Top 10 Lessons Learned So Far About Treating Lung Cancer With Immunotherapy

Get Permission

Roy S. Herbst, MD, PhD

I haven’t felt this type of excitement around new drugs since EGFR inhibitors were developed more than a decade ago.

—Roy S. Herbst, MD, PhD

Immunotherapy agents “really work” in treating lung cancer, but they have unique toxicities, are challenging to combine with other therapies, and questions remain about dose and duration, Roy S. Herbst, MD, PhD, stated at the 2014 Chicago Multidisciplinary Symposium in Thoracic Oncology. “There are many agents—agonists, antagonists—some antibodies activate, some inhibit. We are quite lucky that there is such a fertile field,” he said. “It is very promising for our patients. We’re helping 20% now. How can we help the other 80% in the future?”

Dr. Herbst is Chief of Medical Oncology, Yale Cancer Center and Smilow Cancer Hospital at Yale-New Haven, and Associate Director for Translational Research and Clinical Research Program Leader, Thoracic Oncology Program, Yale Cancer Center, New Haven, Connecticut. At the Chicago meeting, he presented a countdown of the “top 10 lessons” learned about immunotherapy in treating lung cancer and shared his concerns and optimism about tackling the unknowns.

10. “These agents really work. It is clearly a breakthrough for the patients,” Dr. Herbst noted. “The activity is even a little bit better” than that displayed by epidermal growth factor receptor (EGFR) inhibitors when they were developed 10 or 15 years ago,” he said. “What is really striking is the durability [of immunotherapy agents],” he added. While resistance does develop, “many patients have a very long durability of the response, because the immune system can adapt.”

Dr. Herbst cited a phase I study presented at the symposium, which found that the anti–programmed cell death protein 1 (PD-1) antibody pembrolizumab (Keytruda) provided robust antitumor activity as first-line therapy for patients with programmed death ligand-1 (PD-L1)–positive advanced non–small cell lung cancer (NSCLC).1 The overall response rate was 26% by RECIST criteria and 47% by immune-related response criteria, and 100% of RECIST responders and 90% of immune-related response criteria responders had ongoing responses.

In a phase II trial with another anti–PD-1 antibody, nivolumab (Opdivo),2 “while the response rate was 15%, a little bit lower than targeted, the median overall survival of 8.2 months and the 1-year survival of 41% are wonderful results in the third-line setting in squamous cell lung cancer,” Dr. Herbst said.

9. “It is still unclear what is the most appropriate endpoint,” Dr. Herbst pointed out. “Delayed response is something we need to keep in mind, and it does change the way we think about clinical trials. The endpoint should be response rate, sure. Do we need to have a 50% response rate to get these drugs approved in a breakthrough setting? I would think not, because of the large durability of response, but what number you should have is unclear. We should probably look at other endpoints, including survival endpoints, and perhaps survival endpoints at landmarks,” he suggested. “This needs to be thought out in an innovative way.”

8. “All checkpoint antibodies are not the same.” Dr. Herbst observed that the checkpoint antibodies all have different binding affinities, slightly different targets, and different abilities to avoid antibody-dependent cell-mediated cytoxicity. “Probably the major difference,” is between the anti–PD-1 and the anti–PD-L1 antibodies. “Which is better, only the clinical trials can tell. We need robust experiences to really compare them, perhaps someday head-to-head experiences,” Dr. Herbst said.

7. “These agents, while different from chemotherapy, do have unique toxiocities,” Dr. Herbst noted. “As medical oncologists, lung cancer oncologists, we are now dealing in ‘itises’—colitis, renal nephritis, pneumonitis, hepatitis,” Dr. Herbst said. “Dermatologic effects do occur,” he added, “along with some neurologic autoimmune-type phenomena in rare cases” and “endocrinopathies, diseases of the pituitary, thyroid, and adrenal glands. You have to think about it, look for it, and manage it. Sometimes you need to talk about it with your endocrinologist.”

6. “The PD-L1 biomarker has some flaws,” Dr. Herbst said. With four to five companies working on this, each one with a different marker, “this is really a mess,” Dr. Herbst said. There are a lot of issues to be considered, including heterogeneity of the tumor, type of biopsy, interval between biopsy and treatment, and primary vs metastatic disease. “This is a continuous variable. Patients can be 5% positive, 10% positive, 20% positive,” Dr. Herbst noted. “It’s a moving target,” he added, and “the assay is going to help define the field.”

5. “Science can help ‘drive the show.’ This is [a lesson] that is close to my heart,” Dr. Herbst admitted. He reminded colleagues that lung cancer is linked to many mutations and that biopsies and immune monitoring should be done when possible. “Now we can make the clinic our lab and ask what is going on with the tumor,” he said.

“With pretreatment and posttreatment biopsies, you can see dynamically what is happening in the immune process,” he continued. “You can also see when it doesn’t work. So if a patient is not responding, for example, you can see a rim of T cells not getting to the tumor.”3 Several companies are developing new assays for dynamic monitoring.

4. “Questions remain regarding the dose and duration of therapy,” Dr. Herbst said. “How long do you treat—1 year or 2 years? What is the right dose? No one really knows,” Dr. Herbst acknowledged. “Do you need to retreat if someone becomes refractory? These are all issues that might be helped with more immune monitoring.”

3. “Combination therapy is a must,” Dr. Herbst said, but what should immunotherapy be combined with? “Are we going to combine with chemotherapy, targeted therapy, immune therapy, other checkpoint inhibitors? Probably all of the above,” he predicted. “We have to think about this very carefully. We can’t just combine therapies like we used to. We have to say, ‘What is that small-molecule agent going to do to the T cell?’ We don’t want to do something to the tumor cell that is also going to inhibit the T cell. It is very complicated. We need personalized immunotherapy,” Dr. Herbst said.

“The one combination that certainly deserves all of our interest in lung cancer is CTLA-4 and PD-1. Why? Because it works great in melanoma.,” Dr. Herbst said. “Our patients are different, but still, the idea that you target CTLA-4, you target the T cell and prime it when it is in the lymph node, and then you work with the PD-1/PD-L1 in the tumor microenvironment—that is a proven mechanism for melanoma, with 1-year survival rates of up to 80%.”

2. “Immunotherapy will be used in all lines of therapy,” Dr. Herbst predicted. This will “absolutely” happen.

1. “It’s a horse race,” Dr. Herbst admitted. “I haven’t felt this type of excitement around new drugs since EGFR inhibitors were developed more than a decade ago,” he said. He couldn’t predict who would get the first checkpoint drug approved for lung cancer and what that drug would be, but the “winning group will be the one with the best drug, the best biomarker, the best strategy, and a little bit of luck.” The true winners, he noted, will be “our patients.” ■

Disclosure: Dr. Herbst reported no potential conflicts of interest.


1. Balmonoukian AS, Rivzi AN, Garon EB, et al: Safety and clinical activity of pembrolizumab (MK-3475) as initial therapy in patients with advanced non-small-cell lung cancer. Chicago Multidisciplinary Symposium in Thoracic Oncology. Abstract 2. Presented October 31, 2014.

2. Ramalingam SS, Mazières, Planchard D, et al: Phase II study of nivolumab (anti-PD-1, BMS-936558, ONO-4538) in patients with advanced, refractory, squamous non-small-cell lung cancer. Chicago Multidisciplinary Symposium in Thoracic Oncology. Abstract LBA2. Presented October 31, 2014.

3. Herbst RS, Soria JC, Kowanetz M, et al: Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients. Nature 515:563-567, 2014.