It seems rare these days where people's PhD field is directly related to people's field after residency. So in that sense this concern is moot.
OTOH, compared to translational/applied research, basic science seems to move really really slowly. I frankly don't think you'll have a problem catching up if you take a break for 5 years in terms of the content of the science--I read papers posted on Twitter in basic science journals and they look very similar to papers that were written 10 years ago. As to whether what you do generate grants, seems that the part that's "impact" is all about how the research is applied to clinically relevant questions (so hopefully your MD training will be more applicable), and the part about "innovation" is mainly reviewed by people who are unfamiliar with the field at the level of detail to adjudicate the actual innovation. With basic science, you get grants by working with a famous mentor. Your concerns are not really relevant.
The real concern to drop out is when you are in your mid 30s and as a mostly clinician you'll make 3-5x $$$ vs. your basic science colleague. I'd estimate at least 50-80% of drop out at any of the transition points are due to this. IMO once you get to postdoc stage, if you REALLY want a job doing 50%+ research you'll be able to get it--you'll essentially just do it for free until you get grants (the school pays you a low salary that gets covered by clinical revenue). This effect is amplifying as grants are not inflation adjusted.