Stratification by PD-LI expression
Stratifying the patients by the absence or presence of programmed cell death–ligand-1 (PD-L1) expression, the team found that the performance of nivo+ipi was “the same” as that of chemotherapy, Bass said.
“But in cases where there is any expression of PD-L1, the experimental arm performs better,” at an overall survival 18.0 months vs 13.3 months for chemotherapy and a hazard ratio of 0.69, he said.
There was no difference between the two treatment arms in progression-free survival. Chemotherapy performed better in the first 6 months of treatment, after which the nivo+ipi arm had lower event rates.
Nivo+ipi was also associated with a greater duration of response, at a median of 11.0 months vs 6.7 months for standard-of-care chemotherapy.
Moreover, at 24 months, 32% of nivo+ipi patients were still experiencing a response, whereas 8% of those in the chemotherapy arm were.
Treatment-related adverse events rates were almost identical between the two treatment groups, although treatment with nivo+ipi was associated with more grade 3/4 serious treatment-related adverse events, at 15 vs six for chemotherapy.
Choosing immunotherapy vs. chemotherapy
In his discussion of the new study, Fennell compared the current results with those from two studies, INITIATE and MAPS2. “What’s very clear is the response rate is slightly higher,” as is the disease control rate, he said.
This, he explained, “is perhaps not surprising, given that these two previous trials were in the relapse setting.”
He pointed out, however, that the progression-free survival data from those previous trials were “not a million miles away” from results seen in CheckMate 743, “suggesting that this immunotherapy does have significant activity in the relapse setting.”
For Fennell, the “pivotal data” are in patients with nonepithelioid tumors, particularly inasmuch as chemotherapy performed “poorly” in this setting, whereas it performs “as expected” in epithelioid mesothelioma.
He believes that the driver for this is the poor prognosis associated with sarcomatoid biphasic disease, a subtype characterized by increased expression of vimentin and ZEB1, proteins both associated with EMT.
“What does this mean?” Fennell asked.
“If you have have enrichment of EMT, what you see is increased drug resistance, increased invasiveness, something we know well with sarcomatoid mesotheliomas in particular, and this drug-resistance phenotype may account for the drug resistance that we see in CheckMate 743 with chemotherapy.
“This does not appear, however, to impact in any way the efficacy of the immunotherapy,” he noted.
Fennell believes that, with regard to both efficacy and safety, the balance is “very much in favor” of nivo+ipi in epithelioid mesothelioma, although there is less to choose between immunotherapy and chemotherapy in the nonepithelioid setting.
Indeed, the choice is “possible tilting slightly towards chemotherapy” in patients with the nonepithelioid tumors, owing to the lower rates of grade 3/4 serious treatment-related adverse events in comparison with combination immunotherapy.
The study was supported by Bristol-Myers Squibb. Bass has served on the advisory boards of MSD, AstraZeneca, and Takeda. Fennell has received research support from AstraZeneca, Bristol-Myers Squibb, Clovis Oncology, Eli Lilly, MSD, and Roche; research funding from Astex Therapeutics, Bayer, and Boehringer Ingelheim; has served on the speaker bureau of AstraZeneca, Boehringer Ingelheim, and Roche; has acted as a consultant for Bayer and Lab 21; and has served on the advisory board of Atara Biotherapeutics, Boehringer Ingelheim, and Inventiva.
This article first appeared on Medscape.com.