Overall survival from the KEYNOTE-001 trial in patients with advanced non–small cell lung cancer showed treatment with pembrolizumab monotherapy produced more favorable rates of response at 5 years when compared with historical, pre-immunotherapy standard-of-care regimens.
Edward B. Garon, MD, MS
Overall survival (OS) data from the KEYNOTE-001 trial in patients with advanced nonsmall cell lung cancer (NSCLC) showed treatment with pembrolizumab (Keytruda) monotherapy produced more favorable rates of response at 5 years when compared with historical, pre-immunotherapy standard-of-care regimens. Updated results demonstrated a 5-year OS rate of 23.2% in treatment-naïve patients; in previously treated patients, that rate was 15.5%.
These results were presented at a press briefing that took place during the 2019 ASCO Annual Meeting of the phase Ib KEYNOTE-001 trial, the first trial ever to evaluate the PD-1 inhibitor in patients with advanced NSCLC. These results represent the longest safety and efficacy data presented for patients with NSCLC treated with pembrolizumab.
“In total, the data confirmed that pembrolizumab has the potential to improve long-term outcomes for both treatment-naïve and previously treated patients with advanced nonsmall cell lung cancer,” commented lead study author Edward B. Garon, MD, MS, when presenting the findings.
The phase Ib trial (NCT01295827) enrolled 550 patients with confirmed locally advanced or metastatic NSCLC in the lung cancer cohort and treated them with 2 mg/kg of pembrolizumab every 3 weeks or at a 10-mg/kg dose every 2 or 3 weeks. More recently, the protocol for the trial was changed to administer a single dose of 200 mg of pembrolizumab for all patients regardless of body weight every 3 weeks, which is typically used now in clinical practice.
All patients were required to provide a contemporaneous tumor sample, which was tested for PD-L1 expression levels using the immunohistochemistry 22C3 pharmDx test and categorized by tumor proportion score (TPS). However, only the treatment-naïve patients were required to have PD-L1positive disease. The primary endpoint for the trial was objective response rate (ORR) and OS was a key secondary endpoint.
As of the data cutoff date of November 5, 2018, the median follow-up was 60.6 months (range, 51.8-77.9), and at this point 82% of the participants (450 of 550) had died.
Of the 550 enrolled participants, 449 were previously treated and 101 were treatment naïve. The investigator-assessed ORRs by immune-related response criteria were 42% (95% CI, 32%-52%) in the treatment-naïve patients and 23% (95% CI, 19%-27%) in the previously treated patients. However, it was noted that most of these responses were partial responses. The median duration of response was 16.8 months (range, 2.1+ to 55.7+) for the untreated patients and was 38.9 months (range, 1.0+ to 71.8+) in the previously treated patients.
The median OS for the treatment-naïve patients was 22.3 months (95% CI, 17.1-32.3). The untreated patients who had a PD-L1 TPS of 1% to 49% (n = 52) had a 5-year OS rate of 15.7% and those with a TPS of ≥50% (n = 27) had a rate of 29.6%. The 5-year OS rate for treatment-naïve patients with a lower TPS (1%-49%) was 15.7% and was 29.6% for those with a higher TPS (≥50%).
In the previously treated patient population, the overall median OS was 10.5 months (95% CI, 8.6-13.2). Ninety previously treated patients had a TPS below 1%, and the 5-year OS rate was 3.5% in this subgroup. Of the patients with a TPS of 1% to 49% (n = 168), the 5-year OS rate was 12.6%, and the OS rate was 25.0% in the patients with a TPS ≥50% (n = 138).
“[Advanced] nonsmall cell lung cancer generally has been a disease where we have considered patients to be very unlikely to be living 5 years…and it is certainly encouraging to have seen an over 15% 5-year survival in this group,” said Garon, an associate professor of medicine at the David Geffer School of Medicine, University of California, Los Angeles. “It was particularly impressive to see the 25% 5-year survival rate in patients who received pembrolizumab who had PD-L1 of ≥50%. Patients who have this level of PD-L1 staining, often in clinical practice, do receive single-agent pembrolizumab and to have these 5-year survival data is certainly encouraging.”
Sixty patients had received treatment for ≥2 years, which is the length of time pembrolizumab is given in current clinical trials, Garon explained. These patients, 46 of which were still alive at data cutoff, were treated for a median duration of 36 months (range, 17.3-75.9). Of the 60 patients, the 14 who had not received prior treatment had an ORR of 86% and a 5-year OS rate of 78.6%. Seven of these patients (58%) were still in response at data cutoff. Forty-six patients were previously treated before receiving pembrolizumab, and the ORR in this patient population was 91% and the 5-year OS rate was 75.8%. The median duration of response was not reached in these patients and 30 (71%) had an ongoing response at data cutoff.
Immune-mediated adverse events (AEs) were observed in 17% of the patients at 5 years, which was similar to the rate seen when the 3-year results were reported. “We have looked to assess whether there is a late-onset of toxicities, particularly immune-related toxicities...there is a fairly minor increase in the immune-mediated adverse events between 3 and 5 years,” Garon noted.
The most common AE observed was hypothyroidism, which showed a slight increase of rates at 3 years, and the most common serious AE was pneumonitis, which was grade 3 to 5 for almost half of all cases at both 3 years and 5 years.
Earlier results from this trial led to the accelerated approval in October 2015 for pembrolizumab as a treatment for patients with previously treated advanced NSCLC. The PD-1 inhibitor was later approved for the first-line setting as well for patients with advanced NSCLC who do not harborEGFRorALKaberrations.
Garon EB, Hellmann MD, Costa EC, et al. Five-year long-term overall survival for patients with advanced NSCLC treated with pembrolizumab: Results from KEYNOTE-001.J Clin Oncol. 2019;37(suppl; abstract LBA9015).