12007 Background: CRICKET (NCT02296203) was designed to investigate the activity of the rechallenge with cet and iri as 3rd-line treatment in RAS/BRAF wild-type mCRC pts with acquired resistance to 1st-line cet- and iri-based therapy. The role of liquid biopsies as a tool to identify pts more likely to benefit from this strategy was investigated. Methods: Eligibility criteria included RAS/BRAF wild-type status on tissue samples; prior 1st-line iri-based, cet-containing regimen with at least RECIST partial response (PR), 1st-line PFS ≥6 months, and progression within 4 weeks after the last cet; prior 2nd-line oxaliplatin- and bevacizumab-based treatment. Pts received 3rd-line cet + iri until PD. The primary endpoint was response rate (RR) according to RECIST v1.1. With p0 = 5%, and p1 = 20%, 1-sided-α and β errors of 0.05 and 0.20, 27 pts were required. The null hypothesis can be rejected if responses are observed in ≥ 4 pts. Liquid biopsies were collected at the rechallenge baseline. ctDNA was analyzed with ddPCR for specific RAS/BRAF mutations (mut), and then by ultra-deep NGS with Ion Torrent S5 XL. Results: Between Jan 2015 and Jun 2017, 28 pts were enrolled in 9 centres. The primary endpoint was met. Six PRs (two unconfirmed) and 9 disease stabilizations (RR: 21%,95%CI: 10-40%, disease control rate: 54%,95%CI: 36-70%) were reported. RAS mut were found in liquid biopsies collected at the rechallenge baseline in 12 (48%) out of 25 evaluable pts (6 KRAS G12D, 5 KRAS G12V with 1 harboring also Q61H and 1 NRAS Q61L). No RAS mut were detected in samples from pts who achieved a confirmed PR. Pts with RAS wt ctDNA, had significantly longer PFS than those with RAS mut ctDNA (mPFS: 3.9 vs 1.9 mos; HR: 0.48 [95%CI 0.20-0.98], p = 0.048). No BRAF or PIK3CA mut were found. Conclusions: This is the first prospective demonstration of the activity of rechallenge with cet + iri in some mCRC pts initially sensitive and then resistant to first-line iri- and cet-based therapy, with no RAS/BRAF mut in pre-treatment liquid biopsies. Partially funded by Merck Serono SpA. Clinical trial information: NCT02296203.
Background: In gastroenteropancreatic (GEP) high-grade neuroendocrine neoplasms (H-NENs), Ki-67 threshold of 55% defines three prognosis subclasses: neuroendocrine tumor (NET) G3, neuroendocrine carcinoma (NEC) <55%, and NEC ≥55%. We investigated whether the molecular profiling of H-NENs differs among these subcategories and evaluated potential therapeutic targets, including PD-L1. Methods: In GEP-NEN patients, we evaluated: (i) 55% threshold for Ki-67 labeling index for further stratifying NEC and (ii) immunoreactivity and gene mutations by immunohistochemistry and targeted next-generation sequencing (T-NGS). Results: Fifteen NETs G3 and 39 NECs were identified. Ki-67 labeling index was <55% in 9 NECs and ≥55% in 30 NECs. Gene mutations by NGS (TP53, 32.9%; KRAS, 5.5%; BRAF, 4.1%) were detected in 46.6% NENs, significantly enriched in NEC ≥55% (76.7%) compared to NEC <55% (55.6%) or NET (20.0%). PD-L1 staining in tumor-infiltrating lymphocytes was observed in NEC ≥55% (36.7%; p = 0.03). Median OS was 4.3 years in NET G3, 1.8 years in NEC <55%, and 0.7 years in NEC ≥55% (p <0.0001); it was 2.3 years with NGS wild-type, 0.7 years with ≥1 mutation (p <0.0001), 0.8 years in PD-L1-positive patients, and 1.7 years in PD-L1-negative subjects (p = 0.0004). In multivariate analysis, only the proposed subclassification approach yielded statistically significant differences between groups (NEC <55% vs. NET G3, HR 14.1, 95% CI 2.2–89.8, p = 0.005; NEC ≥55% vs. NET G3, HR 25.8, 95% CI 3.9–169, p = 0.0007). Conclusions: These findings identify NEC ≥55% as a biologically and prognostically distinct subtype and pave the way for more personalized treatment.
e16507 Background: The phase II BONSAI trial ( n = 25 ; NCT 03354884) met the primary endpoint demonstrating activity of cabozantinib in untreated metastatic collecting ducts carcinoma (mCDC), a rare and biologically poorly characterized disease. Here we report on molecular analyses of baseline tissue samples. Methods: Formalin-fixed paraffin-embedded (FFPE) samples from 18 mCDC patients enrolled in BONSAI were sequenced by TruSeq RNA Exome kit (Illumina). The data were mapped and quantified using STAR and htseq, respectively. Globaltest and edgeR packages in R were used to assess the correlation between transcriptional profiles and survival data. Nineteen samples underwent DNA Sequencing with the Oncomine Comprehensive Assay Plus (ThermoFisher Scientific). The reads were aligned to the human genome reference (hg19) and analyzed with Opencravat and IonReporter software. Germline variants were excluded based on 1000 Genomes, GnomAd and Exac databases, and Clinical annotation of somatic variants was performed using ClinVar. Results: The global expression levels of thirty-one genes have been found significantly associated with overall survival (OS). The natural grouping of the 18 tumor samples based on the 31-gene signature identified a main group of 11 cases, showing global higher expression levels in 22 out of the 31 genes. This group displayed overall a significant higher OS rate in comparison to the remaining 7 patients, carrying opposite expression trend and mostly undergoing to disease progression. The identified signature was enriched in biological processes like cell junction, cytoskeleton organization and methylation. FOS oncogene was among the 9 genes negatively associated to OS, showing common higher expression in the poorer survival rate group. Furthermore, a 22-gene signature was found significantly associated with progression-free survival (PFS), involving mainly genes associated to cell cycle regulation or recognized as components of Golgi apparatus. Only three genes were globally up-regulated in the group of 9 patients characterized by shorter PFS. Finally, heterogeneous pattern of somatic mutations was identified in 19 tumor samples, with at least 8 genes recurrently affected in more than three patients. Notably, mutated genes were mostly involved in DNA repair and chromatin modification processes. Conclusions: Our findings for the first time define specific molecular signatures that differentiate therapy-specific outcomes in first-line mCDC, warranting further investigation of their involvement in the tumor biology. Clinical trial information: NCT 03354884.
