Longitudinal impact of asymptomatic malaria/HIV-1 co-infection on Plasmodium falciparum gametocyte transcript expression and transmission to Anopheles mosquitoes

Abstract

Despite significant developments towards malaria reduction, parasite transmission in the common context of HIV-1 co-infection and treatment for one or both infections has not been fully characterized. This is particularly important given that HIV-1 and malaria chemotherapies have the potential to alter gametocyte burden and mosquito infectivity. In this study, we examined 782 blood samples collected from a longitudinal cohort of 300 volunteers with asymptomatic parasitemia seeking HIV testing or treatment in the endemic region of Kisumu, Kenya, to define the impacts of HIV-1-malaria co-infection, antiretroviral therapy (ART) plus trimethoprim-sulfamethoxazole (TS) and the antimalarials artemether/lumefantrine (AL) on Plasmodium falciparum gametocyte transcript prevalence and parasite transmission to the African malaria mosquito Anopheles gambiae. Volunteers were assigned to three distinct HIV-1 groups: HIV-1 positive on treatment, HIV-1 positive newly diagnosed, and HIV-1 negative. Volunteers were monitored monthly over the course of six months. Using our highly sensitive digital droplet PCR (ddPCR) assay of three gametocyte specific transcript markers, we detected gametocyte transcripts in 51.1% of 18S positive volunteers across all study groups and time points. After correcting for multiple comparisons, the factors of HIV-1 status, time, CD4+ T-cell levels and hematocrit were not predictive of gametocyte prevalence or transmission. However, among those volunteers who were newly diagnosed with HIV-1 and malaria positive by rapid diagnostic test (RDT) at enrollment, the initiation of ART/TS and AL treatment was associated with a significant reduction in gametocyte transcript prevalence in the subsequent month when compared to HIV-1 negative volunteers treated with AL. To assess gametocyte transmissibility, volunteer blood samples were used in standard membrane feeding assays (SFMA) with laboratory-reared A. gambiae, with evidence of transmission confirmed by at least one of 25 dissected mosquitoes per sample positive for at least one midgut oocyst. HIV-1 status, CD4+ T-cell levels and hematocrit were not significantly associated with successful transmission to A. gambiae. Analysis of SMFA blood samples revealed that 50% of transmission-positive blood samples failed to test positive by Plasmodium-specific 18S ribosomal RNA quantitative PCR (qPCR) and 35% failed to test positive for any gametocyte specific transcript marker by droplet digital (ddPCR), documenting that transmission occurred in the absence of molecular parasite/gametocyte detection. Overall, these findings highlight the complexity of HIV-1 malaria co-infection and the need to further define the unpredictable role of asymptomatic parasitemia in transmission to mosquitoes.

Keywords: HIV; co-infection; ddPCR; gametocytes; malaria; mosquito infectivity; plasmodium falciparum; transmission.

Figure 1

Study Profile Overview. Participant enrollment by study group and corresponding molecular data for asexual and sexual parasite biomarkers. Groups below the solid black line were included in this study for gametocyte analysis. A subset of volunteers was enrolled in a transmission study using the Standard Membrane Feed Assay (dotted line). Positive samples are abbreviated as (+) or POS. Negative samples are abbreviated as (-) or NEG.

Figure 2

Forest Plot of the Log Odds of Gametocyte Positivity in Month 0 and Month 1. The log odds and accompanying exponentiated adjusted odds ratio of being positive for at least one gametocyte specific transcript between each study group in month 0 (M0) and in month 1 (M1). “HIV Neg” represents HIV-1 negative volunteers. “HIV Pos OT” represents HIV-1 positive volunteers on treatment. “HIV Pos ND” represents HIV-1 positive newly diagnosed volunteers. P-values were adjusted using the Holm’s method for multiple comparisons. Log odds are arranged in descending order. Blue confidence interval lines are comparisons within month 1 and black confidence interval lines are comparisons within month 0.

Figure 3

Percent Positivity of Each Gametocyte-Specific Marker Over Time for All Study Groups. The prevalence of each gametocyte-specific transcript biomarker (as a percentage of total analyzed) at each time point/visit.

Figure 4

Percent Positivity of Each Gametocyte-Specific Marker Over Time for Each Study Group. The prevalence of each gametocyte-specific transcript biomarker (as a percentage of total analyzed) at each time point for (A) HIV-1 positive newly diagnosed volunteers (B) HIV-1 negative volunteers and (C) HIV-1 positive volunteers on treatment.

Figure 5

Forest Plot of the Log Odds of Pfs25 Positivity in HIV-1 Positive Newly Diagnosed Volunteers. The log odds and accompanying exponentiated adjusted odds ratio of being positive for at least one gametocyte-specific transcript within the HIV-1 positive newly diagnosed study group between month 0 (M0) and between month 1 (M1). P-values were adjusted using the Holm’s method for multiple comparisons. Blue confidence interval lines are comparisons with month 0 and black confidence interval lines are comparisons with month 1.

Figure 6

Forest Plot of the Log Odds of Pfs25 Positivity in HIV-1 Negative Volunteers. The log odds and accompanying exponentiated adjusted odds ratio of being positive for at least one gametocyte-specific transcript within the HIV-1 negative study group between month 0 (M0) and between month 1 (M1). P-values were adjusted using the Holm’s method for multiple comparisons. Blue confidence interval lines are comparisons with month 0 and black confidence interval lines are comparisons with month 1.

Figure 7

Forest Plot of the Log Odds of PfMGET Positivity in HIV-1 Positive Newly Diagnosed Volunteers. The log odds and accompanying exponentiated adjusted odds ratio of being positive for at least one gametocyte specific transcript within the HIV-1 positive newly diagnosed study group between month 0 (M0) and between month 1 (M1). P-values were adjusted using the Holm’s method for multiple comparisons. Blue confidence interval lines are comparisons with month 0 and black confidence interval lines are comparisons with month 1.

Figure 8

Forest Plot of the Log Odds of PfMGET Positivity in HIV-1 Negative Volunteers. The log odds and accompanying exponentiated adjusted odds ratio of being positive for at least one gametocyte specific transcript within the HIV-1 negative study group between month 0 (M0) and between month 1 (M1). P-values were adjusted using the Holm’s method for multiple comparisons. Blue confidence interval lines are comparisons with month 0 and black confidence interval lines are comparisons with month 1.

Figure 9

Oocyst Positivity by Study Group and Time. A distribution of the total number of oocyst positive samples as percentage of the total analyzed at each time point for (A) HIV-1 positive newly diagnosed volunteers and (B) HIV-1 negative volunteers. As part of the study design, HIV-1 negative volunteers did not have a scheduled visit at week 1 or week 2, so no samples were analyzed at these time points.

Figure 10

Presence of Gametocytes and Asexual Parasites at Month 1 by Study Group. The column graph on the top left (A) shows the presence of gametocytes by ddPCR after AL treatment at month 1 if the volunteer was positive for malaria by RDT at month 0. The column graph on the top right (B) shows the presence of gametocytes by ddPCR without AL treatment at month 1 if the volunteer was negative for malaria by RDT at month 0. The column graph on the bottom center (C) shows the presence of asexual parasites by 18S qPCR after AL treatment at month 1 if the volunteer was positive for malaria by RDT at month 0.