Despite the issues in biomarker status, dosage, and cost-effectiveness, we cannot deny the fact that KEYNOTE-010 has established pembrolizumab as second- or third-line therapy for advanced-stage NSCLC.— Tony S.K. Mok, MD
“Have you received your immune checkpoint inhibitor yet?”
I suspect St. Peter may have started asking this question routinely at the Pearly Gates to Heaven. If St. Peter has not, I am sure most oncologists have.
With extensive media coverage on the approval of nivolumab (Opdivo) and pembrolizumab (Keytruda) as second-line therapies, many patients with advanced non–small cell lung cancer (NSCLC) will inquire whether immune checkpoint inhibitor treatment is a better treatment option. Attending oncologists will likely refer to a recent study reported by Herbst et al in The Lancet, and summarized in The ASCO Post (June 10, 2016), which compared two doses of pembrolizumab (2 mg/kg or 10 mg/kg) with docetaxel in patients who had progressed after first-line platinum-based doublet chemotherapy.1 Investigators selected only patients with PD-L1 (programmed cell death ligand 1) expression (using antibody 22C3), and patients with < 1% expression were excluded.
Three of the four prespecified primary endpoints were met, which included improvement in overall survival in the total population and the ≥ 50% PD-L1–expression stratum and progression-free survival in the ≥ 50% stratum. The difference in progression-free survival in the total population was not statistically significant. These findings are supplementary to two other randomized studies on immune checkpoint inhibitors, namely the CheckMate studies 017 and 057,2,3 which compared nivolumab with docetaxel in advanced-stage squamous cell carcinoma and nonsquamous cell carcinoma, respectively. Based on these three randomized studies, doctors can say with confidence that immune checkpoint inhibitor therapy is superior to chemotherapy as second-line therapy. This implies that almost all patients with advanced NSCLC may include an immune checkpoint inhibitor as part of their treatment. The change in paradigm is exciting and yet can be very expensive.
A Valid Biomarker?
Although we are excited about these recent developments on new treatment options, a number of controversial issues remain unsettled. Should all patients be tested for PD-L1 expression prior to treatment? Is pembrolizumab at 2 mg/kg the correct dose? With tumor response rate limited to less than 20%, is pembrolizumab actually cost-effective?
Biomarkers are the key to precision medicine. An effective treatment cannot be considered “precise” unless treatment efficacy is correlated with a biologically sound and validated biomarker that can effectively distinguish the responder from the nonresponder. In KEYNOTE-010, the prevalence rate of PD-L1 expression was 29% in patients with a PD-L1 tumor proportion score ≥ 50%, 34% in those with a score of 1% to 49%, and 34% in those with a score < 1%. (These data were based on 2,222 patients whose tumor samples were assessable for PD-L1 expression, including 1,188 who did not meet eligibility criteria.) The distribution is similar to that of patients from KEYNOTE-001.4
Hazard ratios (HR) for overall survival in the ≥ 50% and 1% to 49% tumor proportion score strata were 0.53 and 0.76, respectively. However, patients with < 1% tumor proportion score were excluded from this study, and it is thus impossible to define the negative predictive power of PD-L1 expression in the study. Recalling data from KEYNOTE-001, there was indeed no difference in either progression-free or overall survival between the < 1% and 1% to 49% tumor proportion score subgroups. A biomarker without a validated negative predictive power cannot be an effective biomarker.
Minimal Effective Dose
Neither the 2-mg/kg nor the 10‑mg/ kg dose was the maximum tolerated dose of pembrolizumab. Only six patients from KEYNOTE-001 trial received pembrolizumab at 2 mg/kg. KEYNOTE-010 confirms the two doses to be similar in efficacy and toxicity profile.
It is only logical to use the lower 2‑mg/kg dose, but the current study failed to address the minimal effective dose. The current endorsed dose is not established according to minimal toxicity or optimal efficacy. We observed a similar controversy with the use of nivolumab for treatment of advanced malignant melanoma. Although the recommended dose is 3 mg/kg every 2 weeks, 8 of 27 patients (30%) who received 1 mg/kg actually responded.5 Further investigation into the minimal effective dose is potentially cost-saving for patients.
Cost-effectiveness is one of the more difficult endpoints to evaluate. Drug efficacy is founded on science, whereas drug cost is arbitrary and not necessarily based on science. The best scenario to save overall cost is to be able to select patients who benefit the most from treatment and exclude those who will not benefit prior to initiation of treatment.
Despite the high cost of epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors, we were able to establish cost-effectiveness of EGFR tyrosine kinase inhibitors treatment by selecting only patients with EGFR mutation for this treatment.6 It will be a challenge to do the same for immune checkpoint inhibitor therapy when the current biomarker using PD-L1 expression is associated with a weak negative predictive value. Patients in the ≥ 50% tumor proportion score stratum may benefit most from pembrolizumab; however, it will be almost impossible to exclude patients in the 1% to 49% stratum from receiving the drug. When more patients with less benefit receive the drug, it will be less likely to be cost-effective.
Despite the issues in biomarker status, dosage, and cost-effectiveness, we cannot deny the fact that KEYNOTE-010 has established pembrolizumab as second- or third-line therapy for advanced-stage NSCLC. Indeed, the upcoming positive data from KEYNOTE-024 may actually put pembrolizumab among the first-line therapies for patients with ≥ 50% PD-L1 expression. With such rapid development, doctors will be even more likely to ask…
“Have you received your immune checkpoint inhibitor yet?” ■
Disclosure: Dr. Mok is on the speakers bureau of AstraZeneca, Roche/Genentech, Pfizer, Eli Lilly, Boehringer Ingelheim, MSD, Amgen, Janssen, Clovis Oncology, GSK, Novartis, BMS, PrIME Oncology; has received honoraria from AstraZeneca, Roche/Genentech, Pfizer, Eli Lilly, Boehringer Ingelheim, Merck Serono, MSD, Janssen, Clovis Oncology, BioMarin, GSK, Novartis, SFJ Pharmaceutical, ACEA Biosciences, Inc., Vertex Pharmaceuticals, BMS, Aveo & Biodesix, Prime Oncology, Amgen; is a stock shareholder in Sanomics Limited; and is on the advisory board for AstraZeneca, Roche/Genentech, Pfizer, Eli Lilly, Boehringer Ingelheim, Merck Serono, MSD, Janssen, Clovis Oncology, BioMarin, GSK, Novartis, SFJ Pharmaceutical, ACEA Biosciences, Inc., Vertex Pharmaceuticals, Aveo & Biodesix, and BMS.
1. 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.
6. de Lima Lopes G Jr, Segel JE, Tan DS, et al: Cost-effectiveness of epidermal growth factor receptor mutation testing and first-line treatment with gefitinib for patients with advanced adenocarcinoma of the lung. Cancer 118:1032-1039, 2012.