Clinical trials of novel cancer drugs miss their marks far more often than they hit them, in part because the drugs themselves may be aimed at the wrong targets or the targets themselves may not be that important in the first place, investigators have found.
Using CRISPR (clustered regularly interspaced palindromic repeats) gene editing to study the effects of 10 drugs that are in development targeting six proteins ostensibly crucial to the health or survival of cancer cells, Jason Sheltzer, PhD, of Cold Spring Harbor (N.Y.) Laboratory and colleagues found that the cancer cells could get along just fine without the targeted proteins, suggesting that the drugs’ alleged efficacy in the lab dish was because of other, off-target effects.
“It seemed like these genes that encode proteins that are being targeted by putative precision agents in clinical trials aren’t actually essential at all for cancer cell growth, so that was one surprise that we found,” Dr. Sheltzer said in a telephone briefing for reporters held prior to publication of the study in Science Translational Medicine.
“The second surprise was that we took the drugs that were supposed to be specific for these proteins and then we treated cancer cells with them, and we found that the drugs continued to kill the cancer cells that totally lacked the target protein expression,” he said.
But the investigators also made a serendipitous discovery that one of the drugs they tested, OTS964, was not – as originally thought – an inhibitor of the PBK protein but instead was an inhibitor of cyclin-dependent kinase (CDK) 11, making it a molecular relative of drugs such as the CDK4/6 inhibitors ribociclib (Kisqali) and palbociclib (Ibrance), both potent inhibitors of hormone receptor–positive, HER2-negative metastatic breast cancer.
Drug development insights
Their findings also indicate that less-precise candidate-drug identification techniques using RNA interference (RNAi) to knock down protein expression may have led earlier investigators down the garden path, resulting in errors that can lead to the all-too-familiar scenario of a seemingly promising compound flourishing in the early drug development process, only to wither on the vine in clinical trials.
“Everyone knows that it’s really hard to make new cancer drugs, but what we’ve been finding out is that, once a new drug is even made, it can be just as difficult to really understand how that drug is working to kill cancer cells,” said coauthor Chris Giuliano, currently a doctoral candidate at the Massachusetts Institute of Technology in Cambridge, who also spoke at the briefing.
“Our study showed us that a potential problem with the cancer drug development pipeline is that the way in which some of these new cancer drugs work is incompletely understood. Ten years ago many of these studies were developed with a tool known as RNAi, which while being the best available tool at the time ultimately led many researchers to arrive at the wrong conclusions about how some of these drug targets actually work,” he added.
Sour on MELK
The investigators had previously found that MELK (maternal embryonic leucine zipper kinase), a protein previously identified as essential to survival in multiple cancer types, could be eliminated from cancer cells using CRISPR gene editing without significant harm to the cells, and that a drug targeted against MELK in phase 2 clinical trials (OTS167) continued to kill the knockout cells in a lab dish with no loss of potency. This finding alerted the researchers to the possibility that drugs in development could be targeting the wrong protein, accounting for at least some of the high failure rate in cancer drug development, and potentially explaining some of the toxicities seen with experimental agents.
“Moreover, clinical trials that use a biomarker to select patients for trial inclusion are about twice as likely to succeed as those without one. Misidentifying a drug’s mechanism of action could hamper efforts to uncover a biomarker capable of predicting therapeutic responses, further decreasing the success rate of clinical trials,” they wrote.
Other false targets
In the current study, the investigators tested whether other drugs were designed to point toward nonessential or “superfluous” targets, and whether the mechanisms of action of the drugs had been mischaracterized.
They focused on five proteins that were thought to be so important to cancer cells that their loss would inhibit or block cell proliferation (HDAC6, MAPK14/p38, PAK4, PBK, and PIM1) and one (CASP3/caspase-3) that was thought to induce apoptosis when activated by a small molecule.
First, the investigators determined that the putative targets – the five proteins listed before – may not be required for actual cancer cell growth or survival, and then found evidence to suggest that misidentification of the proteins may have been caused by the uncertainties of RNAi.
They then used CRISPR to assess the mechanism of action of each of the 10 drugs, and whether the effects they induced were on or off target. They found that PAC-1, a putative caspase-3 activator currently in three clinical trials, actually works in a caspase-3–independent manner, and that all 10 anticancer drugs “exhibited clear evidence of target-independent cell killing in every [knockout] cell line that we examined.”
Finally, as noted before, they determined that the actual mechanism of action of OTS964 was not PBK inhibition, but inhibition of CDK11, a protein that appears to be vital for mitosis in human cancers.
“We think that CDK11 is an exciting target for future therapeutic development, and we found it specifically by looking for the true targets of these mischaracterized agents,” Dr. Sheltzer said.
The investigators acknowledged that their study was limited by the use of well-established cancer cell lines that may not fully reflect how cancer acts in the human body, and they could not rule out that the superfluous proteins they identified might be important for the survival of rare cancers.
“Additionally, we’re not saying that these targets offer no therapeutic potential,” said lead author Ann Lin, who is currently a Fulbright Fellow at the University of Oslo.
“It might be that there are other, unrelated proteins in the cell taking over its role and targeting of both proteins in combination is needed to kill the cancer cells. Furthermore, removal of these proteins may reveal a weakness in the cancer that can be targeted by a second drug, so our experiments showed that uniquely targeting these proteins alone showed little efficacy,” she said.
Research in the Sheltzer Lab is supported by an National Institutes of Health Early Independence Award, a Breast Cancer Alliance Young Investigator Award, a Damon Runyon-Rachleff Innovation Award, a Gates Foundation Innovative Technology Solutions grant, and a CSHL-Northwell Translational Cancer Research grant. The authors reported that they have no competing interests.
SOURCE: Lin A et al. Sci Transl Med. 2019 Sep 11. doi: 10.1126/scitranslmed.aaw8412 .