As the market for immunotherapies grows, reaching over $10 billion already, the interest in biomarkers to guide prescribing has intensified. Doctors want better ways to predict who will benefit from the drugs. “We need to protect patients from unnecessary risk of side effects from drugs that can’t help them and to avoid spending on drugs that aren’t going to work,” said Timothy Chan, chair of the Immunotherapy Center at the Cleveland Clinic. “We also need to guide them to the best drugs for their cancer as quickly as possible.”
Promising leads are plenty. They include germline and tumor genetics, PD-L1 expression, tumor mutational burden (TMB), microsatellite instability (MSI), human leukocyte antigen (HLA), microbiome markers and more. But uncovering biomarkers for immunotherapies does not seem to be straightforward. Unlike the neat pairings of targeted therapies such as Herceptin for HER2 positive breast cancers, immunotherapy has been a much more complicated case.
The poster child for immunotherapy markers is Keytruda (pembrolizumab), which was approved in 2017 for the treatment of adult and pediatric patients with unresectable or metastatic solid tumors that are microsatellite instability-high (MSI-H) or mismatch repair deficient (dMMR). The drug had been previously approved for other indications, but this marked the first time the FDA approved a cancer treatment based on a common biomarker rather than the location in the body where the tumor originated.
“This was a huge shift,” said Heather Tomlinson, director of clinical diagnostics at Promega. “It was the first pan tumor approval.” MSI, she explained measures mismatch repair. “This is very significant for patients.”
Another model, and an early sucesses for immunotherapy, has been non-small cell lung cancer (NSCLC), which has benefitted notably from the inhibition of immune checkpoints programmed cell death protein-1 (PD-1) and programmed cell death ligand-1 (PD-L1). However, despite providing significant survival benefit for some patients with advanced NSCLC, the objective response rates (ORRs) are only 20% to 30%. NSCLC patients are currently selected for first-line treatment based on expression of PD-L1 in tumor cells. But the sensitivity and specificity of PD-L1 expression is modest, and this has prompted the search for additional biomarkers.
PD-L1 was an obvious early target, but scientists are starting to think that finding biomarkers to guide immunotherapy may require a new paradigm. Rather than matching a drug to a patient using a single biomarker for a mutation that drives the cancer, doctors may need to test for multiple markers that characterize the cancer’s sensitivity to immune attack. They may also have to evaluate these markers on a sliding scale.
“We need to start taking an approach that is not as simple as a single cut-off level,” said Chan. “Rather there will be a continuum, and we’ll need to determine where a particular patient falls on that.”
Another theme is turning “cold” tumors to “hot.” Some cancers contain few infiltrating T cells and are not recognized by—and do not provoke—a strong response by the immune system. That makes them more resistant to current immunotherapies. Such “immunologically cold” cancers include glioblastomas as well as ovarian, prostate, pancreatic, and most breast cancers.
In contrast, immunologically hot tumors contain high levels of infiltrating T cells and more antigens, making them more recognizable by the immune system and more likely to trigger a strong immune response.
Among the cancers considered to be “immunologically hot” are bladder, head and neck, kidney, melanoma, and non–small cell lung cancers. However, even within these immunologically hot cancers, only a minority of patients appear to benefit from immunotherapy.
“Anti-PD-1 and anti-PD-L1 antibodies have so far targeted primarily hot tumors,” said Ira Jacobs, asset team leader for early oncology development and clinical research at Pfizer. They work primarily by activating T cells which are supposed to target tumors but are being held in check. Markers to better distinguish these tumor types will thus be critical.
Emerging testing paradigm
Genomic testing is quickly powering up to meet this challenge. Illumina’s TruSight Oncology 500 (TSO 500) is a next-generation sequencing (NGS) assay for DNA and RNA, which targets 523 genes for small variants, TMB, MSI, splice variants, and fusions. “Part of the beauty of the TSO 500 assay is that you can read all the biomarkers out on one comprehensive NGS panel,” said Kevin Keegan, senior director of oncology marketing at Illumina. The test can be used on either a tissue sample or a liquid biopsy using the TSO 500 ctDNA.
As liquid biopsy emerges as an important complimentary testing method to tissue biopsies, their applications are also growing. In the case of TSO 500 ctDNA, a liquid biopsy sample can overcome two challenges experienced with tissue tests when assessing the patient’s eligibility for targeted or IO therapy. First, it enables non-invasive sampling for patients, and second, it enables clinicians to overcome the issue of sampling bias with respect to tumor heterogeneity. Liquid biopsy is also emerging as the sample method of choice in the monitoring space, an area that Illumina is working with in-vitro diagnostic (IVD) development partners to enable.
The company is currently selling TSO 500 for research use only around the world, and clinical labs can validate the test for use as a laboratory developed test (LDT). “Intermountain Healthcare and several of our other customers have done that,” Keegan said. “They have a centralized genomics lab that runs the test and they use it to manage patient care.” Illumina is also working on getting an approval for an IVD version of the test. Over time, they will expand the number of companion diagnostic claims for that test.
Promega is developing its microsatellite instability (MSI) technology as an on-label, solid tumor companion diagnostic (CDx) to Merck’s blockbuster cancer immunotherapy Keytruda (pembrolizumab), through a collaboration. MSI testing measures genomic accumulation of insertion or deletion (INDEL) errors caused by a deficient mismatch-repair system (dMMR) that occurs in certain types of solid tumors. According to the company, such screening may be used to better characterize tumors and guide therapeutic choices for MSI-High cancer types. Promega’s MSI CDx has already received approval in Japan, and the company intends to seek regulatory clearance in the U.S., as well as in China and Europe.
Tumors with MSI-High status, it’s believed, are more susceptible to immunotherapies because of mutation-associated neoantigens that are believed to cause immune cell infiltration into the tumor microenvironment. Promega MSI technology uses five monomorphic mononucleotides, reflecting a recommendation of the NIH’s National Cancer Institute. The company will initially seek regulatory approvals for the Promega MSI CDx in the U.S. and China, with possible plans to add approvals in additional areas of the world.
“There is still so much we have to learn about MSI,” noted Promega’s Thomlinson. “There are lots of clinical trials ongoing, particularly with liquid biopsies.”
Looking forward
Scientists have not given up on finding specific biomarker for immunotherapy in particular tumors. There is evidence, for example, that elevated Δ133TP53β is a marker for prostate cancers with an increased proliferative index, high immune cell infiltrate, and an immunosuppressive phenotype.
But I-Ming Wang, director of Translational Oncology at Pfizer points out more attention now is going toward understanding “pan” and more easily accessible peripheral biomarkers, and how to evaluate them to best guide treatment.
Encouraging more comprehensive sequencing of cancer patients is another goal. “Only about 20% of cancer patients are getting fully genotyped,” said Rebecca Nagy, senior director in Guardant Health’s Medical Affairs group. Usually it is either because physicians have a hard time keeping up with guidelines, difficulty accessing testing, or cost concerns. “There is a trend of physicians just giving immunotherapy alone, or with chemotherapy, without doing comprehensive testing,” she said. “That is a disservice for the patients, since it is just as important to find the ones who will not respond.”
“One thing that has limited gene testing is how long it takes to get results,” said Nagy. A traditional biopsy can take up to four weeks. “The turnabout for our test takes 7 days and you get all the biomarkers in one test,” she added. Guardant offers the liquid biopsy-based Guardant360 and GuardantOMNI tests for advanced stage cancer patients and LUNAR assay for research use. Nagy said that researchers are also looking for ways to use blood-based tests like Guardant’s much earlier, even in people who do not yet have cancer symptoms, but who have a long history of smoking or a family history of cancer.
Another goal is learning how to combine immunotherapy with other cancer therapies, including radiation therapy and chemotherapy, to stimulate a response that triggers the immune system to elicit a more effective response in patients with a variety of tumor types. “We combine the immunotherapy with other modalities to improve response rate,” said Wang. “There could be multiple modes of action involved in the combination treatments including turning cold tumors into hot ones.”
“People are coming up with models to test combination strategies. The challenge is still to be able to go from the preclinical models, to the clinic. We definitely need better pre-clinical models,” Jacobs agreed. “There will be double or triple combinations. We need to find out how to do that.”
The field is also being advanced by continuing work to map the cancer genome. “It helps if you have a roadmap,” says Josh Stuart, of UC Santa Cruz. Stuart was part of one of the teams that recently created Pan-Cancer Atlas, which characterized 33 different types of cancer from more than 10,000 patients (see “Cancer Gets a Global Genomic Map,” page 12). This is the most comprehensive cross-cancer analysis to date and is also the final output of The Cancer Genome Atlas (TCGA) program of the National Cancer Institute and the National Human Genome Research Institute (NHGRI).
The genome, Stuart explained, “Is a book that keeps on being written.” With this new data, it will be easier to determine which mutations come first, which they are followed by, and how they all may be interconnected. “It’s like a molecular clock,” he says. “Now you can order the events and see which mutation comes before an amplification.” This type of information will help further inform the search for new markers to guide immunotherapy.