Just FYI, I wouldn't put too much faith in safety profiles of related family drugs. The classic med chem example of really unpredictable toxicology is mevastatin vs lovastatin, which have exactly the same structure but for one lowly methyl group. The former molecule was basically an unusably toxic drug candidate that never would have been approved by the FDA, whereas the latter was the first safe and effective statin to go to market.
it's a known mechanistic issue. the pyrrolobenzodiazepine molecules are pretty selective DNA alkylators. Cardiotox is caused probably by transport chain involvement when the 9-position is oxygenated (makes it look like Co-Q). Without that oxygen, the PBDs zip on over to the DNA. PBDs have a slight twist to the molecule which allows them to nestle in the minor groove with exactly the right pitch so that the lesion evades base excision repair response, which is what is thought to cause the more serious side effects in DNA-alkylating chemotherapies.
However, you are right, there could be unforseen toxicity issues.
