Brain Metastasis in Patients With NSCLC Correlates With Enrichment of CDKN2A/B and EGFR Mutations

Significant enrichment of cyclin-dependent kinase inhibitor 2A/B (CDKN2A/B) and cell cycle pathway alterations in non-small cell lung cancer (NSCLC) with brain metastasis was observed compared with extracranial disease, and there was a higher fraction of genome-altered brain metastasis in comparison with matched primary tumor controls, according to an analysis of a molecularly profile cohort of NSCLC brain metastasis samples with clinical correlates.¹

The analysis represents that it is the largest cohort of its kind and was conducted based on the 30% prevalence of brain metastasis in NSCLC at presentation, as well as the 50% of patients who develop brain metastasis later. Prior research has provided limited information around brain metastasis genomics. Clinical data have been missing from previous research, and there has been a lack of match specimens.

Investigators, led by Luke R.G. Pike, MD, PhD, of Memorial Sloan Kettering Cancer Center (MSKCC), sought to evaluate a large cohort of resect brain metastases from NSCLC with match extracranial sample, identify unique genomic feature of brain metastases that impact the spread of cancer to the brain, and reveal the genomic biomarkers that have a part in post-treatment intracranial disease progression. Findings from the cohort analysis were presented at the 2022 American Society of Clinical Oncology (ASCO) Annual Meeting by Anna Skakodub of MSKCC.

The cohort assessed in the new analysis was included clinical data and outcomes for 244 patients with NSCLC and resected brain metastases. The brain metastasis samples were evaluated using MSK-IMPACT version 341, 410, 468, or 505, which is an FDA-approved next-generation sequencing assay. Testing was conducting using matched primary site tissue in 20.9% of patients and matched tissue from another metastatic site or cerebrospinal fluid in 18%. Investigators utilized the precision oncology tool, OncoKB, to filter for genomic variations for driver variants.

The patient population had a median age of 66 years (range, 31-91). The most common histology among the patients was adenocarcinoma, which was seen in 78%. Fifty percent of patients presented with brain metastasis. There were 121 treatment-naïve patients, and among those who received prior therapy, 83% received adjuvant stereotactic radiosurgery (SRS) to the resection site, and 28% received SRS to additional brain metastases.

Following tumor resection, 55.1% of patients had central nervous system (CNS) progression, which was regional progression for 42% of them. The most common salvage therapy utilized was SRS to brain metastases (32%).

At a median follow-up of 2.3 years (interquartile range, 1.3-4.3), the median CNS-progression-free survival was 1.2 years (95% CI, 0.9-1.4).

The most frequently altered genes in the brain metastasis sample included TP53 (72%), CDKN2A (34%), KRAS (31%), KEAP1 (26%), and EGFR (21%). Notably, CDKN2A was altered in 56% brain metastasis lesion (56%) compared with 14% of NSCLC primary sample (P =0.002), and the difference in genome altered brain metastasis samples vs NSCLC primary samples was a P value of 0.0001.

During the ASCO Annual Meeting presentation, Skakodub, explained the genomic portion of the analysis, stating, “we categorized types of progression as local, regional with a singular lesion or multifocal disease, and development of leptomeningeal disease, with them went ahead and compared these five groups focusing only on driver mutations this time. An interesting discovery was that EGFR alterations were significantly enriched in leptomeningeal disease group compared to others.”

Patients in the study were grouped based on CNS progression, and investigators discovered an EGFR alteration in 42% patients with leptomeningeal failure compared with only 18% of patients who had no CNS progression (P =0.03), showing an enriched number of EGFR alterations correlating with CNS progression as Skakodub mentioned. The same enrichment in EGFR alterations was demonstrated in individuals with local or regional progression vs those without at 42% vs 19%, respectively (P =0.03).

“Concurrently, we found that there was less expression of KRAS in the same group that probably accounts for mutual exclusivity in RT KRAS pathway,” explained Skakodub. “Another interesting finding was that NKX2 genes showed higher alteration frequency of 20% of patients without intracranial and more when compared to 6% and 8% respectively, in local progression, and leptomeningeal disease group. It is of particular interest because NKX2 gene shows to be a predictor of better disease outcome in early stage non–small cell lung cancer.”

Based on these findings, Skakodub et al recommend further investigation into the role of systemic therapy and time course to make the potential mechanisms for CNS failure in patients with NSCLC clearer.

_REFERENCE:

_{Skakodub A, Walch HS, Tringale KR, et al. Genomic analysis and clinical correlations of non-small cell lung cancer (NSCLC) brain metastasis (BM). J Clin Oncol. 2022; 40 (suppl 16): 2008. doi: 10.1200/JCO.2022.40.16_suppl.2008}