Immunotherapy Challenges in Lung Cancer: From Patient Selection to Clinical and Financial Toxicity

Get Permission

Immunotherapy has been a major advance in lung cancer, but it is not without its challenges, according to Sanjay Popat, PhD, FRCP, a consultant medical oncologist and reader in cancer medicine at the Royal Marsden Hospital, London, UK. He reviewed some of the challenges pertaining to the use of immune checkpoint inhibitors of programmed cell death protein 1 (PD-1) and programmed cell death ligand 1 (PD-L1) for advanced non–small cell lung cancer (NSCLC) in a workshop at the International Association for the Study of Lung Cancer 17th World Conference on Lung Cancer.1

Selecting Appropriate Patients

“It’s quite difficult to know who is the most appropriate patient for immunotherapy because there are conflicting data. But there are some things for which we do have good solid data,” Dr. Popat shared.

Tumor PD-L1 expression is the currently used biomarker for benefit, although the optimal cutoff is still unclear. For example, post hoc analyses of the CheckMate 057 second-line trial in nonsquamous NSCLC showed that although patients whose tumors showed at least 1% positivity for PD-L1 lived longer if given nivolumab (Opdivo) instead of docetaxel—and the drug is approved for use irrespective of PD-L1 status—benefit was almost entirely driven by those having at least 50% positivity.2 “So these are controversial data for the predictive benefit of nivolumab in the second-line setting, although post hoc analyses should be interpreted with caution,” he added.

Every organ in the body can be affected by immune-related toxicity from immunotherapy.
— Sanjay Popat, PhD, FRCP

Tweet this quote

Shorter time (less than 3 months) since the last treatment and progressive disease as the best response to that treatment also portended poorer early survival. In the OAK second- and third-line trial in NSCLC, progression-free survival in the first several months was actually poorer with atezolizumab (Tecentriq) than with docetaxel, although the former became superior in the longer term.3

“This is a recurring theme that we are seeing with a lot of the new therapies,” Dr. Popat observed. About 15% of patients fare more poorly on immunotherapy, whereas the majority see a benefit, underscoring the importance of identifying better predictors of benefit.

Analyses from these trials as well as the KEYNOTE-010 second-line trial of pembrolizumab (Keytruda)4 have determined that never-smokers and those with epidermal growth factor receptor (EGFR) mutations derive little, if any, benefit from immune checkpoint inhibitors.

In fact, among never-smokers, the survival seen in early-phase trial data with pembrolizumab in those with tumors having at least 50% PD-L1 positivity is about the same as that in counterparts with less than 1% positivity.5 “So in the never-smokers, PD-L1 expression tends not to be a very good biomarker of predictive benefit,” Dr. Popat maintained.

For patients whose tumors harbor EGFR mutations, appropriate first-line therapy is typically a tyrosine kinase inhibitor that addresses this molecular aberration. “In my personal practice, immunotherapy is something I give them way down the line. I generally start with a tyrosine kinase inhibitor, tyrosine kinase inhibitor, then chemotherapy, and then immunotherapy down the line,” he detailed.

Limitations of PD-L1 Assessment

“There are lots and lots of limitations when thinking about PD-L1 assessment,” Dr. Popat asserted. “For a start, PD-L1 expression changes over time. If you give chemotherapy, it can upregulate. If you give a CTLA-4 [cytotoxic T-lymphocyte–associated protein 4] inhibitor, it can upregulate. If you give radiotherapy, it can upregulate.”

Additionally, PD-L1 expression is heterogeneous within tumor tissue, and the significance of its presence in tumor-infiltrating immune cells is unknown. And, as previously noted, the optimal threshold for predicting benefit is unclear.

It’s quite difficult to know who is the most appropriate patient for immunotherapy because there are conflicting data.
— Sanjay Popat, PhD, FRCP

Tweet this quote

Another potential limitation is uncertainty as to whether archived material, perhaps from years ago, can be used for PD-L1 testing or a new biopsy sample is needed. “The data would suggest that the archived material’s okay, but there are scientific questions as to how valid that is,” Dr. Popat said. The scalability and reliability of the assays when performed in different institutions are yet to be defined as well.

Currently, at least four assays are used or in late development for PD-L1 testing, each paired with a specific PD-1 or PD-L1 inhibitor. Each assay is performed with its own antibody clone and must be run on a dedicated platform, raising concern about the comparability of results.

Findings of the Blueprint Project suggest that three of the four assays perform similarly in lung cancers; the outlier is the companion assay for atezolizumab.6 “It’s not surprising that it performs slightly differently,” Dr. Popat commented. “It was not developed predominantly for lung cancer, and it usually requires the use of immune cells to evaluate [expression], whereas all the others are entirely based on tumor cells.”

Of note, for some tumors, this assay yielded negative PD-L1 results, where the others yielded positive results, which may have implications for interpretation of findings of trials of atezolizumab, such as the OAK trial, in which even patients with no PD-L1 expression, as assessed with this assay, saw a benefit.3 “One could be provocative and wonder how many of these negative patients who are deriving a benefit from atezolizumab would have been considered positive by other assays,” he proposed.

Treating Patients With Brain Metastases

About one-third of patients given a lung cancer diagnosis already have brain metastases at their initial presentation, according to Dr. Popat. Would it be safe to give these patients immunotherapy?

“We really don’t know because patients with active brain metastases couldn’t go onto these trials,” admitted Dr. Popat. “They had to have them treated with radiotherapy before they could go onto the trials,” he added.

However, a study of a cohort with untreated brain metastases that were small (measuring no more than 2 cm in diameter) suggested that pembrolizumab can be safely administered to this population.7 “There were responses that occurred extracranially and tended to occur intracranially as well,” Dr. Popat noted. “So it seems to be safe for small-volume brain metastases, but our data from trials are quite lacking.”

Immune-Related Toxicity

“Every organ in the body can be affected by immune-related toxicity [from immunotherapy], and the problem is that patients don’t walk in through the door and say, ‘Doctor, I’ve got adrenal insufficiency,’” Dr. Popat said. “They walk through the door with a whole host of different symptoms, so working out what the actual toxicity is can be quite challenging.”

These toxicities are predominantly grade 1 or 2.8 However, the grading system typically used, the Common Terminology Criteria for Adverse Events (CTCAE), was designed for chemotherapy toxicities, not immune-related toxicities. “So immune-related grade 2 diarrhea is actually clinically serious compared to grade 2 anemia,” he noted.

Although the immune-related toxicities with the PD-1 and PD-L1 inhibitors mainly occur in the first 3 months, they can still occur as far out as a 1 or 2 years into therapy.9 Educating patients is key. They need to know about these toxicities and the importance of telling other providers that they are on immunotherapy, so they receive appropriate treatment.

“These toxicities can go on for a long period as well, and we should bear in mind that patients can be on steroids [to treat the toxicity] for quite a long time,” Dr. Popat noted. Although they may be eager to stop the steroids because they worry about compromising the efficacy of immunotherapy, experience with pembrolizumab suggests progression-free and overall survival are unaffected by steroid therapy.8

Reducing Financial Toxicity

“We shouldn’t underestimate financial toxicity,” Dr. Popat commented, noting estimates suggest that a month of immunotherapy in the United States can cost $10,000 or more. “And when you’ve got co-pays, it really is quite crippling,” he admitted.

“Some countries, such as England, have a single health care payer system. In that setting, where you’ve got a payer who can negotiate with pharma, you are actually protected from extreme costs,” he concluded. ■

Disclosure: Dr. Popat is a consultant for AstraZeneca, Boehringer Ingelheim, Roche, Bristol-Myers Squibb, MSD, Eli Lilly, Pfizer, Novartis, Ariad, and Clovis Oncology; has received honoraria from Pfizer, Boehringer Ingelheim, Eli Lilly, Roche, Novartis and AstraZeneca; institutional research funding from Boehringer Ingelheim, Pfizer, Roche; and travel expenses from Bristol-Myers Squibb, MSD, Boehringer Ingelheim, and Pfizer.


1. Popat S: Immunotherapy challenges in lung cancer and mesothelioma. 2016 World Conference on Lung Cancer. Abstract WS07.05. Presented December 4, 2016.

2. European Medicines Agency: Assessment report EMA/246304/2016. February 25, 2016. Available at Accessed February 22, 2017.

3. Barlesi F, Park K, Ciardiello F, et al: Primary analysis from OAK, a randomized phase III study comparing atezolizumab with docetaxel in 2L/3L NSCLC. 2016 European Society for Medical Oncology. Abstract 4380. Presented December 18, 2016.

4. Herbst RS, Baas P, Kim DW, et al: Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): A randomised controlled trial. Lancet 387:1540-1550, 2016.

5. Hui R, Gandhi L, Costa EC, et al: Long-term OS for patients with advanced NSCLC enrolled in the KEYNOTE-001 study of pembrolizumab. 2016 ASCO Annual Meeting. Abstract 9026.

6. Hirsch FR: The Blueprint Project: Harmonizing companion diagnostics across a class of targeted therapies. 2016 American Association for Cancer Research. Presented April 19, 2016.

7. Goldberg SB, Gettinger SN, Mahajan A, et al: A phase II trial of pembrolizumab for untreated brain metastases from non-small cell lung cancer. 2015 World Conference on Lung Cancer. Abstract ORAL31.07.

8. Leighl N: Discussion. 2015 World Conference on Lung Cancer. Mini Oral Session MINI04.05.

9. Reckamp K: Discussion. 2015 World Conference on Lung Cancer. Mini Oral Session MINI04.10.