 1).

Figure 2

Figure 2. . . Departments of Haiti.

Before the earthquake, all 29 isolates had the wild-type PfcrtK76 allele according to analysis by PCR–restriction fragment-length polymorphism. The mean 50% inhibitory concentration (IC₅₀) of chloroquine for the 24 isolates tested ex vivo by the ³H-hypoxanthine uptake inhibition method was 27 nM (95% CI 23–31). These results are consistent with those of an unpublished study conducted in Haiti during 2007 to monitor chloroquine resistance (Jean-François Vely, unpub. data). In that study, Haiti’s National Malaria Program, in collaboration with the National Malaria Reference Centre in France, found the chloroquine-sensitive genotype in 146 P. falciparum–positive samples in 6 departments (Artibonite, Centre, Grand’Anse, Nord, Nord-Ouest, Ouest) (Figure 2) (12). After the earthquake, 2 (11%) of 19 isolates analyzed by pyrosequencing and PCR–restriction fragment-length polymorphism showed a mixed Pfcrt76K+T genotype. The ratios of K to T genotypes before and after in vitro adaptation were 0.75:0.25 and 0.23:0.77, respectively, for patient 1, and 0.58:0.42 and 0.25:0.75, respectively, for patient 2. The Pfcrt72–76 haplotype was CVMNK before adaptation and CVIET after adaptation for both patients by sequencing. Resistance was confirmed by in vitro methods after culture adaptation. Both isolates had high chloroquine IC₅₀ (506 nM and 708 nM, respectively) and high chloroquine IC₅₀ isolate:Pf3D7 (chloroquine susceptible clone) ratio (20 and 27, respectively) (Table 2).

Patient 1, a 58-year-old woman, was in Haiti during October 2009–January 2010; she returned after the earthquake to Canada, where she sought care for malaise, fever, diarrhea, and vomiting. She reported no previous malaria and no other travel during the previous 2 years. For patient 2, a 16-year-old girl, malaria was diagnosed in Canada on February 25, 2010, after 3 days of fever. She had traveled to Haiti in the past 2 months before malaria was diagnosed and did not report any other recent travel.

The number of P. falciparum–infected travelers returning from Haiti has increased since January 2010, probably because of the higher number of aid workers and visitors and increased P. falciparum malaria transmission. Data suggest that the earthquake and ensuing hurricane and floods created the necessary conditions—inadequate shelters, population movement, and still water—to increase the incidence of malaria and possibly spread the recently identified chloroquine-resistant strains of P. falciparum (10). In France and Canada, laboratory surveillance for malaria found that 2 travelers from Haiti carried chloroquine-resistant strains. In vitro culture might have selected resistant strains not observed initially by ex vivo methods. After carefully interviewing these patients about their travels, we found no evidence to cause doubt that they had acquired malaria in Haiti. Alternatively, the resistant strains could have come to Haiti after the earthquake through human activity, as occurred in the cholera outbreak (13). The origin of the chloroquine-resistant strains identified in Haiti is uncertain. The Pfcrt CVIET haplotype is common in Southeast Asia and sub-Saharan Africa and was found in the 2006–2007 study in Haiti (10).

Regardless of origin, containing the spread of chloroquine-resistant parasites is crucial. Malaria elimination is a goal in Haiti, and it has been strengthened after recent events, but the effects of malaria and many other factors affect the achievability of this goal (14). Control measures, possibly mirroring those used to contain artemisinin resistance in Southeast Asia, should be concentrated in Haiti to prevent resistance spreading to the rest of Hispaniola (15). However, lack of consensus on the use of molecular and in vitro data for policy change will hamper decision making. Neither the chloroquine-resistant Pfcrt76T genotype nor the elevated chloroquine IC₅₀ perfectly predicts treatment failure because of confounding factors like acquired immunity.

Our study has several limitations. Returning travelers are not a representative sample of the Haitian population, and the sample of isolates was limited. The origin of the resistant strains is not defined. Also, the precise location of infection is not reported. Nevertheless, travelers are useful sentinels of emerging resistance in areas where little information is available, providing surveillance data in real time with standardized methods. This nonimmune population also facilitates detection of resistant isolates.

Our data highlight the need to implement a therapeutic efficacy surveillance study for assessing in vivo chloroquine sensitivity, which is essential for providing information for rational control strategies and guiding prophylaxis recommendations in Haiti. In addition, physicians treating malaria acquired in Haiti should be aware of the possibility of chloroquine-resistant infections. Patients with persistent fever despite treatment and infected travelers reporting adherence to chloroquine prophylaxis should be treated with alternate antimalarial drug therapy.