During May–September 2011, mosquito larvae were collected in irrigated rice fields surrounding Tiassalé, southern Côte d’Ivoire (5°52′47′′N; 4°49′48′′W) and reared to adults in insectaries on a diet of MikroMin (Tetra, Melle, Germany) fish food. A total of 1,571 adult female An. gambiae s.l. mosquitoes, 3–5 days of age, were exposed to 1 of 5 insecticides (0.1% bendiocarb, 1.0% fenitrothion, 0.75% permethrin, 0.05% deltamethrin, 4% DDT) or a control papers for 1 hour, according to standard WHO procedures (7). Mosquito deaths were recorded 24 hours later. DNA was extracted from individual mosquitoes according to the LIVAK method (8), and a subsample of 500 mosquitoes were all found to be the M molecular form of An. gambiae s.s. by using the SINE-PCR method (9). The target site mutation G119S in the Ace-1 gene (Ace-1^R) and L1014F and L1014S kdr mutations were screened by using restriction fragment length polymorphism (10) or TaqMan assays (11), respectively.

According to WHO criteria, An. gambiae mosquitoes from Tiassalé are resistant to all insecticide classes, and resistance is extremely prevalent; more than two thirds of mosquitoes survived the diagnostic dose for 4 of the 5 insecticides tested (Table 1). To assess the level of resistance, we exposed the Tiassalé population and a susceptible laboratory population of An. gambiae (Kisumu) mosquitoes to the pyrethroid deltamethrin or the carbamate bendiocarb for a range of exposure times and assessed deaths 24 hours later (Technical Appendix). We found an unexpectedly strong resistance phenotype to the 2 insecticides (Figure 1, Figure 2). For deltamethrin, 4 hours of exposure were required to kill 50% (median lethal time, [LT₅₀]); in comparison, the LT₅₀ for the susceptible Kisumu strain was <2 minutes (resistance ratio = 138) (Technical Appendix). Similarly, the LT₅₀ for bendiocarb was nearly 5 hours for the Tiassalé strain yet <12 minutes for the susceptible strain (resistance ratio = 24) (Technical Appendix).

To investigate the causes of this resistance, we screened a subset of mosquitoes for the target site mutations, kdr 1014F and 1014S. Only the 1014F kdr mutation was detected, and this resistance allele was found at high frequency (83%). There was a significant association between presence of the 1014F kdr allele and ability to survive exposure to DDT but not to either pyrethroid (Table 2). In contrast, the Ace-1^R allele was strongly associated with survival after exposure to bendiocarb and fenitrothion (Table 2).

Pyrethroid resistance in An. gambiae mosquitoes was first reported from Côte d’Ivoire in 1993 (12); carbamate resistance was detected in the 1990s (13). Nevertheless, ≈2 decades later, it is surprising and worrying to find complete resistance to all insecticides tested, particularly—for deltamethrin and bendiocarb—at such high levels. Resistance mechanisms seem to be varied. Ace-1^R is strongly associated with organophosphate and carbamate resistance, and the absence of 119S homozygotes might be attributable to the high fitness cost of the Ace-1^R allele in the absence of insecticide (14). Presence of the 1014F kdr allele alone does not confer the ability to survive diagnostic doses of pyrethroids; thus, alternative mechanisms must be responsible for the high-level pyrethroid resistance in this population.

The selective pressures responsible for this intense multiple-insecticide resistance in Tiassalé mosquitoes are unclear. There is a high coverage of insecticide-treated bed nets, but this coverage does not differ from that in other parts of the continent, and indoor residual spraying has not been conducted in this region. Use of insecticides in agriculture has been linked to resistance in malaria vectors. This use is perhaps the most likely explanation in this district of intense commercial production of rice, cocoa, and coffee.

Whatever the cause, the implications of this resistance scenario for malaria control are severe. With no new classes of insecticides for malaria control anticipated until 2020 at the earliest (15), program managers have few options available when confronted with multiple-insecticide resistance. Assessing the effect of pyrethroid resistance on the efficacy of insecticide-treated bed nets is complex because of the poorly understood associations between net integrity, insecticide content, net usage, and net efficacy. Nevertheless, resistance levels, such as those reported here, combined with continual selection pressure will inevitably lead to suboptimal mosquito control by use of insecticide-treated bed nets and indoor residual spraying. If unchecked, this resistance could spread rapidly and threaten the fragile gains that have been made in reducing malaria across Africa.

Time-mortality curve for Anopheles gambiae mosquitoes, Kisumu strain, exposed to deltamethrin and bendiocarb, and time-death data for adult female A. gambiae s.s. mosquitoes, Tiassalé strain, and standard susceptible colony Kisumu 24 hours after exposure to bendiocarb or deltamethrin.