Revisiting the circulation time of Plasmodium falciparum gametocytes: molecular detection methods to estimate the duration of gametocyte carriage and the effect of gametocytocidal drugs

Abstract

Background: There is renewed acknowledgement that targeting gametocytes is essential for malaria control and elimination efforts. Simple mathematical models were fitted to data from clinical trials in order to determine the mean gametocyte circulation time and duration of gametocyte carriage in treated malaria patients.

Methods: Data were used from clinical trials from East Africa. The first trial compared non-artemisinin combination therapy (non-ACT: sulphadoxine-pyrimethamine (SP) plus amodiaquine) and artemisinin-based combination therapy (ACT: SP plus artesunate (AS) or artemether-lumefantrine). The second trial compared ACT (SP+AS) with ACT in combination with a single dose of primaquine (ACT-PQ: SP+AS+PQ). Mature gametocytes were quantified in peripheral blood samples by nucleic acid sequence based amplification. A simple deterministic compartmental model was fitted to gametocyte densities to estimate the circulation time per gametocyte; a similar model was fitted to gametocyte prevalences to estimate the duration of gametocyte carriage after efficacious treatment.

Results: The mean circulation time of gametocytes was 4.6-6.5 days. After non-ACT treatment, patients were estimated to carry gametocytes for an average of 55 days (95% CI 28.7 - 107.7). ACT reduced the duration of gametocyte carriage fourfold to 13.4 days (95% CI 10.2-17.5). Addition of PQ to ACT resulted in a further fourfold reduction of the duration of gametocyte carriage.

Conclusions: These findings confirm previous estimates of the circulation time of gametocytes, but indicate a much longer duration of (low density) gametocyte carriage after apparently successful clearance of asexual parasites. ACT shortened the period of gametocyte carriage considerably, and had the most pronounced effect on mature gametocytes when combined with PQ.

Figure 1

Models describing change in gametocyte density and gametocyte prevalence over time. S, sequestered gametocytes; G, circulating gametocytes; E, gametocyte-negative infected individuals (i.e. gametocytes not yet released into circulation); I, gametocyte-positive individuals; ρ rate of release of gametocytes from sequestration into the bloodstream; μ, rate of decay/removal of gametocytes; f, rate at which gametocyte-negative individuals become gametocyte-positive; r, rate at which gametocyte-positive individuals become gametocyte negative.

Figure 2

The mean gametocyte circulation time based on Pfs25 QT-NASBA gametocyte density after non-ACT and ACT. Log gametocyte density/μL is given on the Y-axis, the day of follow up after initiation of treatment. Symbols and error bars indicate the field data with 95% confidence interval, lines fitted values. The dotted line indicates the lower threshold for gametocyte detection by QT-NASBA, 0.02 gametocytes/μL. The trial was conducted in Kenya; treatment regimens were non-ACT (open diamonds; SP+AQ administered on day 0-2) and ACT (closed triangles; SP+AS or AL administered on day 0-2). The estimated mean circulation time of gametocytes in this trial was 6.53 days (95% CI 4.84-8.80) after non-ACT treatment and 5.04 days (95% CI 4.20-6.06) after ACT treatment, based on data between d3 and d28.

Figure 3

The mean gametocyte circulation time based on Pfs25 QT-NASBA gametocyte density after ACT and ACT-PQ. Log gametocyte density/μL is given on the Y-axis, the day of follow up after initiation of treatment. Symbols and error bars indicate the field data with 95% confidence interval, lines fitted values. The dotted line indicates the lower threshold for gametocyte detection by QT-NASBA, 0.02 gametocytes/μL. The trial was conducted in Tanzania; treatment regimens were ACT (closed triangles; SP+AS administered on day 0-2) and ACT-PQ (open squares; SP+AS administered on day 0-2 followed by a single dose of PQ on day 2). The mean circulation time of gametocytes in this trial was 4.61 days (2.92 - 7.26) after ACT treatment and 0.53 days (0.24-1.19) after ACT-PQ treatment.

Figure 4

The duration of gametocyte carriage based on the Pfs25 QT-NASBA gametocyte prevalence after non-ACT and ACT treatment. Symbols and error bars indicate the field data with 95% confidence interval, lines fitted values. The trial was conducted in Kenya; treatment regimens were non-ACT (open diamonds; SP+AQ administered on day 0-2) and ACT (closed triangles; SP+AS or AL administered on day 0-2). The estimated average duration of gametocyte carriage was 55.6 days (95% CI 28.7-107.7) after non-ACT treatment and 13.4 days (95% CI 10.2-17.5) after ACT treatment.

Figure 5

The duration of gametocyte carriage based on the Pfs25 QT-NASBA gametocyte prevalence after ACT and ACT-PQ treatment. Symbols and error bars indicate the field data with 95% confidence interval, lines fitted values. The trial was conducted in Tanzania; treatment regimens were ACT (closed triangles; SP+AS administered on day 0-2) and ACT-PQ (open squares; SP+AS administered on day 0-2 followed by a single dose of PQ on day 2). The estimated average duration of gametocyte carriage was 28.6 days (95% CI 17.0 - 48.0) after ACT treatment and 6.3 days (95% CI 4.7-8.5) after ACT-PQ treatment.