DetermaIO Assay Delineates Single-Agent Immunotherapy Benefit in NSCLC

David L. Saltman MD, PhD

Patients with advanced non–small cell lung cancer (NSCLC) who score positively on the Determa immune-oncology (IO) assay are more likely to experience survival benefits with frontline immunotherapy than negative responders, according to David L. Saltman MD, PhD.

The 27-gene DetermaIO assay is designed to assess tumor, inflammatory effector cell, and surrounding cancer-associated fibroblasts in the tumor microenvironment. In 2 cohort studies that were presented at the 2022 SITC Annual Meeting, a positive DetermaIO score corresponded with progression-free survival (PFS) and overall survival in patients with advanced NSCLC who received first-line immune checkpoint inhibitor monotherapy, including those with an ECOG performance status (PS) of 2. These studies found that DetermaIO provided important supplementary information to PD-L1 immunohistochemistry (IHC) and identified patients who derived deeper benefit from single-agent immune checkpoint inhibition vs combination therapy, regardless of their ECOG PS.^1,2

“[Immunotherapies] are expensive treatments that are associated with severe, potentially life-threatening adverse effects [AEs], so we want to select the right patients, especially when [treating a] group with compromised PS,” Saltman said.

in an interview with OncLive^®, Saltman discussed the unique features of the DetermaIO assay, noted the assay’s applications beyond lung cancer, and emphasized that biomarker combinations potentially have greater predictive benefits than single biomarkers. Saltman is a medical oncologist at BC Cancer Agency in British Columbia, Canada.

OncLive^®: What are the advantages and limitations of existing checkpoint inhibition biomarkers in lung cancer?

Saltman: Immunotherapy is a huge advance in the treatment of patients with lung cancer and many other types of cancer. [It has benefits in the] advanced setting, the adjuvant setting, and, excitingly, in the neoadjuvant or preoperative setting.

Unfortunately, most patients who are treated with immune checkpoint inhibitors don’t respond, so predictive biomarkers are important when trying to determine who’s more likely to respond. These drugs are expensive and are associated with severe, sometimes life-threatening AEs, so it’s important to know who’s going to benefit and who will not benefit to spare them the cost and the potential toxicities.

Currently, 1 widely used biomarker is PD-L1, which is [detected through] an IHC test. Its advantages are that it’s relatively inexpensive and pathology labs are quite familiar with it. It’s been used in many clinical trials testing immune checkpoint inhibitors.

The disadvantages [with PD-L1] are that different scoring systems and antibodies are currently in use, which complicates [our practice]. The other problem is that although patients who are PD-L1 positive are more likely to respond to immune checkpoint inhibitors, a significant number of patients don’t [respond]. [Since patients] who are PD-L1 negative may still respond, it’s not an ideal biomarker.

The other biomarker that’s widely used is tumor mutational burden [TMB], which is approved for many histologies and different cancers. A disadvantage [with TMB] is the cost to acquire whole-exome sequencing or, more frequently, next-generation sequencing [NGS]. NGS is usually available at academic centers, and many commercial enterprises provide it as well.

[Along with the cost, there is the question of] interpretation. Is high TMB greater than 10 mutations per megabase? Is it [greater than] 20? There are different scoring systems, which can be confusing. For community oncology centers, these can be expensive tests compared with PD-L1 IHC.

What served as the impetus for the 2 DetermaIO studies you led?

[We aimed] to see if we could find another biomarker that could either be useful on its own or added to existing biomarkers, such as PD-L1. I also have an interest in lung cancer, and at the BC Cancer Agency, we have many patient and archival samples that we can use for this type of retrospective study.

What differentiates the DetermaIO assay from traditional biomarker assessment?

The DetermaIO test is used on formalin-fixed, paraffin-embedded [FFPE] tumor biopsy samples. It measures the presence of immune cells and inflammatory cells subtypes and looks for the presence or absence of stromal microenvironment. This is a 27-gene expression assay, and the data that we get from it are put into an algorithm that is used to determine if a patient is IO positive and therefore more likely to respond to immune checkpoint inhibitors, or IO negative and therefore less likely to respond.

The advantage of this test is it can either be done on existing RNA sequence data, or on FFPE tumor samples that are readily available, and it can be tested with a simple reverse transcription-polymerase chain reaction [RT-PCR]. This test is CLIA [Clinical Laboratory Improvement Amendments] lab approved and is now available in the United States.

What predictive benefits of DetermaIO did these studies reveal, and what are the implications of this research in tumor types other than lung cancer?

The results did not come as a surprise because the DetermaIO assay has been previously tested in a smaller cohort of patients with NSCLC at The West Clinic in Memphis, Tennessee.

It has also been tested in many other tumor types. Interestingly, this test appears to be predictive of response to immune checkpoint inhibitors in many different histologies and settings. For example, it was initially tested in breast cancer and has been useful in predicting pathological complete responses in the neoadjuvant treatment of patients with breast cancer with immunotherapy. It’s been studied in advanced metastatic bladder cancer, and it’s been studied by an Italian group in metastatic colorectal cancer in patients who have not been selected for their mismatch repair [MMR] gene status. That was an exciting study.

Our study was not a large study. One of the problems has been trying to find archival samples that have not exhausted their tumor content or their RNA. We screened many patients for our study and were eventually able to use 147 samples.

The results were consistent with the previous smaller study done at The West Clinic and suggest that the IO test by itself can predict PFS. Interestingly, combining the DetermaIO test with the PD-L1 test in patients who have a PD-L1 score greater than 50% is more predictive than either the PD-L1 or DetermaIO tests by themselves. The utility of this test may be either on its own or in combination with PD-L1 or other biomarkers.

How might these data affect standard practice or parameters around clinical trial eligibility?

[We recruited] patients who had an ECOG PS of 0 to 2. The small number of patients in the study who had an ECOG PS of 2 were [more likely to be IO positive] and more likely to respond and have a better PFS than the patients who were IO negative. This was exciting because possibly up to 40% of patients with NSCLC may have an ECOG PS of 2, and historical data from prospective and retrospective studies have suggested that patients with an ECOG PS of 2 or higher are less likely to respond [to immunotherapy], so these patients are excluded from clinical trials.

The [phase 3] IPSOS trial [NCT03191786] suggested that selected patients with an ECOG PS of 2 [or higher] or comorbidities [including older] age may respond [to immunotherapy]. The DetermaIO test may be helpful for sorting out which of those patients with an ECOG PS of 2 may respond to these treatments.

What are the next steps for this research?

We would like to explore the findings in the group with an ECOG PS of 2 and validate those with a larger study. The difficulty is trying to access tumor samples that have adequate nucleic acid content so we can do these studies. It will be important to try to validate that finding either in our group or other groups.

We’d also like to see the results of larger studies with other tumor histologies. Oncocyte, the company that has developed this assay, is going to [participate] in a large SWOG trial with triple-negative breast cancer in the neoadjuvant setting. It will be exciting to see if previous results can be validated in a larger prospective trial. That’s exciting for the assay.

What main message would you like to leave with colleagues about this assay and the importance of biomarker testing?

We need more and better predictive biomarkers. Using more than 1 biomarker may be more useful in predicting responses to immune checkpoint inhibitors than a single biomarker by itself.

[Additionally,] we need to look more closely at the group of patients with an ECOG PS of 2. In both clinical trials and our practice, we’ve been excluding these patients from [receiving] immune checkpoint inhibitors. However, it’s possible that predictive biomarkers such as DetermaIO can help determine in which patients these agents could be used and which patients are more likely to respond [to immunotherapy].

References

1. Saltman DL, Croteau NS, Lockyer HM, et al. Use of a 27-gene immune-oncology (IO) assay to associate response to single-agent immune checkpoint inhibitor (ICI) therapy in advanced-stage NSCLC patients from a Canadian cohort. Presented at: 2022 SITC Annual Meeting; November 8-12, 2022; Boston, MA. Abstract 59
2. Saltman DL, Seitz RS, Nielsen T, et al. DetermaIO (IO score) is associated with efficacy of ICI monotherapy in advanced NSCLC patients with ECOG performance status 2. Presented at: 2022 SITC Annual Meeting; November 8-12, 2022; Boston, MA. Abstract 101