Projected malaria incidence by age and transmission intensity under PMC and/or RTS,S
To generate varying age-incidence curves, simulations were run in a geography-agnostic setup under annual EIRs ranging from one to 128 ibpa with fixed clinical treatment coverage of 60%.
In the absence of PMC or RTS,S, clinical malaria cases peaked at around two years of age at the highest simulated transmission intensity and shifted to older ages for lower transmission intensities. Severe malaria cases were highest between six months to one year of age across the simulated transmission levels and decreased to low numbers by the end of two years of age, and to very low numbers (< 1 case per 1000 population U2) by the end of three years (Fig 2A, Additional file 1 Fig A1.2.3A).
The overall impact of PMC-3 (PMC with 3 doses, Fig 1B) was greatest among children under the age of one year, and the impact of RTS,S after the first year of life, until four to five years of age (Fig 2B, Additional file 1 Fig A1.2.3B). Under PMC-3 at 80% coverage per dose, cases dropped by around 60% after every dose for one month before resurging to pre-dose levels. PMC-3 only averted cases in children U1 but not beyond, since the expected duration of effect of the last PMC-3 dose at nine months wanes before the child reaches one year of age. Under RTS,S at 80% coverage, and 80% coverage with the booster dose among those who received the primary series, cases were reduced by around 60% after the third priming dose and by 50% after the booster dose at high transmission. Protection after the priming series and booster lasted from several months to a few months depending on transmission intensity (Additional file 1 Fig A1.1.4).
For each of the scenarios, the cumulative number of cases averted in children U2 increased with increasing transmission intensity (EIR) until reaching highest level of simulated transmission of 128 ibpa (clinical incidence of 5,500 per 1,000 population U2, PfPRU2 75%), before dropping for clinical cases, whereas for severe cases continued increasing, although at lower rate (Fig 2C). The trends in impact by transmission were different for impact measures on relative scale, with the PE in clinical cases remaining constant before gradually decreasing after transmission intensity reached 16 ibpa (clinical incidence of 2,000 per 1,000 population U2). Whereas the PE in severe cases increased gradually by transmission, before dropping after transmission intensity reached 32 ibpa (clinical incidence of 3,000 per 1,000 population U2 (Fig 2D). These trends were more pronounced for RTS,S and the combination of both interventions but less so for PMC-3 alone which had a smaller impact (PEclinicalU2: PMC-3 5.7-8.8%; RTS,S 10-32%; PEsevereU2: PMC-3 6.1-13.6%; RTS,S 24.6-27.5%).
Under the PMC-3 scenario in combination with RTS,S, a larger impact than for either intervention alone was projected, with a greater additional impact by RTS,S than by PMC-3 (Fig 2C-D). For instance, compared to PMC-3 alone, the combination averted 2.8-4.5 times more clinical and 3.2-7.1 times more severe cases in children U2 on average across the simulated transmission levels. Whereas PMC-3 in combination with RTS,S compared to RTS,S alone averted 1.2-1.7 and 1.2-1.3 times more cases in children U2 for clinical and severe malaria respectively.
Influence of age at first PMC dose in a PMC-3 schedule
PMC was initially recommended with three doses given at 2.5, 3.5, and nine months of age [8]. At the time of the first dose, at 2.5 months of age, children might still be protected against severe disease by maternal antibodies [43], and the next dose closely follows at 3.5 months of age. Therefore, we assessed the single contribution of the 2.5-month dose to the overall PMC-3 impact in children U2 and evaluated the impact of changing its delivery to instead occur at 6, 12, or 15 months of age.
Overall, three doses were more impactful than two doses, indicating that the dose at 2.5 months does provides additional protection against clinical malaria although to a lesser extent against severe malaria (Fig 3). This finding is likely influenced by assumptions on maternal antibody protection in the model, which has a stronger effect on severe than on clinical malaria (Additional file 1 Fig A1.3.5). Shifting the 2.5-month dose to older ages resulted in an increased reduction in cases; the amount of increase varied by disease severity and transmission intensity. For clinical malaria, timing this dose to later ages (6, 12 or 15 months) was more impactful at low-to-high transmission, but not at very high transmission (EIR >=62 ibpa). For severe malaria, shifting the dose to 6 or 12 months of age increased impact across all transmission levels, whereas a shift to 15 months reduced its impact, especially at very high transmission (Fig 3).
Impact of additional PMC doses with or without RTS,S in children under 2 years of age
Five PMC schedules (Fig 1B) were compared to each other as well as to RTS,S in a simulation with high transmission (EIR=32 ibpa, 3000 cases per 1000 population U2), assuming a constant target coverage of 80% for each dose of PMC and for RTS,S (Fig 4A). Aggregated clinical and severe cases averted for children 0-2 years (children U2) and disaggregated into 0-1 and 1-2 years are shown in Fig 4B.
In children U2, RTS,S was projected to avert the most cases, followed by PMC scenarios based on how many doses were given. PMC with three to seven doses in the first 18 months of age (PMC-3, PMC-4, PMC-5, PMC-6, and PMC-7) averted on average 251 to 669 clinical cases per 1000 children U2 (PE ranging from 8.1% to 21.6%), compared to 774 (25.1%) cases averted by RTS,S and 1003 (32.6%) by the combination of RTS,S and PMC-3. For severe malaria, between two and nine cases were averted per 1000 children U2 (PE 7.3% - 31%) across the same PMC scenarios, compared to eleven (PE 39.8%) cases averted by RTS,S, and thirteen by the combination (PE 46.4%). Overall, PMC averted most cases during the first year of life, and RTS,S after the first year (Fig 4B). For all the PMC schedules tested, RTS,S combined with PMC had a synergistic effect, even under PMC-7, where overlap was greatest (Additional file 1 Fig A1.2.5).
The maximum PE reached in children U2, assuming a 100% coverage, ranged from 10% to 41% across the PMC-RTSS scenarios for clinical malaria and from 10% to 57% for severe malaria. (Fig 4C).
Operational impact of PMC and RTS,S in Southern Nigeria
We ran a second set of simulations where transmission intensity, seasonality, and PMC coverage correspond to the 20 States in Southern Nigeria that include PMC-eligible areas (Fig 5A). Malaria prevalence from the NDHS 2018 and PfPRU5-EIR relationship from the previous simulations were used to obtain annual EIR values for each State. Malaria prevalence ranged between 3.4 to 54.9% across States (mean 30.3%) and matched EIRs ranged from 1.1 to 27.5 ibpa (mean 11.9 ibpa) (Fig 5B, Additional file 1 Fig A1.4.3). The simulated malaria incidence in the absence of PMC or RTS,S ranged between 152 and 2,510 clinical cases and between 0.735 and 33.71 severe cases per 1000 children U2 per year across the States (Fig 5C). For a population of 6.5 million children U2, estimated for 2019 in Southern Nigeria, the projected malaria burden was around 7.6 million clinical and 69,000 severe cases in one year. The presented malaria case estimates include untreated and unreported cases. The operational PMC coverage was based on State-level estimates of EPI coverage reported in NDHS 2018 (mean 73.1%, range 40.8-95% for 3 doses across States), and downscaled to account for expected gaps between immunization and PMC coverage [3] (mean 60.8%, range 28.5-88.8%) (Fig 5D,F).
At operational coverage, PMC-3 was projected to avert annually on average 447,258 (95%CI 329,041 - 565,476) clinical and 4,409 (95%CI 2,950 - 5,868) severe cases in the U2 population (Fig 5E,G). As expected, the relative impact was strongly correlated with coverage, and the total number of cases averted was highest in States with higher population and malaria burden. PMC-5 was projected to avert nearly twice as many cases as PMC-3, with 779,263 (95%CI 589,395 - 969,131) clinical and 8,931 (95%CI 5,619 - 12,244) severe cases averted. PMC-7 was projected to avert 1,125,500 (95%CI 856,435 – 1,394,566) clinical and 13,360 (95%CI 8,314- 18,404) severe cases. In comparison, RTS,S was projected to avert 1,225,010 clinical cases (95%CI 951,960- 1,498,060) and 15,419 (95%CI 9,310 - 21,527) severe cases. Finally, the combination of RTS,S plus PMC-3 was projected to avert 1,647,743 clinical cases (95%CI 1,262,623 - 2,032,864) and 19,394 (95%CI 11,946 - 26,841) severe cases annually in children U2, a protective efficacy against clinical cases of 23.4% (95%CI 21.1-25.7%) and against severe cases of 29.9% (95%CI 27.3-31.6%).
If coverage were to increase to target levels of 80%, more cases could be averted (Fig 5G). For instance, an increase in the mean coverage of PMC-3 from 61% to 80% averted 41-46% more cases with additional 185,014 clinical and 2,063 severe cases averted per year per population U2. The additional cases averted when increasing coverage from operational to target levels increased with the number of doses. For PMC-5, an increase in coverage from 56% to 80% averted on average 442,936 additional clinical cases per year and 5,993 additional severe cases. For PMC-7 an additional 658,421 clinical and 8,983 severe cases were averted per population U2 per year when coverage increased from 55% to 80%. A similar trend was projected for RTS,S with 930,503 additional clinical and 12,004 severe cases averted, at target compared to operational coverage and for the combination of RTS,S and PMC-3 with additional 1,066,371 clinical and 12,926 severe cases averted, corresponding to protective efficacies of 36.6% and 46.4% against clinical and severe cases respectively (Fig 5G).