New Hope for Patients with Advanced Lung Cancer?

Phoebe Starr

May 2015, Vol 6, No 4 - AACR Meeting Highlights


Philadelphia, PA—Immunotherapy with pembrolizumab (Keytruda) was safe and effective in the treatment of patients with previously treated as well as treatment-naïve patients with advanced non–small-cell lung cancer (NSCLC), according to results of the KEYNOTE-001 trial, said Edward B. Garon, MD, Assistant Professor of Medicine, David Geffen School of Medicine, University of California, Los Angeles, Santa Monica, who presented these data at the 2015 American Association for Cancer Research (AACR) meeting.

The overall response rates (ORRs) and progression-free survival (PFS) correlated with high levels of PD-1 ligand 1 (PD-L1) expression in tumor tissue, which may turn out to be a biomarker for response to pembrolizumab. No significant differences in safety and efficacy were observed at the different dose levels of pembrolizumab used in this study.

Studies of other anti–PD-1 therapies have been disappointing with regard to use of PD-L1 expression as a biomarker. The Merck assay developed specifically for pembrolizumab used staining in ≥50% of tumor cells as a cutoff point for high PD-L1 expression. Other companies developing anti–PD-1 therapies have their own assay, using different cutoff points and different techniques.

“With these data, we can confidently say, in previously treated patients with NSCLC and PD-L1 expression in at least 50% of tumor cells, that pembro­lizumab is associated with superior outcomes compared to what can be expected with cytotoxic chemotherapy,” Dr Garon stated. The study was published online in the New England Journal of Medicine (Garon EB, et al. 2015 April 19. Epub ahead of print) to coincide with his presentation.

The phase 1b KEYNOTE-001 study included 495 treatment-naïve and previously treated patients with NSCLC to 2 different dose levels of pembrolizumab (2 mg every 3 weeks or 10 mg/kg every 2 weeks); the patients were assigned to a training group (N = 182) or a validation group (N = 313). PD-L1 expression was assessed in all tumor samples.

The overall results showed that the ORR was 19.4%, with no clear difference in response, regardless of previous treatment, dose of pembrolizumab, or histology. Among the patients who responded, the median duration of response was 12.5 months. The median duration of PFS was 3.7 months, and the median duration of overall survival (OS) was 12 months.

Pembrolizumab was generally well-tolerated. The most common treatment-related adverse events were fatigue, pruritus, and decreased appetite, and these effects were similar at all ­dose levels.

In the training set, a proportion score of 50% (≥50% of cells that stained positive for PD-L1) was found to be highly positive. In the validation set, patients were categorized for PD-L1 expression as follows: a proportion score <1% was considered low; a score of 1% to 49%, intermediate; and a score ≥50%, high. The validation set included 313 patients (223 previously treated, and 90 treatment-naïve to systemic therapy for NSCLC).

PD-L1 staining correlated with response, with no difference in previously treated or treatment-naïve patients. In the highest expressers, those with a proportion score ≥50%, the ORR was 42.5%, the median PFS was 6.3 months, and the median OS had not yet been reached.

The ORR was significantly better in the group with a proportion score ≥50% than in patients with a score from 1% to 49%, or <1% (P <.001 and P = .01, respectively).

Pooled data from the training and validation set showed that PFS correlated with high PD-L1 expression. (The median PFS was 6.3 months for the group with a proportion score >50% versus 3.3 months for a score from 1% to 49%, and 2.3 months for a score <1%.)

At a median follow-up of 10.9 months, the median OS was not yet reached in the group with a score ≥50% in previously treated and treatment-naïve patients. In the previously treated patients, the median OS was 7.3 months for those with a score from 1% to 49% and 8.6 months for those with a score <1%.

In treatment-naïve patients, the median OS was 16.2 months for those with a proportion score from 1% to 49% and 10.4 months for the proportion score <1% group.

Dr Garon emphasized that PD-L1 is an imperfect biomarker, because a proportion of patients with a proportion score <1% respond to pembrolizumab, as did some of those in the proportion score 1% to 49% group. It would be important not to exclude those with low or intermediate PD-L1 expression who failed with other treatments from being considered for pembrolizumab treatment at this point.

“The results are impressive in this difficult-to-treat patient population. We don’t expect to see improvements in survival in this population,” stated ­Suzanne L. Topalian, MD, Professor of Surgery and Oncology, Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University. Dr Topalian moderated a press conference where these data were discussed.