When skin cancer has spread to other organs, even the most promising class of drugs fails in half of all patients.

But in a new University of Pennsylvania study, authors report they can pinpoint why the drugs fail with a simple blood test, and they can do it just six weeks after starting the therapy – allowing for rapid deployment of a second kind of treatment.

The study, published Monday in the journal Nature, is the first successful collaboration to emerge from the Parker Institute, a cancer research consortium established in April 2016 with $250 million from the foundation of tech-industry giant Sean Parker.

Researchers at Penn's Perelman School of Medicine pooled their results with similar data from Memorial Sloan Kettering Cancer Center, thereby accelerating the hunt for answers, said the study's senior author, E. John Wherry, a professor of microbiology at Penn.

"We're taking advantage of the collaborative environment that Sean Parker set up," said Wherry, who co-led the study with Penn physician Tara C. Gangadhar. "We just called our friends at Sloan Kettering."

Penn and Sloan Kettering, in New York City, are among six academic medical centers that have received funds from the institute. The new study also had funds from the government and other sources, but the Parker connection enabled the collaboration, Wherry said.

The drugs analyzed in the study are called checkpoint inhibitors, which include Keytruda, made by Merck, and Opdivo, from Bristol-Myers Squibb. Briefly, they work by taking the brakes off a pathway in the immune system, ideally allowing the patient's body to attack the cancer with weapons called T cells.

The patients in the study had advanced melanoma. Among those for whom the drugs did not work, researchers found that with the blood test, they could identify some cases in which the drug prompted a partial response, and others for whom the drug did not work at all.

The data suggest that patients in the first category would likely be candidates for a second type of checkpoint inhibitor, whereas those in the second category would need another option, Wherry said.

Currently, patients generally wait 12 weeks after starting checkpoint inhibitors before getting a scan to see if the drug is working. But time is of the essence with the aggressive tumors, so the blood test at six weeks could help guide treatment sooner, Wherry said.

In patients for whom the drugs elicit a partial response, a fast determination of that fact would allow physicians to attack the cancer with a second checkpoint inhibitor before the disease recovers, he said.

Patients whose cancers do not respond may be candidates for a different approach. Various research labs are exploring ways to remove a patient's T cells, activating or re-engineering their cancer-fighting ability, then putting them back in the patient.