BioPharma, Pharma

What no one is telling you about BMS’ Opdivo immunotherapy trial failure

After Bristol-Myers Squibb reported that its Opdivo lung cancer trial unexpectedly failed, perhaps a wholesale reset of cancer immunotherapy’s drug development, approval and payment is necessary.


The coverage that followed the Aug. 5 news of the failed trial of the much-heralded Opdivo (nivolumab) immuno-therapy drug from Bristol-Myers Squibb has focused on everything but the actual strategy that Big Pharma is employing to develop these novel drugs.

We’ve read about the “shock” that greeted this announcement as BMS stock plunged like an airplane in free fall. (After all, a trial failure is so rare in pharma and Opdivo was going to be such a sure shot in non-small cell lung cancer, given its near miraculous effects in treating advanced melanoma, that it was so surprising it failed, right?)

And then there were stories that looked at how Merck’s competing Keytruda drug that corners only a small share of the lung cancer market compared to Opdivo would surely benefit from its rival’s flailing fortunes. (Both Keytruda and Opdivo are cleared for second-line treatments in non-small cell lung cancer but doctors need to administer a companion diagnostic test for the former while Opdivo can be prescribed to any patient; hence the latter’s commanding market share.)

Still others concluded that BMS’ trial design was at fault — the study was too broad in the way patients were selected leading to a negative result, wrote Chris Scott, an analyst with JP Morgan in his research note to clients. Scott was forced to adjust his revenue model for BMS.

But in all the fast and furious coverage of the fallout, there appears to be no questions raised about whether standard drug development strategies should be used when immunotherapy — both novel and exciting — is concerned. One person who was bewildered at the response to the news of the trial failure believes there should be.

“What happened (last) week is very informative about the cost of hype,” declared Michael Kolodziej, an oncologist and former head of Aetna’s oncology program, at the Precision Medicine Leaders Summit in San Diego, Aug. 11.

Kolodziej is now national medical director, Managed Care Strategy at Flatiron Health, which backed by Google Ventures, Roche and others, has developed an oncology software platform.

In a phone interview after the conference, he expanded upon why current drug development strategies are not suitable to the development of immunotherapy drugs.

“What they’ve done is applied this very encouraging, very exciting, very novel therapeutic approach to a huge heterogeneous population of patients where only a small percentage really benefit, but there’s all this discussion about how this is a multibillion-dollar product for manufacturer XYZ,”Kolodziej said. “That’s hype, if you ask me.”

Currently only about 20% patients or thereabouts really respond to immunotherapy, even though oncologists are toying with the unheard term of “cure” to describe the response in some melanoma patients with advanced disease. This means that more effort needs to be expended in selecting the patients who can respond to immunotherapy drugs. But properly identifying which patient is best suited requires pharma companies to invest time and money in researching predictive biomarkers and developing companion diagnostic tests.

“I really do believe we will be able to identify the subset and that will result in a way to make sure that the right people get the treatment but if you look at the development strategy from the pharmaceutical companies with immuno-oncologic agents, they look just like the strategies for conventional chemotherapy drugs — basically every disease, every line of therapy,” Kolodziej said.

In fact even though having a biomarker as part of the drug development process provides a three-fold increase in the probability of ultimate FDA approval, only 20 of 159 cancer drugs today come with a predictive biomarker or companion diagnostic. That’s according to Nicholas Dracopoli, vice president and head of Oncology Biomarkers at Janssen Research & Development, part of Johnson & Johnson, who also spoke at the San Diego conference.

Recalling this meager datapoint, Kolodziej characterized the lack of predictive biomarkers in drug development as a problem. The fault, he believes, lies largely with Big Pharma’s reluctance to spend time, effort and money in identifying which patients are most suited for immunotherapy drugs.

“My personal opinion is that the most important thing is getting to market,” he charged of drugmakers. “Nothing else matters. Coming up with a biomarker limits their market size – why would they do that?”

A BMS spokeswoman didn’t address the speed-to-market or patient-selection issue but provided a statement. Here’s the relevant portion from Priyanka Shah:

BMS recognized the unique mechanism of action of immuno-oncology drugs early, and that the efficacy and side effect profiles of these agents would differ from traditional drugs such as chemotherapy.  In order to demonstrate the clinical value of immuno-oncology, we took a different approach to how we discover these medicines, how the efficacy of these agents is assessed, how the clinical trials are designed, how adverse events are managed and, ultimately, how we treat cancer patients.

It’s a historical time in cancer research, and we believe we have an unprecedented opportunity to transform cancer treatment across many tumors for patients waiting for new medicines, and will continue to take bold steps to help even more patients defeat cancer.

An ononcologist — Timothy Byun with the Center for Cancer Prevention and Treatment at St. Joseph Hospital in Orange, California — described Big Pharma’s race to develop meaningful immuno-oncology drugs as a “gold rush.” While agreeing with Kolodziej that the advances made in this therapy is truly exciting for patients and physicians alike, he said there is still not enough known about which patients respond best to immunotherapy.

Take for instance the biomarker probes that Merck’s Keytruda uses and BMS’ Opdivo trial used. Those probes are used to select patients based on how much their tumor cells express PD-L1.

“There’s three  or four PD-l1 immunohistocheimcal probes and they are not all the same. Merck has one probe and BMS used their own probe that was different than the Merck probe,” Byun said. “And it’s not comparing apples to apples. You don’t know if the 50% staining on the tumor cells correlates to a 50% staining using another probe. And to further complicate the matter, you don’t know whether you should be testing and looking for the PD-L1 staining in the cancer cell or the immune cells or both.”

In other words, there’s a lot we don’t know and perhaps it’s time to slow down. And yet Kolodziej is only too aware that it is not what patients battling cancer want to hear.

“You as the patient and you as the physician want to get access to that drug as quickly as possible. I get that,” he said.

So what gives?

A reset of the current regulatory and reimbursement where approval and payment of novel drugs can be conditional until evidence clearly shows benefit for a certain group of patients seems to be in order.

“We’ve developed accelerated approval, breakthrough designation all this stuff to get drugs to market faster but it’s relative uncommon for approval to be conditional and more uncommon for approval to be reversed based on evidence,” Kol0dziej said. “Could we see a universe in which conditional approval and payment based on encouraging clinical results but the requirement that there be some sort of date by which further evidence or better evidence regarding the population that would most likely benefit be mandated?”

He amplified on this theme arguing for a lower level of reimbursement for a provisionally approved drug but affording drugmakers the ability to increase prices once the evidence threshold is met. This is not unlike what happens in some European markets, Kolodziej pointed out.

“We need to find a way to solve for getting access to therapies as quickly as possible without granting basically carte blanche to a pharmaceutical manufacturer based on a registration trial where there is no biomarker,” he said.

Perhaps this approach may prove to be a win-win for patients, who have been disappointed by achieving no response to a therapy that seems to be marketed as a silver bullet in the battle against cancer — and it well may be when we identify the patient population. Not to mention for those folks whose wealth got wiped out after BMS’s stock plunge.

Photo Credit: Getty Images, Meriel Jane Waissman