Mass campaigns combining antimalarial drugs and anti-infective vaccines as seasonal interventions for malaria control, elimination and prevention of resurgence: a modelling study

Background The only licensed malaria vaccine, RTS,S/AS01, has been developed for morbidity-control in young children. The potential impact on transmission of deploying such anti-infective vaccines to wider age ranges, possibly with co-administration of antimalarial treatment, is unknown. Combinations of existing malaria interventions is becoming increasingly important as evidence mounts that progress on reducing malaria incidence is stalling and threatened by resistance. Methods Malaria transmission and intervention dynamics were simulated using OpenMalaria, an individual-based simulation model of malaria transmission, by considering a seasonal transmission setting and by varying epidemiological and setting parameters such as transmission intensity, case management, intervention types and intervention coverages. Chemopreventive drugs and anti-infective vaccine efficacy profiles were based on previous studies in which model parameters were fitted to clinical trial data. These intervention properties were used to evaluate the potential of seasonal mass applications of preventative anti-infective malaria vaccines, alone or in combination with chemoprevention, to reduce malaria transmission, prevent resurgence, and/or reach transmission interruption. Results Deploying a vaccine to all ages on its own is a less effective intervention strategy compared to chemoprevention alone. However, vaccines combined with drugs are likely to achieve dramatic prevalence reductions and in few settings, transmission interruption. The combined mass intervention will result in lower prevalence following the intervention compared to chemoprevention alone and will increase chances of interruption of transmission resulting from a synergistic effect between both interventions. The combination of vaccine and drug increases the time before transmission resurges after mass interventions cease compared to mass treatment alone. Deploying vaccines and drugs together requires fewer rounds of mass intervention and fewer years of intervention to achieve the same public health impact as chemoprevention alone. Conclusions Through simulations we identified a previously unidentified value of deploying vaccines with drugs, namely the greatest benefit will be in preventing and delaying transmission resurgence for longer periods than with other human targeted interventions. This is suggesting a potential role for deploying vaccines alongside drugs in transmission foci as part of surveillance-response strategies.

Includes realistic levels, but also out-ofrange levels for a better understanding of the vaccine -MDA interactions and relative benefits Congruency between interventions and covered population The 3 vaccine doses are given to the same population (given coverage) 1 , but the fourth and the fifth dose are given to random population MDA is given to random proportion of the population (given coverage) for each round; and independent from vaccination 2 unless otherwise specified 3 1 assuming 100% adherence to the 3 immunization doses 2 both when the vaccine is delivered before or simultaneously to the MDA rounds 3 in a subset of the simulations, vaccination and MDA are delivered simultaneously to the same proportion of the population, given coverage

Target age of mass vaccination
Minimum age is 9 months for third dose, with first dose from 5 months of age Minimum age of 5 months at first vaccination is assumed, as intended following RTS,S implementation (5) Target age from MDA All ages but with minimum age from 6 months of age As defined previously (3), pregnant women were not explicitly excluded but would reflect lower coverages Population size 10'000 MDA has shown to have better efficacy in targeted elimination strategies of small populations (6) Monitored outputs from the simulations

Definition of transmission interruption for each simulation
On average across the 5-10 years post intervention deployment period, less than 1 infected individual in 10'000 Table S2: Overview of the main and supplementary simulated strategies. The strategies include: MDA alone (strategies 1 and 2), vaccine alone (strategies 3, 4, s3, s4), or MDA with vaccine (strategies 5 to 8, and s5 to s8). MDA application alone is 3 rounds coinciding with the pattern of seasonal transmission, with 2-3 years of 3 rounds or 3 rounds for only the first year followed by 1-2 years of 1 round at the beginning of the transmission season (strategy 1 and 2); RTS,S-like-duration vaccine or longer duration vaccine application alone is 3 rounds coinciding with the pattern of seasonal transmission with 1-2 years of 1 dose at the beginning of the season (strategy 3 and 4) or as 3 rounds before the pattern of seasonal transmission, with 1-2 years of 1 dose at the beginning of the season (strategy s3 and s4); and strategies combining MDA with RTS,S-like-duration vaccine or longer duration vaccine are a combination of all MDA and vaccine implementations combined together (strategies 5 to 8 and s5 to s8). Intervention coverage was assumed at 60%, with an initial yearly average PfPR 2-10 ≈ 3% -4% with peak PfPR 2-10 ≈ 10% -15% (corresponding to an initial EIR of 2 and effective access to care E 14 =45%). Simulations were chosen at random, full variation of predictions for each strategy are shown in Figure S3.
Prevalence in total population (proportion) Figure S2: Median and range of estimated yearly average all age prevalence following different intervention strategies. Plots (a-c) are for 2 years of mass intervention and plots (d-e) for 3 years of mass intervention. (a) and (d): estimated all age prevalence following mass vaccination with RTS,S-like-duration vaccine (purple) or mass vaccination with longer duration vaccine (pink), (b) and (e) estimated all age prevalence following full rounds of MDA alone (orange) or in combination with mass vaccination with RTS,S-like-duration vaccine (green) or mass vaccination with longer duration vaccine (blue), (c) and (f) estimated all age prevalence following reduced rounds of MDA alone (yellow) or in combination with mass vaccination with RTS,S-like-duration vaccine (brown) or mass vaccination with longer duration vaccine (light blue). Strategies where vaccination was performed before the transmission season are represented by dashed lines. Each intervention is represented by the median and minimum-maximum range across 10 simulations per a strategy. Intervention coverage was assumed at 60%, with initial average annual PfPR 2-10 ≈ 3% -4% with peak PfPR 2-10 ≈ 10% -15% (corresponding to an initial EIR=2 and effective access to care E 14 =45%).
ative impact of combined strategies for different coverage levels of each intervention. The x-axis indicates the coverage of MD mass vaccination. Colour represents the impact calculated as the relative maximum prevalence reduction of the combined interventi m prevalence reduction when using MDA alone (strategy 2) at the same coverage levels (relative impact =1 means that the co s the same impact, and a level of 2 means that mass vaccination with MDA is 2 times greater than MDA alone). A Represents th RTS,S-like-duration vaccine and B represents the impact of mass vaccination with longer duration vaccine. From left to write, results of reduced MDA rounds with mass vaccination for 2 years, the relative impact of reduced MDA rounds with mass vaccination for 3 A rounds with mass vaccination for 2 years and the relative impact of full MDA rounds with mass vaccination for 3 years. From top to , 2, 3 and 5.  Table   Time  deployment   EIR E 14 PfPR 2-10  Strategies  1  2  3  4  5  6  7  8  s3  s4  s5  s6  s7  s8  2years  0.5  45  <1  20  40  70  30  0  0  0  10  60  30  0  0  0  10  2years  1  45  2  80  90 100 80  10  30  0  10 100 70  10  40  10  30  2years  2  45  4  100 100 100 100 60  90  40  30 100 100 50  90  10      resurgence parameters for 2 years deployment of MDA, mass vaccination, or combination of both MDA and ma imported infections. The estimated parameters to individual regressions to each 10 simulations in each strategies are fidence intervals for the estimated half-life, λ 50 , representing the years after maximum prevalence was reached where pr evalence; the Hill's slope, representing the steepness of the logistic curve; and the 10% resurgence threshold, λ 10 , repres lence was reached where 10% of the resurgence occurred. The interventions are deployed during 2 years, initial preva R=5 and E 14 =25%). In the combined strategies, estimates when vaccination and drugs are given to the same proportion pecified with A , if not specified, a random coverage of the population is selected independently for each intervention. Vacci %, and in the combined strategies specified by B initial vaccine efficacy against infection is lower at 50%.  Figure S10: Relationship between entomological inoculation rate, EIR, and effective access to care, E 14 , with PfPR 2-10 and prevalence in all population. 5 different levels of EIR, from 0.5 to 5, are represented in the x-axis and 6 different levels of effective treatment E 14 = 15%; 25%; 45%; 60%; 70% 80%. The corresponding prevalence with given access and EIR levels is shown by the color gradient, the upper plots representing the maximum prevalence in the seasonal setting and the lower plots the average yearly prevalence. The plots on the left show PfPR 2-10 levels and the plots on the right the prevalence in all population. Results are across 10 simulations for each setting, the median value across the simulations is indicated by the color gradient, and the 95% range is indicated in brackets