Optimally timing primaquine treatment to reduce Plasmodium falciparum transmission in low endemicity Thai-Myanmar border populations

Abstract

Background: Effective malaria control has successfully reduced the malaria burden in many countries, but to eliminate malaria, these countries will need to further improve their control efforts. Here, a malaria control programme was critically evaluated in a very low-endemicity Thai-Myanmar border population, where early detection and prompt treatment have substantially reduced, though not ended, Plasmodium falciparum transmission, in part due to carriage of late-maturing gametocytes that remain post-treatment. To counter this effect, the WHO recommends the use of a single oral dose of primaquine along with an effective blood schizonticide. However, while the effectiveness of primaquine as a gametocidal agent is widely documented, the mismatch between primaquine's short half-life, the long-delay for gametocyte maturation and the proper timing of primaquine administration have not been studied.

Methods: Mathematical models were constructed to simulate 8-year surveillance data, between 1999 and 2006, of seven villages along the Thai-Myanmar border. A simple model was developed to consider primaquine pharmacokinetics and pharmacodynamics, gametocyte carriage, and infectivity.

Results: In these populations, transmission intensity is very low, so the P. falciparum parasite rate is strongly linked to imported malaria and to the fraction of cases not treated. Given a 3.6-day half-life of gametocyte, the estimated duration of infectiousness would be reduced by 10 days for every 10-fold reduction in initial gametocyte densities. Infectiousness from mature gametocytes would last two to four weeks and sustain some transmission, depending on the initial parasite densities, but the residual mature gametocytes could be eliminated by primaquine. Because of the short half-life of primaquine (approximately eight hours), it was immediately obvious that with early administration (within three days after an acute attack), primaquine would not be present when mature gametocytes emerged eight days after the appearance of asexual blood-stage parasites. A model of optimal timing suggests that primaquine follow-up approximately eight days after a clinical episode could further reduce the duration of infectiousness from two to four weeks down to a few days. The prospects of malaria elimination would be substantially improved by changing the timing of primaquine administration and combining this with effective detection and management of imported malaria cases. The value of using primaquine to reduce residual gametocyte densities and to reduce malaria transmission was considered in the context of a malaria transmission model; the added benefit of the primaquine follow-up treatment would be relatively large only if a high fraction of patients (>95%) are initially treated with schizonticidal agents.

Conclusion: Mathematical models have previously identified the long duration of P. falciparum asexual blood-stage infections as a critical point in maintaining malaria transmission, but infectiousness can persist for two to four weeks because of residual populations of mature gametocytes. Simulations from new models suggest that, in areas where a large fraction of malaria cases are treated, curing the asexual parasitaemia in a primary infection, and curing mature gametocyte infections with an eight-day follow-up treatment with primaquine have approximately the same proportional effects on reducing the infectious period. Changing the timing of primaquine administration would, in all likelihood, interrupt transmission in this area with very good health systems and with very low endemicity.

Figure 1

Schematic illustration of the malaria transmission model. The diagram shows the compartments of human population (Top) and mosquito population (Bottom) and transmission processes. For human population: S_h = susceptible; L_h = liver-stage only; A_h = asymptomatic infection with gametocyte carriage; C_h = clinical episode; and G_h = gametocyte carriage only. For mosquito population: M = susceptible; Y = infected; and Z = infectious.

Figure 2

Schematic illustration of the model for gametocytogenesis process. An asexual parasite produces 16 new parasites at every erythrocytic schizogony cycle. A proportion of these asexual parasites convert to immature gametocytes that appear in the peripheral blood for a day before sequestrating on the blood vessel. The gametocytes become visible again when the maturation process is completed. A combination of schizonticidal and gametocidal treatment affects the gametocyte production process by eliminating the asexual parasites and the mature gametocytes.

Figure 3

Relationship among initial gametocyte density, gametocyte longevity, and duration of infectiousness. (A) Changes in gametocyte density over time according to the initial gametocyte density (per μL): 10⁵ (black line), 10⁴ (blue line), 10³ (red line). (B) Infectivity to mosquito as a function of gametocyte density (per μL). (C) Probability of infectivity to mosquito over time at different initial gametocyte densities. (D) Duration of infectiousness related to initial gametocyte density without (solid line) and with (dotted line) primaquine follow-up treatment.

Figure 4

Plasmodium falciparum incidence and dynamic changes in proportion of the human population. (A) Incidence of P.falciparum malaria in seven villages of Ratchaburi province. Solid line represents the observed incidence from 1999 to 2006. The incidence from 1999 to 2009, estimated by the transmission model, is shown by the blue dashed line. (B) Estimated changes in proportion of population in each human compartment over the 10-year period. Solid line indicates population with clinical symptoms, C_h; Red dashed line represents population with gametocytaemia, G_h; Blue dotted line indicates asymptomatic population, A_h.

Figure 5

Gametocyte dynamics and duration of infectiousness with different timings of primaquine administration. (A) Changes in the density of each parasite developmental stage over time with different timings of primaquine administration. Pharmacokinetic curve of primaquine (dotted line) related to waves of asexual parasite, A_g(dashed line), immature gametocytes, I_g (grey line), and mature gametocytes, M_g (solid red line). Initial asexual parasite (A_g) 10/μL (B) Duration of infectivity of humans to mosquitoes over different timings of primaquine administration after acute attack.

Figure 6

The R₀/Rc ratio for ACT and follow-up primaquine treatment. The relationship between R₀/R_c and the proportion of P. falciparum patients receiving follow-up primaquine treatment (Q) for different proportions of patients treated with artesunate-mefloquine combination therapy (P).

Figure 7

Estimated P. falciparum malaria incidence for different control policies. Solid line represents the incidence estimated by the initial model when the standard drug regimen (2-day ACTs and Primaquine at day 2) was used (solid line). The incidence significantly decreased when the timing of primaquine administration was shifted to the eighth day after initial attack (short-dashed line). The elimination of malaria transmission can be reached when the combination of both primaquine follow-up treatment and control of imported asymptomatic cases was implemented (long-dashed line).