Yes.
Malware running in a VM can exploit bugs in the hypervisor (which runs as a privileged process on the host) to infect the host or other VMs.
See
https://www.cnet.com/news/crisis-malware-targets-vmware-virtual-machines/#! for just one example.
Crisis was not actually able to infect the host machine from within the VM. It was the other way around. It could infect a VM,
after it had already infected the host machine.
In order for malware to jump from the VM to the host, one of two things would have to happen.
One is that the user would have to have set up a very insecure working environment... for example, allowing the system on the VM free write access to the entire user folder on the host Mac. Doing that could conceivably result in an infection, if there were malware capable of running on the virtualized system, detecting this situation and dropping a Mac payload into the right places in the user folder. To my knowledge, such malware does not exist.
The other is that there would have to be a serious bug in the VM to allow the virtualized system free access to the host system, plus the same requirement for theoretical malware capable of running in the VM and dropping a Mac payload through that bug. To my knowledge, neither of these situations exist.
Could one of those things happen? Sure. Theoretically.
Has it happened? No. As someone who has actually used VMs to run just about every piece of Mac malware from the last few years, I can say that with as much certainty as anyone can.