4513 Background: Patients (pts) with advanced renal cell carcinoma with sarcomatoid features (sRCC) have poor prognosis and suboptimal outcomes with anti-VEGF targeted therapy. Nivolumab plus ipilimumab (N+I) demonstrated superior objective response rate (ORR) and overall survival (OS) vs sunitinib (S) in previously untreated pts with International Metastatic RCC Database Consortium (IMDC) intermediate/poor (I/P)-risk, clear-cell, advanced RCC in the phase 3 CheckMate 214 trial. Methods: We performed a post-hoc exploratory analysis of N+I vs S in CheckMate 214 sRCC pts. The presence of sarcomatoid features was assessed by keyword search for “sarcomatoid” in pts with available local pathology reports accompanying pretreatment tumor samples. Results: 842 (77%) of 1096 intention-to-treat pts had local pathology reports available, including 112 randomized pts with I/P-risk sRCC (N+I, n = 60; S, n = 52). Baseline characteristics of sRCC pts were balanced between arms. Notably, 47% vs 53% of I/P-risk sRCC pts in the N+I and S arms had tumor PD-L1 expression ≥1% at baseline, which was higher than in all I/P-risk pts (N+I, 26% vs S, 29%). In descriptive analyses performed at a minimum follow-up of 30 months, confirmed ORR and complete response rate per investigator (RECIST v1.1), OS, and progression-free survival (PFS) per investigator were improved with N+I vs S in I/P-risk pts with sRCC (Table). No new safety signals were seen in sRCC pts. Conclusions: In this post-hoc descriptive subgroup analysis of CheckMate 214, N+I demonstrated promising efficacy and prolonged survival vs S, with consistent safety, in previously untreated, I/P-risk, advanced clear-cell RCC with sarcomatoid features. Prospective studies of N+I that include pts with sRCC are ongoing. Clinical trial information: NCT02231749. [Table: see text]

e21511 Background: Non-small cell lung cancer (NSCLC) is the leading cause of cancer mortality worldwide. Tumor biopsy represents the standard for molecular diagnosis, nevertheless it is not always feasible to obtain. Liquid Biopsy (LB) may offer an alternative in the first line setting. Methods: Retrospective review of case files and next generation sequencing report (NGS) from LB from patients in a Validation Trial, performed from January 2018 to September 2019. We present the molecular alterations and clinical characteristics of the population. Results: 147 records and LB reports of patients 18 years and older with metastatic lung adenocarcinoma without prior treatment were included. 40% of the population had been selected based on a commercial local PCR test for EGFR (Idylla). 49% of them were female with Mean age of 60.9 ± 12.7. 56% had a history of smoking with an average packs/year of 20.5 (1-156) and 39% had exposure to wood smoke, 92% had an ECOG 0-1. 65 % of the Patients were diagnosed in Clinical Stage (8th Edition) IVA and 22% in IVB. 13% had SNC metastasis. The LB detected genomic alterations in 85.5% of the cases. 78% represent pathogenic mutations and 7.5% variants of uncertain significance (VUS).14.5% of the biopsies could not detect any ctDNA. The most frequent aberrations reported were TP53 in 51.7%, KRAS 16%, EGFR 16%, ALK 9%, PIK3CA 4%, RET 4%, BRAF 3%, BRCA 3, and HER2 3%. In addition, 20% had bTMB ≥10. Only 20% of the patients could receive a targeted therapy or immunotherapy of the potential ≈50%. Most cases had co-mutations (1-6 x case). Looking for factors associated with the presence of bTMB by a logistic regression model, it was possible to identify the presence of CNS metastases and smoking as identified with an OR of 3,688 (p = 0. 0.042) and 3.952 (p = 0. 024) respectively. Conclusions: It is feasible to have tumor molecular analysis through liquid biopsy in most cases, with genetic alterations previously reported in tumor tissue in the first line setting. [Table: see text]

e21544 Background: A relationship between EGFR signaling pathway expression in skin and the use of targeted cancer therapies has been consistently demonstrated. Nonetheless, consistent evidence to support the use of skin biopsies as a surrogate for therapeutic evaluation. Methods: The present study is a prospective single-blind analysis of skin biopsies of patients with confirmed advanced EGFR mutated lung adenocarcinoma. Immunohistochemistry was performed with EGFR, p27, Ki67, STAT3, and MAPK, as well as an H&E histopathological analysis, looking for their relationship with the response to treatment with tyrosine kinase inhibitors. ROC curve analysis was used to determine the cutoff value for each biomarker selected dichotomizing the response to treatment as mentioned in the tissue samples section (adequate response or no response). Kaplan Meier analysis for progression-free survival was performed. Results: From the 35 biopsies obtained, 21 (60%) of the patients were women and 14 (40%) men; the mean age of participants was 60.6±11.7 years. Twelve patients (34.3%) were at the pre-treatment group, 12 (34.3%) had an adequate response to treatment and 11 (31.4%) were at the no response to treatment group. The median progression-free survival was 9 months. The next biomarkers were significantly related to an adequate response to treatment by using a bivariate correlation test: EGFR (p = 0.025), Ki67 (p = 0.015), STAT3 (p = 0.017), stratum corneum thickness (p = 0.039) and the number of layers of the stratum corneum(p = 0.041). A better median of progression-free survival was obtained on those with a value above of the cutoff preestablished of EGFR (21 months versus 7 months, 95% CI 0-46 versus 4.23-9.77, p = 0.025) and number of layers of the stratum corneum (21 months versus 8 months, 95% CI 0-43.81 versus 6.72-9.28, p = 0.030), however, for p27 a better median of progression-free survival was shown in those with a value below the cutoff before mentioned (21 months versus 8 months, 95% CI 8.17-33.83 versus 6.87-9.13, p = 0.031). Conclusions: We found a relationship between EGFR, Ki67, STAT3, stratum corneum, number of layers of stratum corneum, with the response to treatment, and better progression-free survival for high expression EGFR, number of layers of the stratum corneum and low expression for p27. The present study should incite to perform a further investigation to validate these markers as potential prognostic and predictive factors.

e14570 Background: Cancer in Mexico represents the third cause of death nationally. Immunotherapy agents have changed management and prognosis of cancer patients. Immune-checkpoint agents are now being used as both first- and second-line treatment improving cancer patient prognosis, nevertheless it is not always accessible. Methods: Retrospective review of patients diagnosed with any type of cancer that received immunotherapy during their disease course from the last 5-years. We present the most common malignancies, immunotherapeutic regimens received, population characteristics and survival analysis. Results: 130 records of patients ≥18 years that received immunotherapy were included. 52.3% female with mean age of 59 years (range, 22-89 years), 48.5% former smokers (average pack/year of 22.5) and 31.5% asbestos exposure. 53.1% were ECOG 0 and 82.3% clinical stage IV. 62.3% of cases were lung adenocarcinoma, followed by lung epidermoid carcinoma (13.1%) and 10% small-cell lung carcinoma. Immunotherapeutic agents used included nivolumab in 63.1%, pembrolizumab 12.3%, nivolumab + ipilimumab 8.5%, durvalumab 6.2%, pembrolizumab + ipilimumab 4.6%. 36.2% of patients received immunotherapy as second line treatment, 30.0% as third and 23.8% as first line treatment. The best Response Evaluation Criteria in Solid Tumors (RECIST) was RECIST 3 in 46.2% followed by RECIST 0 with 25.4%. Median progression-free survival (PFS) was 5 months (95% CI; 3.883-6.117) and median OS of 13 months (95% CI; 10.210-15.790). Analysis per immunotherapy on PFS (p = 0.0414) and OS (p = 0.0046) demonstrated pembrolizumab had the longest median PFS with 19-months and OS with 22-months. Analysis between tumor types was significant for both PFS (p = 0.0018) and OS (p = 0.0090) with melanoma having the longest median PFS (42-months) and OS (46-months). Conclusions: Immunotherapy has changed cancer management; however, its use depends on specific biomarkers and adequate patient selection. Not all patients benefit from immunotherapy, in a country like ours where resources are limited, it is of vital importance to properly select candidates for immunotherapy. Even though, all patients had an FDA-approved indication at the time of receiving immunotherapy, PFS and OS are not as significant as Phase 3 studies demonstrate, this may be due to late stages, advanced ECOG, correct biomarkers availability and adequate patient selection. It is important to mention that about 40-60% of patients with immunotherapy do not respond adequately. In our study, because there is more evidence with pembrolizumab, the patients who received it were better chosen and therefore there was an impact in PFS and OS. Immunotherapy selection also depends on physician experience to the different immunotherapy regimens.