For clinicians, a central message from 2015 ASCO is that a framework for managing immune-related adverse events in patients who receive these immune checkpoint agents is taking shape.
Jedd D. Wolchok, MD, PhD
Three years ago, Jedd D. Wolchok, MD, PhD, sat in the lobby of a Chicago hotel at midnight during the ASCO annual meeting with several of his colleagues and sketched out on a napkin a “dream trial” that would simultaneously seek answers to key questions about immune checkpoint therapies just then moving toward center stage in oncology drug development.
This year, Wolchok was back at the 2015 ASCO Annual Meeting with the results from that study—a three-arm trial that compared the PD-1 inhibitor nivolumab (Opdivo) as monotherapy and in combination with the anti-CTLA-4 antibody ipilimumab (Yervoy) versus ipilimumab alone in patients with advanced melanoma.1
The fact that emerging agents employing novel therapeutic strategies could vault so quickly from promising concepts to phase III data involving 945 patients treated at 137 sites in the United States and approximately 12 other countries seems to amaze even Wolchok.
“It’s a very strong indication of how the field is accelerating and I think it demonstrates a very, very firm commitment on the part of investigators and the sponsor as well to find the best treatments as quickly as possible,” he said in an interview at this year’s conference. “This was a truly outstanding accomplishment.”
Indeed, the CheckMate-067 trial on which Wolchok reported emerged as one of the most significant pieces of research presented at ASCO this year—and not simply in terms of efficacy data. The trial findings are part of a growing body of evidence about how best to administer checkpoint-modulating therapies to patients with melanoma and other tumor types in which the therapies likely will be increasingly employed.
Experts say agents that inhibit PD-1/PD-L1 activity are poised to become a backbone of anticancer therapy as important as chemotherapy and, in some malignancies, may someday be a replacement for it.
For clinicians, a central message from 2015 ASCO is that a framework for managing immune- related adverse (iAEs) in patients who receive these immune checkpoint agents is taking shape.
“This is something we’re going to all need to learn how to do and we’re going to all need to embrace because it’s really not just for melanoma patients and not just for ipilimumab anymore,” said Michael A. Postow, MD, a colleague of Wolchok’s at Memorial Sloan Kettering Cancer Center in New York City. “These drugs will likely be FDA approved in many different indications and, even though the side effects are different and the use of the agents may be a little bit different, it’s something that everyone’s going to need to be familiar with.”
The PD-1 Landscape
As research unfolds, inhibiting PD-1/PD-L1 activity as an anticancer strategy continues to expand into solid tumor types previously considered unlikely candidates for such an approach. That was among the themes of 2015 ASCO, according to Lynn M. Schuchter, MD, FASCO, a melanoma specialist who served as an ASCO expert commentator at the meeting.
“There were results presented in using PD1 antibody in lung cancer, in head and neck cancer, and in liver cancer, with encouraging results,” Schuchter, chief of Hematology Oncology at Penn Medicine in Philadelphia, said in an interview. “I would say that one theme was the surprise that there is going to be broader applicability of the PD1 antibodies and immunotherapy in more cancers than we had anticipated.”
The Society for Immunotherapy of Cancer (SITC) concentrated on nivolumab and the PD-1 inhibitor pembrolizumab (Keytruda) in developing its list of immunotherapy highlights from this year’s conference (Table 1). The list illustrates the list of tumor types in which PD-1 inhibitors are under study as well as the search for biomarkers that might help define the most appropriate patient population for the drugs.
CRC indicates colorectal cancer; DCR, disease control rate; HCC, hepatocellular carcinoma; KLH, keyhole limpet hemocyanin; MMR, mismatch repair; mOS, median overall survival; mPFS, median progression-free survival; NR, not reached; NSCLC, non—small cell lung cancer; ORR, objective response rate; OS, overall survival; PD-L1, programmed death ligand-1; PFS, progression-free survival; RCC, renal cell carcinoma.
One notable exception on the SITC list is rindopepimut (CDX-110), a vaccine designed to stimulate an immune response against the epidermal growth factor receptor (EGFR) variant (EGFRvIII) in glioblastoma. In February, the FDA granted a breakthrough designation for rindopepimut in adults with glioblastoma based in part on the phase II ReACT trial results, according to Celldex Therapeutics, which is developing the drug.2 Updated results were presented at the meeting.3
Nevertheless, PD-1/PD-L1 inhibitors continue to constitute the most rapidly developing class of immunotherapy agents, as evidenced by the speed with which nivolumab and pembrolizumab have been moving rapidly through the regulatory process. The FDA approved nivolumab as a treatment for patients with unresectable or metastatic melanoma in December 2014 and added an indication in metastatic squamous non—small cell lung cancer (NSCLC) less than 3 months later. Now, Bristol-Myers Squibb, which is developing both nivolumab and ipilimumab, is seeking the FDA’s approval for the two drugs to be used in combination for patients with previously untreated advanced melanoma based on clinical trial results from CheckMate-069, which demonstrated an objective response rate of 61%, including a 22% complete response rate.4 The company anticipates a decision by September 30.
Meanwhile, Merck is expecting the FDA to decide by October 2 on its application for approval of pembrolizumab in progressive, advanced NSCLC based in part upon the KEYNOTE-001 trial, in which the response rate was 19.4% overall and 45.2% among patients with high PD-L1 expression levels (≥50% tumor cells).5
The FDA initially approved pembrolizumab in September 2014 as a treatment for patients with unresectable or metastatic melanoma and disease progression following other therapies. Merck also has an application pending for the use of pembrolizumab in ipilimumab-refractory advanced melanoma.
Although the clinical trial data are not as advanced as they are for nivolumab, pembrolizumab also is showing broad efficacy across multiple tumor types. Merck said pembrolizumab has now demonstrated antitumor activity in 13 different malignancies.6
Managing Adverse Events
Amid such wide-ranging potential for checkpoint inhibition therapy, oncology leaders are focusing on bringing oncologists up to speed on how best to translate these new drugs to clinical practice.
During an education session at 2015 ASCO,7 Postow delved into three central questions in managing AEs experienced by patients receiving checkpoint inhibitors: (1) What are the side effects?; (2) When do they happen?; and (3) What do you do if they happen?
Postow said in an interview that the AEs could be grouped into four major buckets: dermatologic, gastrointestinal, endocrine, and neurologic. Inflammatory hepatitis also may be concerning, although most instances of hepatic iAEs in the CheckMate-067 trial were laboratory values as opposed to side effects with clinical impact, Postow said.
In fact, the CheckMate-067 trial data offer a summary of the most frequent iAEs associated with the nivolumab and ipilimumab.1 As expected, fewer patients receiving nivolumab monotherapy experienced iAEs of any grade than did participants who received ipilimumab alone while those who took both drugs had a markedly higher incidence of iAEs (Table 2). The most frequently reported iAEs fell into three of the four buckets Postow mentioned: skin problems including pruritus and rash, gastrointestinal iAEs including diarrhea, and the endocrine iAE of hypothyroidism.
The majority of grade 3-4 iAEs in these categories were resolved with the exception of endocrine events, Wolchok noted in his presentation. Specifically, the resolution rates were 85%-100% for the immunotherapy combination, 50%- 100% for nivolumab monotherapy, and 83%-100% for ipilimumab alone.
In the neurologic category, iAEs are a rarity, said Postow. “Some people have had neurologic symptoms like aseptic meningitis, which can [lead to] inflammation of the meninges,” he said.
Symptoms including posterior reversible encephalopathy syndrome, aseptic meningitis, enteric neuropathy, and transverse myelitis have surfaced among patients who have received ipilimumab, Postow noted in an article he wrote for the 2015 ASCO Educational Book.8
Drawing on accumulated research, Postow indicated in his presentation that most grade 3-4 iAEs occur at varying times depending on the symptom and the regimen.7 For instance, the median time to onset of grade 3/4 problems with the combination of nivolumab and ipilimumab was approximately 2 weeks for skin problems, compared with 6.9 weeks for gastrointestinal AEs and 9.4 weeks for endocrine AEs. For nivolumab alone, the median time to onset of any grade of treatment-related AE was 5 weeks for skin problems, 7.3 weeks for gastrointestinal difficulties, and 10.4 weeks for endocrine AEs.
Regardless of when such events occur, strategies for managing these AEs have been developed. In his presentation, Postow outlined approaches to some of the more common AEs, including these recommendations:
Rash—Use topical corticosteroid cream, Benadryl or hydroxyzine for pruritus, and oral steroids.
Diarrhea/colitis—Remedies depend on baseline stool frequency; <4X, Imodium or budesonide; <7X, prednisone; >7X, hospitalize with IV Solu- Medrol, consider colonoscopy and CT scan, or consider infliximab.
Endocrinopathy—Replace the missing hormones with levothyroxine or hydrocortisone therapy and be aware of adrenal crisis. Use of higher doses of steroids during acute hypohysitis to prevent longterm pituitary dysfunction remains controversial.
For managing patients who experience iAEs on ipilimumab therapy, Postow noted that the readily available Risk Evaluation and Mitigation Strategy guidelines approved along with the drug offer detailed advice for clinicians.
Overall, Postow said, the side effects of anti-PD-1 therapy are less frequent and manageable. “The rate of significant side effects is much greater with chemotherapy than PD1 agents,” Postow noted. “So, these are not necessarily worse side effects, but just different. And in many contexts, they are less than with chemotherapy.”
Michael A. Postow, MD
The toxicity profile is much higher with the combination of nivolumab and ipilimumab but it is “way too premature” to say such regimens will be used widely, Postow said. “Right now in my opinion, the combination has mostly demonstrated benefits in melanoma and so I’m not sure what the combination will be doing in other diseases,” he said. “The side effect management of combination immunotherapy at least in melanoma, is a consideration and is something we have to be careful with in patient selection and during the treatment.”
Wolchok noted with satisfaction that there were no treatment-related deaths in the combination arm of the CheckMate-067 trial despite the higher rate of toxicities, which is particularly noteworthy considering the global nature of the study.
In all, 36.4% of patients in the combination arm of the trial discontinued treatment due to AEs. Yet remarkably, 67.5% of the 120 patients in that group developed a response, Wolchok said. That suggests some patients do not have to take checkpoint inhibitors until progression to experience the benefits.
“That’s a very important attribute of biologic therapy,” said Wolchok. “We’re treating the patient and their immune system not the tumor directly, so the idea that more drug for longer time equals better response is not necessarily true with immunotherapy.
“The goal is to get the patient’s immune system activated to a point where it can then treat the tumor and in different patients that may occur at different times,” he said. “Toxicity is telling you that you’ve achieved what you wanted to achieve.”
aEvents reported in ≥10% of patients
Wolchok JD et al. 2015 ASCO Annual Meeting; May 29-June 2, 2015; Chicago, IL. Abstract LBA1. Reprinted with permission.
Selecting Patients for Therapy
Beyond the question of managing toxicities looms the matter of which patients are most likely to benefit from anti-PD-1/PD-L1 therapy. The binding of the PD-1 protein to its ligand PD-L1 transmits an inhibitory signal to T cells, thus suppressing an immune response, and correlating PDL1 expression levels with response data has become a prime research goal.
Clinical trial findings, however, continue to present a complex picture concerning PD-L1 expression even as companies developing these drugs establish their own assays for measuring PD-L1 levels. Although patients with higher PD-L1 expression levels are more likely to respond to therapies targeting that checkpoint, individuals who are not PD-1-positive also have benefited from therapy.
“There is still a lot of debate about what is the right biomarker to use for some of these agents, such as the PD1 antibody and whether expression of PDL1 on a tumor is going to be a useful biomarker,” said Schuchter. “The jury is still out on that.”
In the CheckMate-067 trial, participants with higher levels of PD-L1 expression (≥5%) achieved a longer median progression-free survival (PFS) and a greater overall response rate (ORR) in both nivolumab-containing arms than did patients with lower PD-L1 levels (<5%).
Lynn M. Schuchter, MD, FASCO
The median PFS in the nivolumab/ipilimumab arm was 14 months for higher PD-L1 expression versus 11.2 months for lower PD-L1; in the nivolumab monotherapy arm, the median PFS was 14 months vs 5.3 months, respectively. Similarly, the ORRs in the combination arm for higher compared with lower PD-L1 levels were 72.1% versus 54.8%, respectively, in the combination arm, and 57.5% versus 41.3%, respectively, in the nivolumab-alone group.
Nevertheless, Wolchok said he would not use PD-L1 expression levels to exclude patients from therapy. Instead, he said the levels should be considered along with other clinical factors in consultation with the patient.
“As patients try to make decisions, especially based upon their comorbidities, their degree [of disease], their performance status, their relative frailty—this is important information to consider and for the first time in immunotherapy, it allows us to start to take a precision approach to who to offer this therapy to,” he said.
Schuchter said a clinical trial that correlated responses to pembrolizumab in patients with colorectal cancer and other advanced cancers with whether their tumors were deficient in DNA mismatch repair (MMR) represents an intriguing avenue of exploration for defining which patients to recommend for anti-PD-1 therapy.9
MMR deficiency results in a buildup of genetic mutations in a tumor, increasing the probability that the immune system will recognize and attack the tumor, researchers explained at ASCO. The deficiencies are found in 15%-20% of sporadic colorectal cancers and to varying degrees in other malignancies including stomach, small bowel, endometrial, prostate, and ovarian cancers.
“There is an emerging understanding that there may be a relationship between mutation profile and the kind of mutation, and hypermutated cancers that may be the group of cancers responding to treatment,” said Schuchter.
Although MMR deficiency is relevant in certain cancers, the implications for recommending patients most likely to benefit from checkpoint therapies in an increasingly value-based equation loom large, said Schuchter. Testing for MMR can easily be accomplished on colon cancer specimens for less than $1000, researchers have said.
“Our thinking in the past was more related to whether there are things about how the immune system is functioning that’s predicting how these therapies work,” said Schuchter. “I think it [MMR] is a new dimension in terms of thinking about why certain patients are responding to treatment and others are not.”