NEW YORK – Autologous tumor-infiltrating lymphocytes selected to target patients' unique neoantigens shrunk metastatic breast tumors in a small but potentially impactful trial conducted at the National Cancer Institute.
The research, recently published in the Journal of Clinical Oncology, indicates that selected TIL therapy could benefit these heavily pretreated patients. The NCI is now expanding the trial to further explore these findings.
Autologous, selected TIL therapies involve a highly complex but personalized approach. They require resecting solid tumors, sequencing them, isolating TILs from tumor fragments, identifying the specific TILs that recognize bespoke tumor neoantigens, expanding ex vivo a product of TILs that have the identified, neoantigen-targeting T-cell receptor, then reinfusing them into the patient following lymphodepleting chemotherapy. These therapies can potentially be given in combination with immune checkpoint inhibitors.
"You can't get more personalized than this," explained Steven Rosenberg, chief of the surgery branch at the NCI and lead author on the recent JCO paper. "We create a new drug for every patient using their own peripheral lymphocytes, into which we've inserted the T-cell receptor to recognize their neoantigens."
According to Rosenberg, the pilot study results are particularly significant since breast cancers have not historically demonstrated durable responses to immunotherapy treatments. Breast cancer is far less immunogenic than cancers such as melanoma and smoking-related non-small cell lung cancer, in which autologous TIL treatments have led to durable responses in past studies.
Although just six metastatic breast cancer patients were ultimately treated in the NCI's trial — a basket study that also includes brain, colorectal, pancreatic, and ovarian cancer cohorts — over half of the 42 patients initially screened were found to have TILs that recognized at least one of the tumor-specific neoantigens on their tumors, which points to the potential feasibility of treating patients with this approach.
As described in the recently published JCO study, researchers screened treatment-refractory metastatic breast cancer patients by whole-genome sequencing their tumor tissue, isolating TILs from tumor fragments, and then analyzing the TIL-neoantigen interaction in vitro. The fact that 28 of the initial 42 screened patients, or 67 percent, had TILs that recognized their tumor-specific neoantigens, gives Rosenberg and his team the proof of concept they need to begin enrolling additional breast cancer patients to the ongoing Phase II study.
"On the basis of this pilot study, and the knowledge that the majority of breast cancer patients have T-cell immunity against the mutations in their cancer, we've now begun to increase the accrual of those patients with breast cancer to [TIL] studies so we can find out just how good it is and what modifications we need to make it even better," said Rosenberg, who was involved in the trials for the first autologous CAR T-cell therapies that are available now for hematologic malignancies.
Thanks in part to a $33 million cell therapy center that has just opened at the NCI, the team can expand the trial significantly. Rosenberg expects to accrue an additional 30 to 40 metastatic breast cancer patients as a next step.
Patient responses
Although the NCI screened 42 patients for the trial, ultimately the selected and expanded TILs were reinfused into six patients who were still clinically eligible. These patients underwent a lymphodepleting regimen alongside Merck's immune checkpoint inhibitor Keytruda (pembrolizumab) in the hopes of increasing efficacy.
Out of the treated patients — of which four had hormone receptor-positive tumors and two had triple-negative cancer — one patient experienced a complete response and is still in remission 5.5 years later, and two patients experienced partial tumor regressions. One of the two patients with a partial response saw her tumor shrink by 69 percent and the other patient experienced a 52 percent tumor reduction.
Both patients relapsed at 10 months and 6 months, but the tumors were small and easily removed with surgery. "They … are now living disease-free as well," Rosenberg said, adding that he hopes that enrolling additional patients to the trial will allow for a more illustrative analysis of why some patients responded but others didn't.
"We've only treated six patients [but] we are going to expand considerably," Rosenberg said. "In the first 30 to 40 patients [treated], we will at least get an idea of the absolute frequency of their responses and their durability."
By analyzing each patient's response deeply, Rosenberg hopes to garner insights into why the one patient who had sustained a complete response for over five years had HR-positive disease, a tumor type not particularly responsive to checkpoint inhibitors. Past studies have found few, very short-term responders among HR-positive breast cancer patients. Because of this, Rosenberg and colleagues believe this patient's exceptional response was unlikely to be due to Keytruda, which she received with the TILs. Rosenberg's team hopes to tease out the relative contribution, if any, from additional checkpoint inhibitor treatment in the expanded phase of the trial.
Commercial directions, scalability
There are no US Food and Drug Administration-approved autologous TIL therapies on the market, even though in clinical trials some melanoma patients have experienced durable responses to unselected TILs, which are more simply harvested, expanded, and reinfused all together. While part of this lag is due to the challenges of defining the TIL product for regulatory purposes and securing a reliable and accepted potency assay to ensure the treatment will be active against tumor cells once infused, there are also turnaround time and scalability challenges.
The selected TILs that patients received in the NCI trial took an average of 24 days to manufacture. During this time, many patients initially enrolled went on to receive other treatments. Some patients' cancers were too advanced, and they couldn't wait for the TILs to be engineered. Outside of the clinical trial setting, the lengthy manufacturing time as well as the scientific expertise needed to create these products stand to be significant access barriers.
In the case of FDA-approved CAR T-cell therapies like Novartis' Kymriah (tisa-cel) and Gilead's Yescarta (axi-cel) for lymphoma, these barriers have resulted in a steep price tag and deep disparities in patient access, which drugmakers say they are trying to address. But even with these hurdles, the fact that autologous CAR T-cell therapies reached the market at all is an encouraging sign for Rosenberg.
"When I was approached by pharmaceutical companies back in 2009 about [CAR T-cell therapy], I ran into the criticism that it was too complicated," he said, recalling the chain of events that brought Yescarta to market. After a former fellow with Rosenberg saw the treatment's potential and started Kite Pharma in 2011, the FDA approved Yescarta in 2017, and Gilead Sciences bought Kite for $11.9 billion that year.
"[Yescarta] is now being widely applied in patients with lymphomas and leukemias," Rosenberg said. "If you find something that works, the genius of American industry will find ways to do it, because people are dying due to metastatic cancer."