NEW YORK – Researchers have flagged acquired mutations in human leukocyte antigens (HLA), KEAP1, and STK11, along with changes in the tumor immune environment after treatment, that contribute to non-small cell lung cancer patients becoming resistant to immunotherapy.
Researchers used comprehensive tumor genomic profiling, machine learning-based assessment of tumor-infiltrating lymphocytes, multiplexed immunofluorescence, and HLA-class I immunohistochemistry testing to analyze matched tumor biopsies from NSCLC patients before and after immune checkpoint inhibitor treatment and published their findings in the Journal of Clinical Oncology last month.
The proportion of NSCLC patients who develop acquired resistance to immunotherapy is not clear, but investigators have attempted to estimate it by analyzing the proportion of patients who relapsed after experiencing an initial response to treatment. In one of these estimates, researchers looked at pooled data from several trials of Bristol Myers Squibb's checkpoint inhibitor Opdivo (nivolumab) in NSCLC patients. They found about 30 percent of NSCLC patients relapsed within two years after an initial objective response to Opdivo, and 65 percent of responders progressed at four years of follow-up, suggesting the rate of acquired resistance falls somewhere between these figures.
Mark Awad, senior author of the JCO paper and associate chief of the Lowe Center for Thoracic Oncology at the Dana-Farber Cancer Institute, noted that researchers have taken steps to understand acquired resistance to targeted therapies in lung cancer, but resistance to immunotherapy has not been explored as much.
For instance, previous studies have characterized common mechanisms of resistance to EGFR inhibitors and KRAS inhibitors, among others, in NSCLC, and many pharma companies have begun developing next-generation treatments that try to overcome resistance to earlier targeted therapies. There are several ongoing studies involving newer investigational drugs in this regard, including PF-07284892, a SHP2 inhibitor from Pfizer; patritumab deruxtecan, Daiichi Sankyo's HER3-targeted treatment; and THE-349, Theseus Pharmaceuticals' pan-EGFR inhibitor.
"By contrast, with immunotherapies, this class of medicine certainly has revolutionized care for lung cancer as well as a number of other cancer types, but we had relatively little understanding of how cancers may initially respond to treatment, then develop resistance to [it]," Awad said.
Awad also emphasized the importance of doing repeat biopsies after a patient progresses on treatment to track the tumor's molecular evolution under therapeutic pressure. The practice is more common at academic institutions than in community oncology settings, Awad noted, and more common for patients who receive targeted therapies.
"There is more of a push to do repeat biopsies and repeat sequencing because a lot of research over the last several years has improved the understanding of resistance mechanisms to those therapies, and certain clinical trials are geared towards targeting those resistant mechanisms, but we haven't really had such a pathway or paradigm to do that with immunotherapy yet," he explained. "Through our work and the work of other groups, our hope is that there will be more of a rationale and reason to do these repeat biopsies if it's safe and feasible for the patient."
Awad and his colleagues analyzed matched pre- and post-immunotherapy biopsies samples from 82 patients with NSCLC. They then compared their findings with a control cohort of samples from patients who received non-immunotherapy treatment between biopsies, including 32 who received chemotherapy, 89 who received targeted therapy, and 17 who received both.
They found that 62 percent of patients in the investigational cohort had genomic changes in the pre- and post-immunotherapy biopsies at the time of resistance to immunotherapy. Acquired mutations were identified in 28 percent of cases while acquired copy number alterations were found in 49 percent of cases.
The most common acquired mutations included loss-of-function alterations in STK11, B2M, SMARCA4, NF1/2, APC, CDKN2A/B, MAP3K1, MAP2K4, KEAP1, MTOR, JAK1, and JAK2, and activating mutations in PIK3CA, SOS1, PDGFRA, ERBB2, and BRAF.
The activating mutations in PIK3CA, SOS1, PDGFRA, ERBB2, and BRAF identified in these resistant tumor samples could also be targeted with approved and investigational drugs to re-sensitize tumors to immunotherapy.
When compared to the pre- and post-treatment genomic changes in the control group, the researchers found that none of these same acquired mutations occurred in the patients who received only chemotherapy while those who received targeted therapies, such as EGFR, ALK, or MET inhibitors, had some overlap with acquired resistance mutations in EGFR, PIK3CA, KRAS, and BRAF.
Among patients in the investigational cohort who had acquired copy number alterations, the most common heterozygous losses were in B2M, STK11, SMARCA4, and KEAP1, which were not found in the control groups.
The researchers noted that loss-of-function mutations in STK11, KEAP1, and SMARCA4 have been found to drive primary resistance to checkpoint inhibitors in NSCLC. There are also investigational drugs in development that target these mutations that could be used to re-sensitize these resistant tumors to immunotherapy treatment. Pharma companies like Tango Therapeutics and BerGenBio are developing treatments against STK11-mutant tumors, and Prelude Therapeutics is developing therapies targeting SMARCA4-mutant cancers.
"Right now, we don't have approved targeted therapies for mutations in STK11, KEAP1, SMARCA4, or others, but we do have trials that are open and enrolling for cancers that have these exact mutations in STK11, KEAP1, SMARCA4," Awad said. "You might not think of these trials based on the initial sequencing, but if you repeat the biopsy and see the acquired mutations, that might help steer patients towards [more biomarker-informed] clinical trials."
The researchers also explored tumor immunophenotype differences in a smaller group of 16 patients with adequate pre- and post-treatment samples. They used a machine learning-based computational pathology platform, PathML, to assess tumor-infiltrating lymphocytes (TILs) and found TILs decreased significantly after immunotherapy treatment. Other immune markers, such as CD3e-positive T cells, CD8a-positive T cells, and PD-1-positive cells, also decreased in the post-immunotherapy samples compared to pre-treatment samples.
The researchers further examined HLA class I expression, biomarkers correlated with improved outcomes to immunotherapy, and B2M expression using IHC across the matched samples. They found a decrease in HLA class I expression in the post-immunotherapy samples compared to pre-treatment samples. In the control groups, by contrast, there was a slight increase in HLA class I expression in both the chemo and targeted therapy groups. Of patients who had acquired B2M mutations, half also had abolished B2M expression according to IHC analysis, which has been associated with primary resistance to immunotherapy.
"If B2M is mutated, the tumor likely isn't presenting itself to the body's immune system anymore, and it may be a way of evading or hiding from the immune system," Awad explained. "Similarly, other samples had decreased HLA class I expression or down regulation of HLA class I, suggesting that the tumors may have developed or evolved ways of evading immune detection or surveillance."
The loss of HLA class I expression is also a potentially targetable biomarker in these patients. Previous research has shown that certain treatments, including certain chemotherapies and HDAC or DNMT inhibitors, can restore HLA expression.
The researchers noted that the potential targetable biomarkers identified in this research need to be explored further before implementing any change in the clinic. But Awad added that understanding these acquired resistance mutations may help guide more precise treatment strategies after immunotherapy.
"Right now, we have a lot of clinical trials open for patients who have been on a prior immunotherapy or prior PD-1 inhibitor, but a lot of those trials are biomarker unselected, meaning they're just enrolling all comers," Awad said. "If we understood why some patients were resistant to immunotherapy, maybe we could steer one group to a trial [using these biomarkers] and another group to a different trial that might be more appropriate for their cancer's biology. The goal is to try to understand the biology of resistance more fully, to see if we can either prevent or delay resistance from happening, or if it does develop, how could we overcome resistance to therapy?"