NEW YORK – Agendia is getting closer to commercially launching a 53-gene signature, dubbed ImPrint, to help guide immunotherapy treatment for early-stage breast cancer patients.
The company's commercialization plans for the test, currently available for research use, were recently bolstered by preliminary validation data on the signature's ability to predict which patients with high-risk, HER2-negative early-stage breast cancer respond to neoadjuvant PD-1 immune checkpoint inhibition.
The sensitivity and specificity data are from a cohort of 69 patients enrolled in the I-SPY2 trial. In that adaptive trial, high-risk, luminal- or basal-type early-stage breast cancer patients received four cycles of Merck's Keytruda (pembrolizumab) in the neoadjuvant treatment setting.
Based on whole-transcriptome microarray data from tissue biopsies taken before patients underwent surgery, the 53-gene signature had more than 90 percent sensitivity and over 80 percent specificity for predicting which patients experienced pathologic complete response — defined as no residual invasive cancer in breasts or nodes — with neoadjuvant Keytruda.
When patients were stratified according to hormone receptor status, the ImPrint signature had over 95 percent sensitivity and more than 70 percent specificity in predicting neoadjuvant Keytruda response for triple-negative breast cancer patients, and over 80 percent sensitivity and 85 percent specificity for hormone-positive, HER2-negative patients.
Agendia Chief Medical Officer William Audeh highlighted as a key performance metric that ImPrint had a 77 percent positive predictive value among the hormone receptor-positive subgroup. "That's a group where it's been very difficult to show any real benefit of immunotherapy," he said, adding that it will be important to focus further validation studies specifically in this hormone receptor-positive group, and that this is already planned through additional treatment arms in the next iteration of the trial, dubbed I-SPY2.2.
If the signature holds up in larger cohorts of hormone receptor-positive patients, the implications could be significant, because unlike early-stage triple-negative breast cancer — for which a subset of patients already have neoadjuvant Keytruda as a US Food and Drug Administration-approved option for about a year — there are no anti-PD-1 treatments approved for early-stage hormone receptor-positive, HER2-negative breast cancer patients.
Accordingly, should ImPrint advance toward routine, commercial use, Audeh said that Agendia will likely position the test as a way to identify hormone receptor-positive breast cancer patients who are eligible for immunotherapy.
"I wouldn't look to use this to exclude immunotherapy from triple-negative patients, but rather to add immunotherapy to the treatment of hormone receptor-positive, and perhaps HER2-positive patients, in the right setting," he said. "The main value of the ImPrint signature is to expand the population for whom immunotherapy may be beneficial. … As we all know, there are hormone receptor-positive breast cancers that don't respond very well to either endocrine therapy or chemotherapy."
The use of the test in this way, of course, is contingent on the results of additional validation studies, since the hormone receptor-positive patient numbers were quite small in the I-SPY2 dataset. Only 12 of the patients who experienced a pathologic complete response to neoadjuvant Keytruda had hormone receptor-positive cancers.
If indeed further validation studies confirm the ImPrint signature's predictive performance in this small dataset, Agendia plans to launch ImPrint as a commercially available test. Audeh was not willing to share a timeline for a commercial launch, however, given the early nature of the findings.
"There are still a lot of questions that remain," Rebecca Dent, head of the department of medical oncology at the National Cancer Center Singapore, acknowledged at the American Society of Clinical Oncology's annual meeting earlier this month, where Agendia shared the ImPrint validation data. "[But] I'm actually quite optimistic that we're getting close to identifying patients who might benefit from immune checkpoint inhibitors in the early breast cancer setting, with the caveat that the numbers are small."
Next steps, commercialization plans
According to Audeh, the next steps involve validating the signature prospectively within new arms of the I-SPY2 trial, while also retrospectively validating the test within already-conducted immunotherapy trials. The latter approach will be the "most rapid way to get new information validating the signature," he said, though prospective validation provides greater assurance of the test's real-world performance.
Additionally, Agendia plans to tap into a growing database of patients in its real-world FLEX registry to determine how prevalent this ImPrint signature is among the early breast cancer patient population. The firm has already used the registry to establish the real-world performance of its MammaPrint and BluePrint signatures.
"We have whole-transcriptome data on over 10,000 such patients now in that database," Audeh said. "It's natural for us to apply the 53 genes in the immune response signature … to learn more about the prevalence of the immune-positive state, even in women who were not treated with immunotherapy."
The signature is currently still for research use only, but if validation does end up supporting its commercial use, Audeh said the ImPrint signature would fit easily into the workflow that Agendia has established for MammaPrint and BluePrint. The 70-gene, FDA-cleared MammaPrint test is already used on a commercial scale to help determine whether early-stage breast cancer patients are at high risk of recurrence and need chemotherapy after surgery. The 80-gene BluePrint test, often used in combination with MammaPrint, helps oncologists refine patients' breast cancer subtype, grouping them into luminal-type, HER2-type, or basal-type and, in turn, informing prognosis.
These signatures for guiding breast cancer treatment use one gene expression platform, which Agendia could also use to assess patients' ImPrint signature status.
"We obtain the 70 genes from MammaPrint and the 80 genes from BluePrint at the same time on the same run … so we anticipate that these would all be run at the same time on the diagnostic specimen to really allow for the most precise treatment planning," he said, adding that Agendia isn't looking at ImPrint so much as a standalone test but as a component of a comprehensive gene-expression profiling platform used in tandem with MammaPrint and BluePrint.
The patient population in I-SPY2 that served as the initial validation cohort for ImPrint had already undergone BluePrint and MammaPrint testing, reflecting the likely scenario in a real-world setting. If a patient with hormone receptor-positive breast cancer has a low recurrence risk according to MammaPrint, they are already unlikely to benefit from neoadjuvant chemo or targeted therapy, which is why only MammaPrint high-risk hormone receptor-positive patients went on to receive ImPrint testing.
In the future, Audeh envisions a scenario in which the ImPrint signature is offered as a test alongside the others, as a comprehensive suite to inform the best possible treatment for early-stage breast cancer.
"We at Agendia have always based our approach to breast cancer on gene expression rather than mutation analysis or protein staining," he said. "This really fits into our [goal] of correlating a gene expression pattern with a clinical outcome of interest, and we're going to continue to develop new signatures in this same manner to improve the precision of therapy selection in breast cancer."
Down the line, Audeh added, Agendia also plans to validate the ImPrint signature's ability to predict response to neoadjuvant therapies other than Keytruda. These might include PD-L1 inhibitors, rather than anti-PD-1 therapies. Already, Agendia has run the signature on samples from patients in I-SPY2 who received the combination of AstraZeneca's PD-L1 inhibitor Imfinzi (durvalumab) and Merck/AstraZeneca's PARP inhibitor Lynparza (olaparib).
Researchers will need to find a way to separate the effects of the two agents to evaluate the signature's predictive ability for Imfinzi, but Audeh is optimistic. "We think that this has the potential to be more universal in its application to immunotherapies," he said.