Nondriver mNSCLC: Stratifying to the KEYNOTE-189 Regimen

Mark A. Socinski, MD:Pembrolizumab was the immunotherapy drug that was developed along the lines of a biomarker, that being PD-L1 [programmed-cell death ligand 1]. You had to be PD-L1—positive to get this drug. In the trial from a couple years ago, 2016, KEYNOTE-024, it was shown that single-agent pembrolizumab in patients with the PD-L1 levels greater than 50% of tumor that pembrolizumab as a single agent was superior to standard chemotherapy in terms of progression-free survival, response rate, and overall survival. So, pembrolizumab as a single agent became the standard of care in the high expressors greater than 50%.

KEYNOTE-189 explored the combination of that with chemotherapy. When you look at the KEYNOTE-189 data, and you divide the PD-L1 subsets into greater than 50%, 1% to 49%, and then less than 1%, we would call those negative.

There was a benefit in all 3 subsets, so that’s important to realize. Now, like we’ve seen with the history of PD-L1, the stronger you stain, the greater the benefit. And there clearly was more benefit in KEYNOTE-189 in the greater than 50% relative to the lessor stainers, if you will. However, there was still significant benefit in even the negative patients and in the lesser stainers. So, I think PD-L1 is an important test. Its main importance, to me, is to identify those patients in which I might not have to use chemotherapy. If you’re greater than 50%, I use pembrolizumab alone based on the KEYNOTE-024 data. I wasn’t convinced with KEYNOTE-189 that adding chemotherapy to those patients greater than 50% added benefit relative to the toxicity.

Now I will admit that there may be patients who are greater than 50% where I would use chemotherapy. If you have a very symptomatic patient, if you had a patient with a large tumor volume, and you were worried that if you didn’t get control of that cancer, that patient was going to go downhill pretty quickly. I would then, even in greater than 50%, add chemotherapy, because the response rates are about 20% higher in that population. But most of the patients aren’t in that minority that I just described.

Most of the patients don’t have a high tumor burden, are relatively, I wouldn’t say asymptomatic, but minimally symptomatic. And I think that immunotherapy by itself is so well tolerated because you don’t have to add the chemotherapy in. Not that the adverse effects of chemotherapy are prohibitive nowadays. We do a pretty good job with supportive care in getting patients through chemotherapy with acceptable adverse effects. But if you don’t have to deal with them, why? If you aren’t convinced that chemotherapy’s adding significant benefit, why would you expose the patient to the potential risk of chemotherapy in that subset?

We talked a little bit about the convenience of this KEYNOTE-189 regimen in terms of the infusion time and the fact that it’s once every 3 weeks. I think that’s an advantage of this particular regimen. When I talk to patients about this recommendation—and specifically talk to them about the administration issues, it’s convenient compared with many of the things we do in cancer—I talk to them about the adverse effects to chemotherapy, which we’ve been doing for years, and the side effects of immunotherapy as 2 separate conversations. And I have my standard discussion about nausea, vomiting, alopecia, GI [gastrointestinal] toxicity, blood counts, those sorts of things with the chemotherapy. And then I have my talk with them about immunotherapy, which is all the “-itises” we talked about. And we certainly point out that things like colitis, dermatitis, hepatitis, these sorts of things are relatively common.

I do tell them about thyroid dysfunction, because that’s the most common endocrine abnormality we see in this population. Having said all of that, the vast majority of patients tolerate the triplet regimen relatively well. I think the important aspects of it are being prospective in the management of anticipated adverse effects, and educating patients and our nurses who are the front line of management of adverse effects to recognize the immune-related adverse events early. When they’re grade 1 or 2, to intervene early on, not let them get to be more severe and be grade 3 or 4, where then you’re talking about discontinuing treatment in that particular setting.

Transcript edited for clarity.

A 66-Year-Old Man With NSCLC

May 2018: H&P

• A 66-year-old man presented to primary care with complaints of persistent cough and shortness of breath with easy exertion.
  • PE: Average height, very thin (BMI = 18 kg/m²); says he has been losing weight although not dieting; mild fever (100.6 degrees); intermittent hemoptysis
  • Lab results: CrCL 75 mL/min; A1C 6.8%; WBC 15K/µL
  • PMH: HTN managed on atenolol; former smoker (30 pack-years); attributes cough to smoking but has persisted for 3 years now since he quit
• Primary care suspected bronchitis and prescribed amoxicillin; referred to pulmonology

June 2018: Pulmonology evaluation

• Pulmonologist evaluated patient for COPD: diminished lung function on spirometry
• CT revealed a 3-cm mass in left lung and multiple (<2 cm) masses in right lung, pleura, and axial lymph nodes; patient referred to oncology.

July 2018: Oncology exam

• Biopsy identified adenocarcinoma in left lung with lymph node and pleural involvement
• Molecular testing:
  • ALK& ROS1 rearrangement, negative
  • EGFR, KRAS wild-type
  • KRAS negative
  • PD-L1 TPS: 45%
• Additional testing: Abdominal CT, NED; Brain MRI, NED
• Diagnosis: Stage IVA lung adenocarcinoma without molecular drivers

August 2018

• Patient begins treatment with pemetrexed/carboplatin plus pembrolizumab 200 mg q3 wks