80 Background: In Ireland, 21% of colorectal cancers are metastatic at diagnosis, with 5-year-survival of 15%. Recent first line trial data suggests median overall survival for Stage IV CRC of > 30 months. Assessing extreme phenotypes can allow identification of distinct molecular and clinical pattens often masked in large heterogenous cohorts. We describe our cohort of long-term responders (LTR) with mCRC. Methods: We retrospectively reviewed all pts diagnosed with mCRC from 2015-2019. Inclusion criteria included unresectable disease and LTR defined as alive >30 months post first chemotherapy. All pts had colorectal focused next generation sequencing (NGS) inclusive of extended RAS/RAF/NTRK. Poor prognostic laboratory parameters were defined as alkaline phosphatase (ALP) ≥300 U/l, neutrophil/lymphocyte ratio (NLR) and platelet/lymphocyte ratio (PLR) >3 and >300 based on previously identified prognostic significance. Results: Of 104 pts treated for de novo mCRC, 11 patients were LTR 8 male (73%) and 3 female (27%). Median age at diagnosis was 62 (range R 39-75). 64% were never smokers. Median BMI was 25.7 (R 19-35). ECOG performance status was: 0 (28%), 1 (36%), 2 (36%). Metastatic sites included liver (n=9, 81%), lung (n=7, 64%), pelvic (n=2, 18%), peritoneal (n=1, 9%); 7 pts (64%) had ≥ 2 sites of metastasis (all including liver metastases). Median ALP and CEA at diagnosis were 80 (IQR 65-86) and 42 (IQR 7.4-192). Median NLR and PLR were 2.7 (R 1.27-5.8) and 158 (R 98-381). 73% were moderate (n=8) and 27% were poorly differentiated (n=3). All tumours were left sided (9 rectal, 2 sigmoid), mismatch repair proficient (pMMR) and BRAF wild type (WT). Molecular profiling identified 3 KRAS mutation (MT), 1 NRAS MT and 1 HER2 amplification. Median survival from first chemotherapy cycle was 49 months (R 32-110). All pts received first line chemotherapy with 5-FU backbone FOLFOX (82%), FOLFIRI (9%) or FOLFIRINOX (9%). Median time to first progression was 25 months (R 15-37). All pts received ≥2 lines; 82% had third line (n=9), 64% fourth line (n=7) and 5 patients had >4 lines. All eligible pts received biomarker directed therapy. Three pts had rechallenge regimens >12 months from prior use. Of those with rectal cancer (81% n=9) 7 patients (78%) had radiation to primary (5 short, 2 long course). Conclusions: Through extreme phenotypes, novel prognostic markers may be extracted from routine clinical data. Several favourable prognostic factors were consistently present: left sided tumours, pMMR, BRAF WT. Median ALP, NLR and PLR values fell within non–high risk ranges, indicating potential prognostic value. Broader NGS panels may identify important co-mutations and pathway dependence associated with prolonged response. Tumour characteristics. Colorectal Focused Panel Pts n (%) Variant KRAS MT 3 (27) 2 KRAS c.35G>A p.(G12D) 1 KRAS c.35 G>T p.(G12V) NRAS MT 1 (9) NRAS c.182A>T p.(Q61L) HER 2 amplified (FISH) 1 (9)
Talbot et al. (Sat,) studied this question.