Scientists discovered that the drug, pevonedistat, does not kills cancer cells directly, as originally intended, but targets the proteins that the cells need to replicate and develop.
While it is already being tested in people to combat melanomas, researchers suggested it could also be effective in other cancers that have resisted treatment.
The critical protein blocked by pevonedistat is produced by the gene CDT2, and it is vital for malignant cancer cells to survive. Without it, the cells stop replicating and begin to fall apart.
Dr Tarek Abba, leader of the study at the University of Virginia School of Medicine, said: "We think that this (the protein) is what lets the cancer cells cope with the amount of replication they must undergo.
"We have great understanding now for how the drug works, and we think the drug works not because of its intended target, but rather because it works on (the CDT2 protein) way downstream of that."
As high levels of CDT2 are also found in many other tumours, such as those affecting the brain, breast and liver, researchers are hopeful that measuring these levels could help doctors gauge disease prognosis.
But as it is still being tested, pevonedistat is not yet available as a treatment.
"We have great hope that this drug will have very significant impact on melanoma in general," said Dr Abbas. "In fact, the drug is very effective on all melanomas, including those for which an effective therapeutic is lacking.
"We actually show this drug can work on melanoma that resisted treatment, which is a major challenge in melanoma therapy.
The findings are published in the scientific journal EBioMedicine.
