Defining Progression of NSCLC on Immunotherapy

Corey Langer, MD:So, stage aside, let’s assume that this individual really did have disease that was not potentially curable at contralateral pulmonary metastases or cannot have his tumor contained within a single radiation field. The median survival in this group historically has been in the order of about 8 to 12 months with standard chemotherapy. There have been 2 important wrinkles in the last 5 years. The advent of nanoparticle albumin-bound paclitaxel in combination with carboplatin, particularly in squamous cell, has resulted in much higher response rates in the order of 40% to 45% compared to about 25% from using conventional Q3-week solvent-based paclitaxel with carboplatin. And in the large phase III trial that ultimately led to nanoparticle albumin-bound paclitaxel being approved, in the elderly cohort—about 150 individuals—we observed a 9-month improvement in overall survival, in median survival, from about 10.5 months using conventional 2-week or 3-week paclitaxel and carboplatin to 19.5 months, with the elderly cohort receiving nab-paclitaxel and carboplatin. So, if this individual did not harbor PD-L1 positivity, if the tumor had less than 50% expression, my go-to regimen—at least as of 2017 and really for the last 3 to 4 years—has been combination carboplatin with weekly nab-paclitaxel grafted on to the carboplatin.

It’s important to realize that performance status trumps age. Individuals who are PS 0 or PS 1 do far better than advanced lung cancer patients who are PS 2 or 3, regardless of age. Age alone is not an impediment to therapy. We look across the broad therapeutic landscape to individuals 70 years of age and older who’ve received systemic therapy, whether it’s chemotherapy or, for that matter, now immune checkpoint blockade. Survival rates are as good or nearly as good in that older cohort as they are in younger individuals. So, baseline comorbidities, baseline performance status figures much more strongly in our therapeutic decision making. And remember, this individual was up and about, virtually asymptomatic, golfing, maintaining all his normal activities.

This gentleman’s experience, unfortunately, is a big atypical for those with 50% or higher expression PD-L1. Remember, he was about 65% or so. You look at KEYNOTE-024; nearly half those individuals will have bona fide partial responses. At best, he had stable disease. The median progression-free survival in the pembrolizumab group, the group that received immune checkpoint blockade, was over 10 months compared to 6 months in the control arm. His PFS, at best, was about 5 months. So, unfortunately, he has drawn the short straw, at least from a prognostic or therapeutic standpoint.

In terms of how individuals manifest disease progression, there’s no set standard or no set scenario. It can be intracranial progression alone. This gentleman did have a seizure and that was the tipoff to new sites of disease. It may just be local progression or progression in established sites of cancer. For the minimum individuals who were receiving checkpoint blockade, chemotherapy, or a TKI, I will usually obtain follow-up imaging, generally a CAT scan, at 2 to 3 months for the first 6 months. And then, if the disease is under control, if we’ve seen a response or at least disease stability, we start spacing it out at 3-month intervals. So, that’s our usual standard approach.

Pseudoprogression is actually quite rare. It has been seen in up to 10% to 15% of individuals with melanoma, which is where the checkpoint blockade was originally approved. In lung cancer, it’s a bit lower. I was recently at a meeting with Dr. Brahmer, who’s really one of the gurus of checkpoint blockade, and she cited a 6% incidence. We have looked at 175 consecutive individuals treated with nivolumab from the time of its first approval, March of 2015, to March of 2016. So, that’s 175 patients with advanced disease, all of whom received single-agent nivolumab, and our incidence of pseudoprogression was 2.8%—5 individuals in that number. It’s important to look for, but if the patient is having rapid growth of tumor and is becoming increasingly symptomatic, that almost invariably is real progression, not pseudoprogression. Pseudoprogression usually is associated with the least symptom stability, occasionally symptom improvement. And the pace of progression is, by and large, not uniformly, a bit more indolent.

Transcript edited for clarity.

November 2016

• An 81-year-old male presents to his physician with symptoms of cough, hemoptysis, and fatigue requiring frequent rest during the day
• PMH includes hypercholesterolemia, controlled on simvastatin and hypertension, controlled on a calcium channel blocker; mild osteoarthritis
• He has no history of smoking
• The patient is physically active and plays golf several days per week
• CT of the chest revealed a solid cystic mass in the left upper lobe and lymphadenopathy in the left hilar and bilateral mediastinal nodes
• PET/CT imaging showed 18F-FDG uptake in the lung mass, left hilar and both mediastinal lymph nodes
• Bronchoscopy and transbronchial lung biopsy were performed
• Pathology showed grade 3 squamous cell carcinoma of the lung
• Genetic testing was negative for known driver mutations
• PD-L1 testing by IHC showed expression in 65% of cells
• The patient was started on therapy with pembrolizumab
• Follow up imaging at 3 months showed stable disease

April 2017

• After 5 months on immunotherapy, the patient was hospitalized after having a seizure. He reported worsening fatigue and cough for 1 month
• CT showed increased size of the left upper lobe pulmonary mass
• Brain imaging showed several small intracranial lesions
• WBRT was started
• Immunotherapy was discontinued and the patient was started on carboplatin and nab-paclitaxel