EGFR mutations are detected in 20%~50% of lung adenocarcinoma patients, EGFR exon 20 insertions are the next most common EGFR mutations in NSCLC after classical mutations, accounting for 4-10% of all observed EGFR aberration[5, 6]. Most EGFR exon 20 insertion mutations predict resistance to clinically achievable levels of tyrosine kinase inhibitors (TKIs) in advanced NSCLC, although there are rare exceptions. Some novel EGFR inhibitors have shown encouraging antitumor activity for EGFR exon 20 insertions[7], but commercially obtainable effective targeted drugs remain an unmet need in clinical management. Therefore, platinum-based chemotherapy is considered standard treatment for these patients.
This patient harbored EGFR exon 20 in-frame insertion (P772_H773insYNP+H773Y). exon 20 H773Y was previously detected combined with exon 20 insertions[6, 8]. Preclinical research indicated that the complex mutations of exon 20 insertion combined with H773Y resist 1st-generation EGFR-TKIs[8], while clinical evidence for the effect of 2nd or 3rd generation EGFR-TKIs are lacking.
Although ICIs have revolutionized the systemic treatment of patients of NSCLC with EGFRwild-type (WT), the activity and clinical impact in the subgroup of EGFR mutation-positive tumors have been lower than in the EGFRWT population in large 2nd- and 3rd-line phase III trials and 1st-line trials evaluating the efficacy of ICIs often excluded EGFR-mutation-positive patients. Anecdotal evidence suggests that some EGFR mutation-positive patients benefit from PD-1/PD-L1 inhibitors[9]. EGFR exon 20 mutations demonstrated a higher response rate and longer survival compared to classic EGFR mutations[10]. A case report demonstrated that a heavily pre-treated patient of NSCLC with EGFR exon 20 insertion mutation responded to ICIs-based therapy[11]. But the effect of upfront ICIs in lung cancer patients harboring EGFR exon 20 insertion mutation is lacking. This patient had high PD-L1 expression, PD-L1 is a positive predictive biomarker of ICIs, which has been prospectively validated and approved by the Food and Drug Administration (FDA), and was also positively associated with outcomes for ICIs in EGFR mutant NSCLC[4].
Considering this patient was neurologically asymptomatic, with rare EGFR mutations, high PD-L1 expression, and High TMB[12], therefore, 1st-line immunotherapy plus chemotherapy was applied. A durable response was achieved with impressive control in both extra-cranial and cranial tumor lesions.
Brain metastases, associated with poor clinical outcomes, are diagnosed in approximately 20% of NSCLC patients[13]. For asymptomatic BM, upfront systemic chemotherapy could be an option. However, traditional chemotherapy drugs have a limited role in BM management, owing to the presence of efflux pumps[14, 15], and lacking penetration of the blood-brain barrier (BBB). Here is the rational basis for ICIs-based treatment in patients with BM: first, antitumor T cells are activated and home to the brain from extra-cranial sites, second, tumor neo-vessels are leaky to facilitate ICIs penetrating to stimulate tumor-associated T cells. Prospective trials evaluating ICIs efficacy in previously untreated BM are scarce. It was supported by a few studies of anti-PD-1 mono-therapy that ICIs can induce intra-cranial response[16, 17],[18]. Series of retrospective studies provided evidence that intracranial and extra-cranial ICIs efficacy are comparable [16, 19].
Some questions have yet to be resolved regarding how to identify the patients most likely to benefit from ICIs-based therapy, and how to weigh the probability of immunotherapy benefit when concomitant potential positive and negative factors both exist. For this patient, concomitant alterations in immune-related genes were revealed including CTNNB1 S37F and ARID2 E1056X. CTNNB1 S37F is a gain-of-function mutation that could lead to aberrant activation of the WNT/β-catenin signaling, which is enriched in non-T cell inflamed tumors[20] [21] and has been linked to lack of benefit of immunotherapy in NSCLC[22]. Whereas, ARID2 E1056X is a loss-of-function mutation. ARID2, encoding a PBAF complex subunit, acts as an immunomodulator. ARID2 mutation is a potential biomarker positively indicating ICIs effectiveness in melanoma patients[23-25]. Considering that biomarkers are often complex and non-binary, and we only see part of the picture since the potential heterogeneity has not been evaluated, therefore, it is overly simplistic to seek predictive power from a single marker. An interconnected network of multiple factors would be the eventual biomarker to form a more complete puzzle.
In conclusion, our case demonstrated that upfront chemotherapy plus PD-1 inhibitor might be an option for some NSCLC patients of BM harboring EGFR exon 20 insertion and high PD-L1 expression/high TMB. Additional insights into gene aberrations provide more information. Further study is needed to validate the predictor involved in responders to ICIs-based therapy with EGFR mutations.