NEW YORK – Gritstone Bio is exploring the activity of Granite and Slate, its individualized and off-the-shelf neoantigen immunotherapies, in combination with checkpoint inhibitors in molecularly defined solid tumor indications based on encouraging data from earlier trials that the company unveiled last week.
At the European Society for Medical Oncology Congress on Friday, Daniel Catenacci, assistant professor of medicine at the University of Chicago, reported data from a Phase I/II trial in which patients with advanced solid tumors received Gritstone's Granite individualized vaccine in combination with Bristol Myers Squibb's checkpoint inhibitors ipilimumab (Yervoy) and nivolumab (Opdivo) and highlighted the activity of the treatment in microsatellite-stable colorectal cancers. Simultaneously, Gritstone released data from another Phase I/II trial, in which patients with different types of advanced solid tumors received the Slate off-the-shelf vaccine with ipilimumab-nivolumab and spotlighted the response in KRAS-mutated tumors.
Based on the efficacy signals seen in these earlier trials, the Emeryville, California-based company said it will begin two randomized trials next year involving Granite in MSS colorectal cancer: a Phase II/III trial exploring the activity of the personalized vaccine in combination with immunotherapy as a maintenance treatment and a Phase II trial in the adjuvant setting involving patients who have circulating tumor DNA after surgery and are at high risk for relapse. The company also announced that it has dosed the first patient in a Phase II trial exploring the activity of a second version of the Slate off-the-shelf vaccine in 60 patients with KRAS-mutated tumors: NSCLC patients who have progressed on immunotherapy and chemotherapy, MSS colorectal cancer patients on first-line therapy, and MSS colorectal cancer patients on their third line of treatment.
The Granite trials in MSS colorectal cancer are slated to start in the first half of 2022 and readout sometime in 2023. The ongoing Phase II Slate trial in KRAS-mutated NSCLC and MSS colorectal cancer will readout by mid-2022.
Waking up cold tumors
To date, checkpoint inhibitors have primarily benefited cancer patients who have neoantigen-specific CD8-positive T cells, high tumor mutation burden, or upregulated PD-L1 expression. For example, nivolumab alone or in combination with ipilimumab is approved for previously treated, advanced colorectal cancer patients with high microsatellite instability, or MSI-high, which causes cancer cells to accumulate tumor mutations. Merck's checkpoint inhibitor pembrolizumab (Keytruda) is approved as a first-line treatment for metastatic colorectal cancer patients with MSI-high tumors. Pembrolizumab is also approved for treatment-refractory patients with any kind of advanced solid tumor and a tumor mutation burden of at least 10 mutations per mega base.
The majority of gastrointestinal tumors, however, tend not to have high tumor mutation burden or PD-L1 expression. Moreover, only 5 percent of metastatic colorectal cancers are MSI-high, while 95 percent are microsatellite stable, or "cold" tumors, that cannot mount an immune response, even when exposed to checkpoint inhibitors.
"Unfortunately, the very common solid tumors that kill most people often are cold tumors," Gritstone CEO Andrew Allen said during a call last week with investors to discuss the company's development plans for the Granite and Slate vaccines. "If those patients are treated with a checkpoint inhibitor, little useful happens, apart from sometimes the development of autoimmunity."
Gritstone is developing Granite and Slate as vaccines that can wake up "cold" tumors. The company has designed the vaccines to deliver neoantigens that will induce CD8 T cells to travel to tumors and make them visible to the immune system. The immune system, primed to recognize and attack tumors, will then also respond to checkpoint inhibitors, the company is betting.
To develop the individualized Granite immunotherapy, the company uses its proprietary EDGE platform to predict the neoantigens that go into the vaccine. The machine-learning algorithms the platform applies to predict the neoantigens have been trained using tumor and normal DNA and tumor RNA. Patients provide formalin-fixed paraffin-embedded tissue samples for sequencing, as is standard for tumor profiling, and from this data the EDGE platform identifies 20 patient-specific neoantigens that Gritstone then inserts into viral and RNA vectors to create a personalized vaccine that is administered in combination with immunotherapy.
Allen noted that when developing the EDGE platform, the company wanted to ensure that the neoantigen prediction system could analyze data from FFPE samples and fit into the standard tumor molecular profiling workflow at most cancer centers. The company is using the EDGE platform to develop vaccines for infectious diseases and cancer. On Thursday, Gritstone announced it had raised $55 million in a private investment in public equity financing that it will put toward developing vaccines using the EDGE platform.
Meanwhile, the company's off-the-shelf Slate cancer vaccine includes predefined neoantigens that traffic T cells to tumor cells driven by specific mutations. The first version of Slate included neoantigens specific for p53 and KRAS mutations, but the company wasn't satisfied with the product's ability to induce T-cell responses. According to Allen, Gritstone discovered p53 was immunodominant and was trafficking T cells to tumors that weren't driven by p53 mutations.
"We learned that p53 … was probably crushing the magnitude of the KRAS-specific T-cell responses," Allen told investors. Based on this, Gritstone removed p53 from Slate v2, so it now only includes KRAS mutation-specific neoantigens, and this version has shown immunogenic superiority over the prior version in human HLA-transgenic mice.
Focusing on molecular response
At the ESMO Congress, University of Chicago's Catenacci presented data from a Phase I/II trial of Granite plus ipilimumab-nivolumab in advanced solid tumors. Out of 22 efficacy evaluable patients in the study, one gastroesophageal cancer patient had a complete response according to RECIST criteria and is still responding to the vaccine, while five patients had stable disease. However, Catenacci reflected that objective response rate assessed by RECIST criteria, which is essentially a measurement of tumor size, may not reflect the vaccine's potential to benefit patient's long term, because it takes time for the vaccine to induce CD8 T cells to infiltrate tumors and instigate an immune response.
As such, researchers looked at whether the vaccine with ipilimumab-nivolumab reduced ctDNA levels from baseline by at least 50 percent and were particularly encouraged to find a 44 percent molecular response rate in nine MSS colorectal cancer patients. In the four colorectal cancer patients who had a ctDNA response, the median overall survival was not reached after 17 months compared to a median overall survival of 7.8 months in those who did not have a molecular response.
At the meeting, Catenacci highlighted that the regimen produced neoantigen-specific T cells in tumors that previously lacked it. Further, molecular responses were seen despite the fact that researchers did not enrich the trial with patients who had biomarkers that would make them more likely to respond to immunotherapy. In fact, MSS colorectal cancer patients in the study had low PD-L1 expression and low tumor mutation burden.
After discussing this data and registrational plans for Granite with the US Food and Drug Administration, Gritstone announced last week that it will conduct a Phase II/III trial to explore the individualized vaccine as a maintenance treatment for newly diagnosed MSS metastatic colorectal cancer patients. The company plans to screen all colorectal cancer patients for MSI-high status and enroll those who are negative. Patients will first have induction therapy with FOLFOX and bevacizumab (Genentech's Avastin), during which time Gritstone will sequence patients and use its EDGE platform to predict the neoantigens it will use to develop a personalized vaccine.
Allen estimated that its platform will be able to predict neoantigens for 60 percent of patients who will then be randomized to either a personalized vaccine plus 5FU-bevacizumab or just 5FU-bevacizumab. In the open-label Phase II portion, researchers will track patients' ctDNA levels on treatment from the start of the trial. In the pivotal Phase III portion, the primary endpoint is progression-free survival as determined by independent reviewers.
Gritstone will also conduct a randomized Phase II trial, in which Granite will be given as adjuvant treatment to stage II/III MSS colorectal cancer patients who are at high risk of relapse after surgical tumor resection due to the presence of ctDNA. Up to 30 percent of stage II/III colorectal cancer patients aren't cured by surgery and will relapse, and these patients can be identified early using liquid biopsy testing, Allen said, because they will have persistent ctDNA in plasma.
In the planned Phase II trial in the adjuvant setting, MSS colorectal cancer patients who have ctDNA in blood tests after surgery will be randomized to receive adjuvant chemotherapy, and Granite plus immunotherapy as maintenance treatment or given adjuvant chemo and closely followed. The primary endpoints are molecular response according to reduction in ctDNA levels and disease-free survival.
Since most patients with ctDNA after surgery are expected to relapse within 12 to 14 months, this trial is "efficient" in that "everybody in the trial gives an endpoint," Allen said. "It's also an important trial because these patients need help. They are going to recur and standard-of-care [treatment] is unlikely to be sufficient."
During the call with investors, Allen also shared data from a Phase I/II trial, in which Gritstone treated 26 solid tumor patients with Slate v1 plus ipilimumab-nivolumab who had previously progressed on anti-PD-1 treatment. The efficacy of Slate plus immunotherapy was particularly promising in six NSCLC patients in whom HLA A*02:01 peptides were presenting KRAS G12C mutations on tumor cells.
Based on this experience, the company is conducting a Phase II trial of Slate v2 plus ipilimumab-nivolumab in KRAS-mutated advanced NSCLC and colorectal cancer patients. "You have several different KRAS mutations within the vaccine, but it's not enough to have these mutations. You do need to have the relevant HLA Class I allele to present that mutation," Allen explained to investors.
In the recently launched Phase II trial investigating Slate v2, "patients are screened for the presence of the appropriate pair: KRAS mutation plus the relevant HLA allele," he added. The company announced last week that it had dosed the first patient in this study.
Allen estimated that testing for these biomarkers could identify up to 15 percent of patients with lung adenocarcinomas and around the same proportion of colorectal cancer patients who are eligible for the Slate vaccine.
Although in both the Granite and Slate trials, Gritstone is measuring the efficacy of its vaccines using standard response endpoints (i.e., tumor shrinkage by RECIST criteria), the company is also interested in characterizing therapeutic benefit using serial ctDNA and other imaging measurements. Allen estimated that after receiving a neoantigen vaccine, it can take between eight and 16 weeks for patients to mount a T-cell response, and therefore noted that it makes sense to explore the activity of its vaccines in earlier disease settings, such as adjuvant MSS colorectal cancer, while patients are healthier and have time to mount an immune response.
In Gritstone's vaccine trials to date, high baseline ctDNA has shown to be a reliable predictor of who will not benefit from its neoantigen therapies. "The obvious negative prognostic biomarker is a very high ctDNA at baseline," Allen said, adding that if baseline ctDNA levels are "greater than 500 copies/ml … you're probably going to progress extremely quickly, probably before our therapy has a chance to help you."
As such, Allen agreed with the University of Chicago's Catenacci that lack of initial tumor shrinkage to the Gritstone vaccines doesn't necessarily mean the treatment won't ultimately benefit patients by allowing them to live longer. For example, Immunocore looked at the activity of its T-cell receptor bispecific immunotherapy tebentafusp in a subset of cancer patients who had progressive disease according to RECIST criteria in a Phase III trial and reported that those on the tebentafusp lived longer than patients receiving the investigators' choice of treatment. The company has also reported data that suggests tracking reductions in ctDNA levels in patients with progressive disease on the TCR immunotherapy may predict which patients will likely derive an overall survival benefit.
"We need to open our minds to the fact that with T-cell infiltration into lesions, shrinkage may not be what we see," Allen said, adding that Gritstone is taking a novel approach by treating solid tumors with neoantigen vaccines and it isn't yet clear "how efficacy will manifest. … Molecular response may be particularly important here."