Test for multiplicative effect in mendelian randomization?

Question

[clairvoyant]

A study reported that exposure A and B had multiplicative effect on the outcome C. Let's say I want to prove in MR that A*B had multiplicative effect on outcome C. I already knew that MVMR (multivariate MR) could investigate the causal effect of related exposures and avoid the weak IV bias. But little information could be found about how to determine whether additive effect or multiplicative effect in MR. Could anyone tell me how to do this? Any help would be greatly appreciated.