Most cancer drugs that enter clinical trials never make it past regulatory review. Why? It might be because they do not actually hit their intended molecular targets, scientists at Cold Spring Harbor Laboratory (CSHL) argued in a new study.
After examining the mechanisms of 10 cancer drugs that were thought to target one of six proteins, Jason Sheltzer and colleagues at CSHL made the unexpected discovery that the drugs may indeed kill tumor cells, but not by blocking their intended targets. The medicines they scrutinized came from companies that included Celgene and Pfizer.
The findings, published in the journal Science Translational Medicine, could help explain why seemingly effective treatments never make it out of the clinic. With that knowledge, scientists can improve drug discovery and increase the success rate of personalized medicine, the researchers believe.
The 10 small-molecule drug candidates are designed to target CASP3, HDAC6, MAPK14, PAK4, PBK and PIM1. They include the HDAC6 inhibitors citarinostat and ricolinostat, which Celgene received via its acquisition of Acetylon Pharmaceuticals in 2016 and is now testing in multiple myeloma. The study also included the p38 MAPK inhibitor ralimetinib, which Eli Lilly put up for sale mid-2017 amid a shake-up of R&D priorities. Then there was Pfizer’s PAK4 inhibitor PF-03758309, for which a phase 1 in advanced solid tumors was terminated early due to “undesirable PK characters … and the lack of an observed dose-response relationship.” All told, the drugs the CSHL researchers examined have been used in at least 29 different clinical trials involving more than 1,000 patients.
Five of the six proteins these drugs target are reported to be key for cancer cell’s survival. (The exception is CASP3, which is considered to induce cell death when activated.) But when the CSHL researchers used the CRISPR gene-editing tool to remove the proteins, they found the cancer cells continued to grow normally in lab dishes.
More surprisingly, the drugs killed cancer cells that did not express their protein targets, the team reported.
This shows the drugs “must necessarily be killing the cancer cells through some other unknown off-target effect,” said Jason Sheltzer, Ph.D., the study’s senior author, during a press briefing Tuesday.
“Our results kind of suggest that … you can have drugs that have a good efficacy against a particular protein in an isolated test tube system, but it could be doing something very different in a cancer cell, or it could be inhibiting that protein in the cancer cell and just that inhibition has no efficacy,” he said.
So what exactly causes this misunderstanding of how drugs really work to kill cancer cells? The researchers suspect the tools scientists used to block the production of specific proteins could be to blame.
The method, called RNA interference (RNAi), allows researchers to inhibit gene expression. In fact, RNAi has been used in such therapeutics as Alnylam Pharmaceuticals’ Onpattro to treat polyneuropathy in hATTR amyloidosis. But there’s a chance that in cancer research, the process could somehow interfere with the production of other proteins. CRISPR, on the other hand, produces fewer unintended “knockdowns” than RNAi does, Sheltzer said.
Then there’s another question: Does knowing the exact target really matter, as long as the drugs can still kill cancer cells? Yes, the researchers believe.
“The issue here is doing whatever we can to increase the probability that a drug will be successful,” Chris Giuliano, a co-first author of the study, said during Tuesday’s conference call. Using those targets to find the patients who are most likely to benefit from a therapy is the whole point of precision medicine.
RELATED: Treating resistant prostate cancer by depriving a key hormone of its ‘co-pilot’
Sheltzer and his team picked the drugs for the current study based on the lack of reported gene mutations that could render the cancer cells resistant to them. “The identification of a mutation that grants resistance to a particular agent is really the gold standard for proving an on-target effect,” Sheltzer explained. So, finding these mutations can help clinicians figure out what a drug is really doing, he argued.
That’s what the researchers did for a putative PBK inhibitor dubbed OTS964 by OncoTherapy Science. They had already showed PBK was not required for cancer cell growth. To find out how exactly OTS964 works, they exposed colorectal cancer cells to a lethal concentration of the drug and isolated those that somehow survived. The team found that the cells developed resistance by mutating the gene that produces the protein CDK11. And it happens to be the first drug to target CDK11 ever discovered, according to Sheltzer.
Cyclin-dependent kinases (CDKs) are a large protein family that’s implicated in cancer, with several CDK4/6 inhibitors having already been used in breast cancer. But CDK11 is still understudied. Sheltzer and colleagues are now investigating whether inhibiting CDK11 could work in any specific cancer types so they can run better clinical trials.
Moreover, they intend to figure out the mechanisms of action of the other anti-cancer drugs. “If we are able to successfully do that, then we might find new vulnerabilities in cancer cells that we can target or new ways to identify the patients who are most likely to respond to a particular therapy,” Sheltzer said.