EFFICACY OF MRNA VACCINES IN TROPICAL DISEASES: A QUANTIFIED EVIDENCE GAP, CLINICAL-TRIAL LANDSCAPE ANALYSIS AND HOST-RESPONSE
Keywords:
Mrna vaccines; tropical diseases; dengue; malaria; Zika; chikungunya; transcriptomics; clinicaltrials.gov; vaccine platforms; secondary data analysis; reproducibility Abbreviations: CHMI, controlled human malaria infection; CPM, counts per million; DENV, dengue virus; FDR, false discovery rate; GMT, geometric mean titre; ISG, interferon-stimulated gene; LNP, lipid nanoparticle; nab, neutralizing antibody; PRNT, plaque-reduction neutralization test; TPM, transcripts per million; WHO, World Health OrganizationAbstract
Background: Messenger RNA (mRNA) vaccines transformed prevention of COVID-19, yet no mRNA vaccine against any tropical infectious disease is licensed, and the availability of underlying public data has never been systematically quantified.
Methods: We performed a multi-dataset empirical secondary analysis. We quantified the public evidence base through logged NCBI GEO, NIAID Data Ecosystem and ClinicalTrials.gov queries; downloaded and re-analysed every vaccine-transcriptomic dataset available for dengue, malaria, Zika and chikungunya (GSE146658, TAK-003 live-attenuated dengue vaccine, n=20; GSE211626, PfSPZ malaria vaccine with sterile-protection outcome after controlled human malaria infection, n=15; GSE107672, RTS,S malaria vaccine pooled transcriptomics, 2,129 samples); and downloaded 477 registry records (469 unique vaccine trials) via the ClinicalTrials.gov API.
Results: No public dataset contains clinical efficacy or subject-level data for any mRNA vaccine against a tropical disease (GEO: 0 series). The registry contains exactly three mRNA-vaccine trials, all against Zika (NCT03014089, NCT04064905, NCT04917861), all phase 1-2 and completed with no posted efficacy results; one further trial (NCT03829384) used mRNA to encode a monoclonal antibody rather than a vaccine antigen. No mRNA-vaccine trial reached phase 3 (0 of 3) compared with 61 of 409 trials on other platforms, and across all platform classes advancement to phase 3 was significantly associated with platform class (chi-square 17.01, df 7, p=0.017); in a multivariable logistic model (n=410), enrolment size (OR 12.5 per log10, p=1.1e-13) and start year (OR 1.11 per year, p=0.0015) predicted phase-3 advancement. TAK-003 vaccination induced a transient whole-blood transcriptional response (2,446 up-regulated genes at day 2; 15 up/14 down at day 7, FDR<0.05) with a coherent interferon-stimulated gene module (mean paired log2 fold-change +1.26 at day 7, Mann-Whitney p<0.0001). In PfSPZ vaccinees, post-vaccination expression did not distinguish protected from unprotected subjects at single-gene FDR<0.05, but an exploratory regularised 20-gene model achieved leave-one-out AUC 1.0 (permutation p=0.005; wide Wilson intervals) and the MX2/GPR183 ratiometric signature moved directionally as reported for RTS,S (MoreProtected +0.81 vs LessProtected +0.35; not significant, n=15).
Conclusions: Public data currently permit immunogenicity-oriented and registry-oriented inference only - not efficacy estimation - for mRNA vaccines against tropical diseases. We provide a fully reproducible pipeline, a quantified evidence gap, and benchmark host-response signatures to guide future mRNA-vaccine development for dengue, malaria, Zika and chikungunya.
