Immune-based therapies continue to show promising signals for patients with small cell lung cancer and mesothelioma, but better predictive biomarkers are needed to determine who is most likely to benefit.
Justin F. Gainor, MD
Immune-based therapies continue to show promising signals for patients with small cell lung cancer (SCLC) and mesothelioma, but better predictive biomarkers are needed to determine who is most likely to benefit, according to Justin F. Gainor, MD.
The rationale for immunotherapy in SCLC is growing because it is a very molecularly complex disease, Gainor said in a presentation during the 5th Annual Miami Lung Cancer Conference. While PD-L1 expression is infrequent in this space, SCLC tumors can be characterized by relatively high tumor mutation burden (TMB).
"We've seen responses with PD-1 monotherapy as well as with dual checkpoint blockade regardless of PD-L1 status," said Gainor, assistant professor of medicine at Harvard Medical School and an assistant in medicine at Massachusetts General Hospital. "The combination of PD-1/CTLA-4 is associated with a higher response rate compared with nivolumab alone, but this benefit appears to be most pronounced in high TMB patients."
Signals with checkpoint inhibitors were first reported with pembrolizumab (Keytruda) in the multi-cohort KEYNOTE-028 trial, which showed that the PD-1 inhibitor pembrolizumab was associated with a 33% overall response rate (ORR) in patients with PD-L1-positive SCLC.1 Of all patients enrolled on the study, 31.7% tested PD-L1 positive, defined as ≥1% PD-L1 expression in tumor and inflammatory cells or stroma.
Nivolumab (Opdivo) has also been investigated in several SCLC studies. The open-label, phase I/II CheckMate-032 trial, for example, evaluated 216 patients with progressive SCLC who had received ≥1 prior platinum-containing regimen and were not stratified for PD-L1 expression. Patients were randomly assigned to single-agent nivolumab or the combination of nivolumab and ipilimumab (Yervoy) at 1 of 2 doses.2
In an exploratory analysis of the trial, the combination of nivolumab and ipilimumab (Yervoy) was associated with a 46% ORR in patients with recurrent disease with high TMB compared with a 21% ORR with nivolumab monotherapy. In patients with medium and low levels of TMB treated with the combination, the ORR was 16% and 22%, respectively. The ORR was 7% and 5%, respectively, among those in these populations treated with nivolumab alone.
In the pooled intent-to-treat (ITT) population, the overall ORR was 11% with nivolumab alone and 22% with nivolumab plus ipilimumab.
"We can see that the response rate to nivolumab monotherapy is actually lower than what we see in an unselected patient population in non-small cell lung cancer, for example, so only about 11 to 12% of patients are responding to monotherapy," Gainor explained. "By contrast, the response rate to the combination is about double, at about 21% to 22%."
PD-L1 did not seem to be a reliable predictor for benefit based on an analysis of CheckMate-032.
"If you look at responses based on PD-L1 expression, there are responses in people who are PD-L1 positive and PD-L1 negative, so it is not a fantastic predictor of benefit," Gainor said.
Ongoing phase III studies are taking these clinical data a few steps further. In the phase III CheckMate-451 trial, approximately 810 patients with SCLC are being randomized in a 1:1:1 ratio to receive nivolumab, nivolumab plus ipilimumab, or placebo (NCT02538666). The primary endpoints are overall survival (OS) and progression-free survival (PFS).
Additionally, the CheckMate-331 study compares OS outcomes with nivolumab versus topotecan (Hycamtin) or amrubicin in patients with SCLC who did not previously receive treatment with CTLA-4, CD-137, or PD-1/PD-L1/PD-L2 inhibitors (NCT02481830).
Aside from PD-1 therapy, DLL3 is another target being explored in SCLC, Gainor explained. DLL3 is normally expressed during Golgi development, can be aberrantly expressed in SCLC tumor-initiating cells, and interacts with and inhibits NOTCH1 signaling. It may also mediate NOTCH inhibition downstream of ASCL1.
Rovalpituzumab-tesirine (Rova-T) is a DLL3-targeted antibody-drug conjugate that demonstrated encouraging single-agent antitumor activity with a manageable safety profile in a phase Ib trial for the treatment of patients with recurrent SCLC.3
In the study, 11 of 60 evaluable patients (18%) who received an active dose of Rova-T achieved a confirmed objective response, and 30 patients (50%) had stable disease. The median PFS was 2.8 months (95% CI, 2.5-4.0).
An exploratory analysis of available tumor tissue samples (n = 34) from this trial showed an overall response rate of 38% (n = 10) among 26 DLL3-high patients compared with 0% in DLL3-low patients. The median PFS was 4.3 months (95% CI, 2.8-5.6) versus 2.2 months (95% CI, 1.3-2.5), respectively.
A number of confirmatory trials are exploring Rova-T in other settings. The phase II, single-arm TRINITY study is testing the agent in the third- or later-line setting in patients with ≥1% DLL3-expressing SCLC and has completed accrual (NCT02674568). The TAHOE trial is looking at second-line treatment with Rova-T versus topotecan in patients with high-DLL3-expressing SCLC (NCT03061812). Finally, the phase III placebo-controlled MERU trial is investigating maintenance Rova-T in patients with all levels of DLL3 expression (NCT03033511).
"Antibody-drug conjugates against DLL3 have shown promising clinical activity and we're awaiting additional clinical trials," said Gainor.
Immunotherapy agents are demonstrating mixed activity in the mesothelioma space, he said. The phase IIb, double-blind, placebo-controlled DETERMINE study randomly assigned 571 patients to the CTLA-4 inhibitor tremelimumab versus placebo in the second- or third-line setting. Data showed that tremelimumab was not linked to an OS benefit over placebo (HR, 0.92; 95% CI, 0.76-1.12; P = .41).4
"This was a disappointing result," Gainor said. "Unfortunately, despite going up against placebo, this study was negative. It looks like CTLA-4 blockade by itself is really inactive in mesothelioma."
However, in the mesothelioma cohort of the phase Ib KEYNOTE-028 trial, pembrolizumab was associated with a 20% ORR in patients with PD-L1-positive (≥1%) mesothelioma tumors.
Moreover, nivolumab alone and the combination of nivolumab with ipilimumab are being investigated in patients with unresectable disease in the randomized, noncomparative phase II Mesothelioma Anti-PD-1 Study 2 (MAPS-2) trial. Preliminary findings had showed that, of the first 108 eligible patients, the disease control rate (DCR) at 12 weeks was 44% for those who received nivolumab alone and 50% among those who received it in combination with ipilimumab. In the ITT population, the DCR was 39.7% with nivolumab alone and 51.6% with the combination.5
Longer-term follow-up is necessary to determine the optimal role of these agents in the mesothelioma paradigm, Gainor summarized.
"Single-agent PD-1 inhibitors and combinations of PD-1/CTLA-4 demonstrated promising activity in mesothelioma," he said. "Both pembrolizumab and nivolumab plus or minus ipilimumab are listed on the NCCN guidelines for mesothelioma, so these are being used in clinical practice. As in other disease areas, there is a crucial need for better predictive biomarkers for response."