BARCELONA – Attempting to limit adverse events with the investigational cyclin-dependent kinase (CDK) 4/6 inhibitor trilaciclib in triple-negative breast cancer failed to do what investigators theorized, missing both primary and secondary endpoints, researchers reported here.
Nevertheless, treatment was still associated with improved overall survival relative to a standard chemotherapy regimen.
Median overall survival was 12.6 months for women who were treated with gemcitabine plus carboplatin vs 20.1 months for those who had trilaciclib added twice during each 21-day treatment cycle with the chemotherapy doublet (P=0.0023), and 17.8 months for those who received trilaciclib four times per cycle (P=0.028), reported Joyce O'Shaugnessy, MD, chair of breast cancer prevention research at Baylor-Sammons Cancer Center/Texas Oncology in Dallas.
Taking all the trilaciclib patients together and comparing them still showed a nominal significant effect compared with chemotherapy (P=0.0015), she said at the .
"We can only say that these findings for overall survival are hypothesis generating," O'Shaugnessy told ľֱ. "We did not achieve our primary endpoint, so we have to consider the P values as nominal. It makes us believe that we should pursue treatment with trilaciclib."
The aim of the study was to see if trilaciclib could prevent adverse events caused by gemcitabine by preventing the myelosuppression that follows treatment with the gemcitabine-carboplatin combination. At her oral presentation, O'Shaugnessy suggested that chemotherapy-induced cell toxicity might limit the host immune system response against triple-negative breast cancer.
But things didn't work out that way: "The idea was not to target the breast cancer itself, but to use trilaciclib to put the bone marrow to rest. Trilaciclib like other CDK 4/6 inhibitors puts other cells to rest. We gave this intravenous drug along with gemcitabine-carboplatin with the idea that we could help preserve blood marrow and reduce severe neutropenia in the first cycle and reduce neutropenia that lasted 3 days or longer. Our second endpoint was to reduce the number of patients who got Grade 4 neutropenia at any time during the trial."
For the study, she and her colleagues randomized 102 patients into three treatment groups:
- Conventional gemcitabine/carboplatin chemotherapy administered on days 1 and 8
- The same chemotherapy regimen plus trilaciclib
- Single-agent trilaciclib on days 1 and 8, followed on days 2 and 9 by the trilaciclib/chemotherapy combination
The goal was to reduce marrow myelosuppression so patients could stay on the treatment longer, so more drug could be used to attack the cancer, O'Shaugnessy explained. "Unfortunately, the two primary endpoints were negative."
But there were hints that something important was happening, she continued: "Patients who didn't receive trilaciclib were able to handle four cycles of chemotherapy, and those who received one of the two trilaciclib administration schedules were about to handle seven or eight cycles of gemcitabine-carboplatin, which translated to receiving 50% more chemotherapy in the trilaciclib arms. But there were no differences between the groups in side effects."
The median treatment duration was 3.3 months with chemotherapy alone, vs 5.3 and 5.5 months in the two groups who received trilaciclib. The median cumulative doses of gemcitabine and carboplatin also increased with trilaciclib, O'Shaugnessy said.
"To me, that suggests some degree of bone marrow protection, because we were able to give more chemotherapy without having any statistically greater adverse events," she added.
"We looked at response rates, and they were not different between the arms. Then we looked at progression-free survival, and again there was no statistical difference between trilaciclib and not delivering the CDK 4/6 inhibitor, although there were numerical differences that favored trilaciclib."
'Surprise Came With Survival'
"Then the surprise came with survival," O'Shaugnessy said. "This was completely unexpected. Both trilaciclib arms had a significant improvement in overall survival, and it was not because the gemcitabine-carboplatin arm underperformed – 12 months is about what you would expect in triple-negative breast cancer."
She speculated that somehow the trilaciclib may have activated the immune response of the patients, thus attacking the cancer. "This could also be chance, because we only had about 30-35 patients in each arm. It could also be that the longer duration of chemotherapy with trilaciclib could have had a positive impact on survival."
There is more to the story, O'Shaugnessy said: "There is definitely interest in going ahead with a Phase III trial with gemcitabine and carboplatin plus or minus trilaciclib."
The study's discussant, Mafalda Oliveira, MD, PhD, of Vall d'Hebron Institute of Oncology in Barcelona, said: "This was a 'negative' trial with clinically 'positive' results, with an improved toxicity profile and an overall survival benefit."
"Although the final overall survival analysis is pending, a greater than 5 month gain in overall survival warrants further investigation," she said.
Disclosures
The study was sponsored by G1 Therapeutics, Inc.
O'Shaugnessy disclosed relevant relationships with AbbVie, Agendia, Amgen, AstraZeneca, Bristol-Myers Squibb, Celgene Corporation, Eisai, G1 Therapeutics, Genentech, Genomic Health, GRAIL, Immunomedics, Heron Therapeutics, Ipsen Biopharmaceuticals, Jounce Therapeutics, Lilly, Merck, Myriad, Novartis, Odonate Therapeutics, Pfizer, Puma Biotechnology, Prime Oncology, Roche, Seattle Genetics, and Syndax Pharmaceuticals.
Oliveira disclosed relevant relationships with Roche, GSK, Puma Biotechnology, AstraZeneca, Philips Healthcare, Genentech, Immunomedics, Seattle Genetics, Boehringer-Ingelheim, Pierre-Fabre, GP Pharma, and Grunenthal.
Primary Source
European Society for Medical Oncology
O'Shaugnessy J, et al "Trilaciclib improves overall survival when given with gemcitabine/carboplatin in patients with metastatic triple negative breast cancer in a randomized phase II trial" ESMO 2019; Abstract LBA22.