Monoclonal antibody CIS43LS sets the bar for long-acting malaria protection

In this issue of the JCI, Tran and colleagues present a well-designed study and meticulous analysis of malariometrics, host responses, and malaria genetics on the foundation of pharmacokinetics and pharmacodynamics to yield a benchmark correlate of protection for the mAb CIS43LS (20). The pharmacokinetic and pharmacodynamic analysis utilized data from doses of 5, 10, and 40 mg/kg, which translates to 25, 50, and 200 mg for a 5 kg child and 300, 600, and 2,400 mg for adults. In a larger study of 110 individuals, there were 39 breakthrough infections at the 10 mg/kg dose and 20 at the 40 mg/kg dose. The pharmacokinetic modeling parameters were used to estimate a clinical correlate of protection from the two doses at 64 μg/mL (95% CI 49–93 μg/mL). This protection estimate is 1,000 times higher than the 50 ng/mL correlate that was approximated in vitro using hepatocyte invasion inhibition (IC₆₂). The pharmacokinetics align with almost 6 months of malaria protection.

Lowering to a simulated dose to 30 mg/kg, which would be achievable by a subcutaneous route of administration, afforded protection for up to four months. The authors also noted that presence or absence of concurrent malaria bloodstream parasitemia did not affect antibody levels or outcomes. The authors’ analyses also looked carefully for possible cross-reactive non-CSP malaria proteins expressed during erythrocyte stages and found that only five of the roughly 3,000 expressed P. falciparum proteins displayed minimal binding interaction. These cross-reactive proteins do not localize to the cell surface of merozoites or infected erythrocytes and were not thought to confer any clinical blood stage efficacy. Importantly, sequence data were available from all patients with breakthrough infections and indicated no new mutations in the conserved P. falciparum CIS43LS hinge epitope. Anti-drug antibodies to CIS43LS were rare, minimal, and transient; overall, these anti-drug antibodies were interpreted to be clinically nonexistent.

Estimating the antibody serum level required to prevent infections depends on the assay used. For example, using the Ghent-CEVAC ELISA, the RTS,S vaccine threshold of 121 ELISA units (EU)/mL is estimated to prevent 50% of infections (21, 22). An analysis of nine RTS,S vaccine trials with over 5,000 participants determined a protective level of 51 EU/mL (23) measured in a single laboratory (24). The ProC6C-AlOH/Matrix-M vaccine used a different assay (total IgG against full-length PfCSP) and found a 4.1 μg/mL protective level (25), but differences in assays and laboratories prohibit direct comparisons with the 121 or 51 EU/mL RTS,S values. Indeed, while the correlation of ELISA units among assays is generally good, standardizing is difficult even among anti-CSP ELISAs (26). Furthermore, comparison between vaccine-induced polyclonal antibody responses and protective levels of CIS43LS is inherently complicated by an “apples and oranges” comparison between vaccines and mAbs correlates of protection.