NEW YORK (GenomeWeb) – The US Food and Drug Administration on Friday approved the first lung cancer immunotherapy drug for a molecularly defined subset of patients.
Merck's anti-PD-1 drug Keytruda (pembrolizumab) is indicated for advanced non-small cell lung cancer patients who have progressed despite receiving platinum-containing chemotherapy or agents targeting EGFR or ALK tumor mutations, and whose tumors are positive for PD-L1 expression.
Alongside Keytruda, the agency approved an immunohistochemistry test called PD-L1 IHC 22C3 pharmDx, developed by Agilent Technologies subsidiary Dako to identify NSCLC patients whose tumors express high levels of the PD-L1 protein.
Keytruda blocks PD-1 receptors on activated T cells from interacting with PD-L1 and PD-L2 proteins on tumor cells. When this interaction occurs it hinders the body's immune system from attacking cancer cells. Immunotherapeutics like Keytruda attempt to inhibit this interaction, allowing T cells to attack cancer cells.
Studies have shown that when PD-L1 is highly expressed in certain tumor types, such as NSCLC, patients respond particularly well to anti-PD-1 agents. But these studies also showed that some patients with lower levels of PD-L1 expression also responded to treatment. As such, following the approval of Keytruda, some doctors may have a more flexible interpretation of drug labeling language defining PD-L1-positive patients.
Merck and Dako both announced they had filed their marketing applications for the drug and companion test in April. Two months later, Merck said that the FDA had accepted its supplementary biologics license application for the drug, and granted it priority review as well as breakthrough therapy designation. In granting accelerated approval for the drug on Oct. 2, the agency met its PDUFA date.
A Merck spokesperson told GenomeWeb that since Keytruda was approved under the accelerated pathway, the FDA accepted data on a subset of 61 patients with high PD-L1 expression. Dako retrospectively tested 220 patients using PD-L1 IHC 22C3 pharmDx to determine their PD-L1 status and identify the 61 patients with high PD-L1 expression. In this subset, 41 percent of patients saw their tumors shrink. Of the 25 patients who responded, 21 had ongoing responses at the final analysis cutoff of between 2.1 months and 9.2 months.
The agency considered the safety of the drug in a study involving 550 patients with advanced NSCLC. Common side effects associated with the drug included fatigue, decreased appetite, shortness of breath, difficulty breathing, and cough. The drug, due to its impact on the immune system, also caused adverse effects in the lungs, colon, and hormone-producing glands. Some uncommon immune-related side effects were rash and inflamed blood vessels. The agency noted that in clinical studies, some patients experienced Guillain-Barre Syndrome — a condition where the immune system attacks the nerves and can lead to paralysis.
The sensitivity of the test, the scoring guidelines, and the interpretation guidelines are developed specifically to Keytruda and patient outcomes to the drug.
In announcing the approval of Keytruda, the FDA highlighted it had also approved the "first test designed to detect PD-L1 expression in non-small cell lung tumors." As soon as the approval came, Merck was ready to go to market and announced that three large US reference labs — LabCorp, Quest Diagnostics, and GE Healthcare's Clarient Diagnostics Services — would offer the CDx.
When a therapy has priority review, as Keytruda did, it also speeds up the review timeline for the companion test, which can be challenging if the sponsors are aiming for simultaneous launch of the drug and test. Henrik Winther, VP and general manager of companion diagnostics at Agilent, acknowledged that it took a lot of people under lots of pressure to successfully get the CDx through the FDA, but in his view the process "hasn't been that difficult" from a timeline perspective. He described the regulatory experience as a classic CDx submission where the company made modular premarket approval submissions to the agency covering the analytical, clinical, and manufacturing aspects of the test.
However, Dako's CDx is entering the market while the regulation of lab-developed tests is in flux. Currently, labs can launch LDTs through CLIA-certified labs that gauge the same analytes as FDA-approved companion tests. Companies that have taken the time to take their tests through the FDA object to what they view as an "uneven playing field." The agency has cited these complaints as one reason it wants to begin regulating LDTs and has issued a draft plan for doing so.
While Winther expects that there might be some labs that offer PD-L1 expression testing as LDTs, he doesn't expect it to be as big a problem as it has in the past. From patient safety perspective, he emphasized that Dako's PD-L1 CDx has been designed to precisely fit the drug.
"The sensitivity of the test, the scoring guidelines, and the interpretation guidelines are developed specifically to Keytruda and patient outcomes to the drug," he said. "We really designed the test very specifically, and we submitted data that demonstrated that the test will be safe and effective to use with [certain] sensitivity scoring and interpretation guidelines."
In fact, there are a number of pharmaceutical companies developing immunotherapeutics for NSCLC with PD-L1 companion tests that have unique characteristics, including different cutoffs for defining PD-L1-positive and -negative patients. "Other drugs have other mechanisms of working and even if it is the same indication, they will have different outcome data," Winther said. "Therefore, you have to design that companion diagnostic specifically to that drug."
According to the FDA, when it drew up its guidance of CDx development, the agency didn’t anticipate there would be multiple drugs in the same class, each with its own companion assay. In such a situation it becomes costly for hospitals to adopt all the different FDA-cleared companion tests and confusing for doctors to figure out which test to order.
To address these challenges, the agency held a meeting in March, during which four drugmakers developing immunotherapies — AstraZeneca, Roche/Genentech, Bristol-Myers Squibb, and Merck — said they would compare the IHC tests they were using to stratify patients in drug trials and characterize the analytical variability between the platforms.
If covered, I would definitely prescribe the drug to patients with lower levels of expression as there was definite benefit to some patients in the clinical trials with low or no expression.
Defining likely responders
Data submitted to the FDA on Keytruda suggest that 22 percent of NSCLC patients had PD-L1 expression in 50 percent or more tumor cells. However, a difficulty for physicians on the ground will be figuring out whether to prescribe Keytruda to NSCLC patients when their tumors express PD-L1 in less than 50 percent of tumor cells.
The label for the PD-L1 IHC 22C3 pharmDx test states that if 50 percent or more of tumor cells are stained for PD-L1 expression, then the patient's sample should be considered PD-L1 positive. The indication section of Keytruda's label doesn't specify the cutoff for determining when a patient's tumor is PD-L1 positive, but the clinical studies section of the label details that the data submitted to the agency was from patients who had PD-L1 expression in half or more tumor cells.
As a predictive marker, PD-L1 expression isn't a binary determinant of response in the same way that EGFR mutation status is in NSCLC. Earlier this year, a study presented at the American Association for Cancer Research annual meeting and published in the New England Journal of Medicine reported that around 20 percent of the overall study population had a response to Keytruda. But when researchers evaluated patients based on PD-L1 expression in tumors, it revealed a best responder population, but also different degrees of benefit.
There was a 45 percent overall response rate in patients who had PD-L1 expression in half or more tumor cells. In patients with 1 percent to 49 percent of PD-L1-positive tumor cells (the intermediate group) the response rate was 17 percent, while for patients with less than 1 percent PD-L1-positive tumor cells the response rate was 11 percent. Because patients with lower levels of PD-L1 expression still responded to the drug, experts at the meeting predicted that establishing a cut off for determining best responders would be tricky.
"Certainly, that situation is more clear cut with the kinase inhibitor class of drugs where the biology tells us if the drug is really not likely to work if the patient doesn't have a particular mutation," Suzanne Topalian of Johns Hopkins Medicine said during a presentation of the above data at the AACR meeting. "Here … we heard about a potential biomarker to select patients with lung cancer for anti-PD-1 therapy, but this kind of biomarker is a lot more blurry. And you saw that even patients with lower levels of expression of this marker still had notable response rates."
Ultimately, how payors interpret the drug and test label and the available evidence on PD-L1 as a predictive marker might guide physicians. "We are all anxiously waiting to see how strict the payors will be regarding the biomarker," Lecia Sequist, a lung cancer expert at Massachusetts General Hospital, told GenomeWeb. "If covered, I would definitely prescribe the drug to patients with lower levels of expression as there was definite benefit to some patients in the clinical trials with low or no expression."
One of Sequist's best responding patients that she treated as part of early trials of Keytruda was negative for PD-L1 expression. "We still have a lot to learn about how best to select patients for immunotherapy," she said.
Since Keytruda was approved under the agency’s accelerated approval program, Merck will have to submit data from additional trials to garner full approval for the drug. "An improvement in survival or disease-related symptoms in patients being treated with Keytruda has not yet been established," the FDA said in a statement. The Merck spokesperson noted that the company is conducting a confirmatory study, called KEYNOTE-010, to demonstrate an improvement in survival or disease-related symptoms. The results from this study are expected in November.
Last year, the FDA approved Keytruda for advanced melanoma patients who have stopped responding to other drugs. In the lung cancer space, the FDA in March approved Bristol-Myers Squibb's immunotherapy Opdivo (nivolumab) for advanced squamous NSCLC patients who have progressed on platinum-based chemo. But Opdivo is not approved in a molecularly defined patient subset like Keytruda is.
Based on published reports, Keytruda and Opdivo both cost around $150,000 for a yearly course of treatment. The Merck spokesperson quoted the monthly price of Keytruda at $12,500, but said that the cost to patients will differ based on length of treatment, as well their insurance policies. Merck has assistance programs to help defray the cost for patients who can't afford the drug.
Winther noted that its reimbursement for the PD-L1 CDx will "be in the typical range of a companion diagnostic assay," but would not provide a specific amount. According to reimbursement experts, there is a single CPT code for IHC tests, which is priced at around $100.