As the class of agents targeting the PD-1/PD-L1 pathway expands in nonâ€“small cell lung cancer, so does the potential population of patients who would be candidates for the groundbreaking immunotherapy.
Renato G. Martins, MD, MPH
As the class of agents targeting the PD-1/PD-L1 pathway expands in non—small cell lung cancer (NSCLC), so does the potential population of patients who would be candidates for the groundbreaking immunotherapy, according to leading experts in the field.
Less than a year after the first agent gained the FDA’s approval in NSCLC, systemic immune checkpoint therapy has become the top choice for second-line therapy for patients with metastatic disease that does not harbor mutations for which there are molecularly targeted drugs.
In fact, patients with negative expression levels for the PD-1 ligand, PD-L1, demonstrate response rates to checkpoint inhibitors that compare favorably with the experience of patients who are treated with second-line chemotherapy without the attendant toxicities.
The optimal use of the approved and emerging checkpoint agents was a central theme of the 10th Annual New York Lung Cancer Symposium® that Physicians’ Education Resource (PER®) hosted November 7 in New York City.
“As of now, I don’t think that any patient is clearly not a candidate for these drugs,” said Renato G. Martins, MD, MPH, the medical director for thoracic/head and neck oncology at Seattle Cancer Care Alliance, in an interview. “Every patient with non—small cell lung cancer would potentially be a candidate in the second-line setting.”
Roman Perez-Soler, MD, who served as a program chair for the PER meeting, expressed excitement and confidence about the potential for the PD-1/PD-L1 pathway agents to go beyond shrinking tumors and deliver long-term survival for a subgroup of patients with NSCLC.
“The dream here—and what would make a big difference—is the possibility that a small group of patients with stage IV non—small cell lung cancer can actually be cured,” said Perez-Soler, who chairs the Department of Oncology at Montefiore Medical Center and is deputy director of the Albert Einstein Cancer Center in Bronx, New York.It has been 3½ years since the concept of targeting the PD-1 immune checkpoint pathway was introduced at the American Society of Clinical Oncology Annual Meeting in early research findings that involved nivolumab (Opdivo) in several advanced tumor types, including NSCLC.
Martins said those findings, along with research into EGFR mutations in NSCLC and a report on CT screening for high-risk patients that were released within days of each other, marked three of the most important days for lung cancer treatment.
In terms of immunotherapy advances, “the big breakthrough really was the recognition that non—small cell lung cancer could be a disease treated by immune therapy,” Martins said.
Since then, the PD-1 inhibitors nivolumab and pembrolizumab (Keytruda) have gained FDA approvals in melanoma and NSCLC. In late November, nivolumab also was approved for patients with advanced renal cell carcinoma.
Nivolumab was initially approved in March 2015 for the treatment of patients with metastatic squamous NSCLC with progression on or after platinum-based chemotherapy. In October, the FDA expanded the indication to include patients with progression on or after platinum-based chemotherapy with nonsquamous histologies.
Also in October, pembrolizumab was approved for patients with metastatic NSCLC with progressive disease after platinum-containing chemotherapy with a key difference.
A key difference between the approvals for the two drugs is that the label for pembrolizumab specifies that it is indicated for patients whose tumors express PD-L1 as measured by a companion diagnostic, which thus far is the PD-L1 IHC 22C3 pharmDx test. For nivolumab, there is a complementary test, the PD-L1 IHC 28-8 PharmDx assay, that is not required to use the drug.
Additionally, two agents targeting PD-L1, atezolizumab (MPDL3280A) and durvalumab (MEDI4736), are advancing in clinical development. The FDA has designated atezolizumab as a breakthrough therapy for the treatment of patients with PD-L1—positive NSCLC, meaning the agency will expedite its review. Durvalumab is being evaluated under a fast track designation for patients with NSCLC who have received 2 prior systemic treatments.
As far as Perez-Soler is concerned, clinical trial results thus far support the use of PD-1 inhibitors as the first choice for second-line therapy “based on efficacy across the board and lower toxicity.”
“Patients typically progress in two or three months,” Perez-Soler said in an interview. “The initial drugs, typically chemotherapy, shrink the tumor but then it comes back and the same drug cannot kill it.
“The most important thing now is to try and move these therapies to the front line because they are nontoxic and might be more effective,” he continued. “These therapies could also be used in combination with chemotherapy. That has been studied and so far the results look fairly promising.”
In October, the National Comprehensive Cancer Network updated its NSCLC guidelines to list nivolumab and pembrolizumab as preferred systemic therapies for patients with performance status scores of 0-2 who have progressed after receiving a chemotherapy doublet that could be paired with bevacizumab, if appropriate.1 Nivolumab was given a category 1 recommendation.Although the mantra in oncology is to find a biomarker that would identify the appropriate patient population for a given drug, the leading biomarker candidate for PD-1/PD-L1 inhibitors thus far has not provided a definitive answer for clinical use.
Objective response rates are notably higher among clinical trial participants with positive PD-L1 expression levels for the approved and emerging agents, but patients who test negative for the marker also respond to the therapies.
Martins analyzed the response rates among patients whose tumors were PD-L1—negative in 3 clinical trials involving different agents and found that between 8% and 9% responded to the immunotherapies.
“We have seen responses in patients who are ‘negative’ for PD-L1 and the response rate was still 8.9%, which compares with what chemotherapy would do in this setting—with the difference that even for those patients who are ‘negative,’ it seems that the quality of their response is as good as the quality of responses that we see in those who are PD-L1 positive,” said Martins.
Similarly, Perez-Soler noted that the response rates for patients with PD-L1—negative tumors are similar to those that have been seen when pemetrexed was compared with docetaxel as second-line therapy for nonsquamous NSCLC in 2004.2 “We thought this was a great study because pemetrexed was less toxic,” he remarked.
When it comes to predictive biomarkers in NSCLC, Perez-Soler said EGFR mutation testing has proved to be the “gold standard as it is a binary variable and allows for both inclusion and exclusion of patients to a specific therapy.” He said ALK FISH testing is less accurate because it requires a positive cutoff of 15%, making it prone to errors.
By contrast, Perez-Soler said, PD-L1 expression “is a continuum not a categorical variable” like the distinction between EGFR-mutant and wild-type tumors.
Additionally, he said the inherent limitations of the immunohistochemistry (IHC) methods used to measure PD-L1 expression on tumor samples make it difficult to accurately ascertain for slides stained at less than 50%, which would equate to an IHC3+ score.
“IHC is difficult to standardize,” he said. “Choosing a very high positive cutoff minimizes the concern.”
An IHC assay used in atezolizumab trials that measures PD-L1 expression on both tumor cells and tumor-infiltrating immune cells is “even trickier” to interpret, said Perez-Soler.Ultimately, PD-L1 expression should be analyzed as a predictive biomarker of long-term survival, said Perez-Soler. He said he will test all patients for PD-L1 expression level.
“Each patient is a unique scientific problem and obtaining reasonable biological information without risk and at a reasonable cost is justified,” Perez-Soler said. “PD-L1 IHC can be added to the IHC panel used at the time of initial diagnosis.”
However, many questions remain unanswered. Clinical trial evidence thus far presents a mixed picture of the marker’s clinical utility.For nivolumab, Perez-Soler noted that an analysis of responses in squamous NSCLC demonstrated that the median overall survival (OS) benefit was independent of PD-L1 expression levels stratified at greater and less than 1%, 5%, and 10%.3
In that trial, CheckMate-017, the median OS was 9.2 months with nivolumab versus 6.0 months with docetaxel, which translated into a 41% lower risk of death for those patients who received the immunotherapy agent.
The overall response rate (ORR) was 20% with nivolumab compared with 9% for docetaxel, with the median duration of response with nivolumab not yet reached after a follow-up of 11 months, Martins noted in his presentation.
In nonsquamous NSCLC, the median OS benefit from nivolumab was higher for patients increased with the level of PD-L1 expression: 17.7 months at ≥1%, 19.4 months at ≥5%, and 19.9 months at ≥10%.4
Overall, nivolumab demonstrated a median OS of 12.2 months compared with 9.4 months for docetaxel (P = .0015). The immunotherapy also resulted in a 19% ORR with a 17.2-month median duration of response, while docetaxel demonstrated a 12% ORR that lasted a median of 5.6 months.For pembrolizumab, an examination of the KEYNOTE-001 clinical trial outcomes by PD-L1 expression levels has indicated a markedly higher OS for participants with PD-L1 expression levels ≥50%.5
Survival data were presented at the 2015 AACR Annual Meeting for 356 patients in the total population whose PD-L1 levels were evaluable using an IHC assay. After a median follow-up of 10.9 months, OS was not yet reached in the high PD-L1 group (n = 119) and was 8.8 months in both the intermediate (n = 161) and low (n = 76) PD-L1 groups.
Based on these findings, Perez-Soler said a PD-L1 expression level >50% is a “reasonable cutoff” for pembrolizumab. However, he said, that may change based on the results of KEYNOTE-010, a phase II/III trial that compared 2 doses of pembrolizumab with docetaxel in 1034 patients with NSCLC who experienced disease progression after platinum-containing therapy and whose tumors expressed PD-L1.
In October, Merck announced that pembrolizumab was associated with longer OS among all patients, who needed a PD-L1 score ≥1% to enroll in the trial, as well as among the first group of patients with PD-L1 scores ≥50% whose outcomes were analyzed.6 Details have not been released yet.
The FDA’s approval for pembrolizumab in NSCLC, granted on an accelerated basis pending confirmatory clinical trial data on survival, indicates that the drug should be administered to patients with positive PD-L1 expression but does not specify a particular level.Perez-Soler noted a similar relationship between higher OS outcomes and PD-L1 expression levels with atezolizumab in the POPLAR trial, which compared atezolizumab with docetaxel.
In updated results reported at the 2015 European Cancer Congress in September, the activity of atezolizumab correlated with the extent of PD-L1 expression on tumor cells (TC) and tumor infiltrating cells (IC), with the highest activity seen in patients demonstrating expression on both cell types.7
Patients with the highest level of PD-L1 expression (TC/IC 3) achieved median OS with atezolizumab of 15.5 months compared with 11.1 months for docetaxel (HR, 0.49; P = .068). Across all levels of PD-L1 expression, the median OS was 12.6 months for atezolizumab versus 9.7 months for docetaxel (HR, 0.73; P = .040).
In patients without PD-L1 expression (TC/IC 0), a difference was not observed between the two groups.The PD-L1 inhibitor durvalumab is the subject of a broad clinical testing program in NSCLC, including several phase III trials as monotherapy and in combination with tremelimumab, an inhibitor of the CTLA-4 immune checkpoint.
The PD-L1 expression level for durvalumab is measured with the SP263 IHC assay, and patients with positive staining in ≥25% of tumor cells at any intensity are considered to have a positive score, said David R. Spigel, MD, of the Sarah Cannon Research Institute, during a presentation at the NY Lung symposium.
There was a difference in response rates based on PD-L1 levels in phase I testing, Spigel indicated. The ORR rate was 27% among 84 patients with positive PD-L1 expression, compared with 5% among the 92 participants considered PD-L1—negative.8
Survival data are not yet available.