Abstract 471P
Background
Survival of patients with metastatic colorectal cancer (mCRC) has improved in the last decade but remains heterogeneous. The two main prognostic scores for mCRC (Köhne and GERCOR) were developed before targeted therapy approval and do not account for resectability of liver-only metastases. We propose a new prognostic score - GEMCAD score.
Methods
Patients with mCRC treated with first-line doublet/triplet chemotherapy (CT) with or without monoclonal antibodies (mABs) were analyzed. The score was first tested in a training set of 634 patients from three phase II trials and one prospective cohort and later validated in a larger cohort of 1088 patients (validation set) from a prospective observational study (GEMCAD 14-01). Patients were classified as: (low-risk) if liver metastases were considered resectable (<10 nodules), PS 0-1 and LDH ≤1.5 ULN; (intermediate-risk) if liver metastases were not resectable (>10 nodules) or with extrahepatic spread, PS 0-1 and LDH ≤1.5 ULN; (high-risk) if PS 2 or LDH >1.5 ULN. GEMCAD score discrimination capacity was evaluated and compared to Köhne and GERCOR scores using Harrel’s C index (HCI).
Results
In the training and validation sets, distribution of patients according to GEMCAD score was: 21.1% and 16.6% in low-risk, 53.6% and 50.1% in intermediate-risk, and 25.1% and 33.3% in high-risk. Median overall survival (OS) was 30.5, 19.2 and 15 months (p<0.0001, HCI 0.607) and 25.3, 19.3 and 14.2 months (p<0.0001, HCI 0.621) in low, intermediate and high-risk, respectively. Similar results were obtained when applying Köhne and GERCOR scores in the training (p<0.0001, HCI 0.604; p<0.0001, HCI 0.604) and validation sets (p<0.0001, HCI 0.620; p<0.0001, HCI 0.623). When comparing patients treated with CT + mABs (n=1025) versus patients treated with only CT (n=697), performance of three scores (GEMCAD, Köhne and GERCOR) remained significant but with lower discrimination capacity (HCI 0.590 vs 0.647, 0.588 vs 0.641, and 0.599 vs 0.635, respectively).
Conclusions
GEMCAD score defines clearly three prognostic groups for death and allows treatment guidance. Its performance is comparable to Köhne and GERCOR scores. All three scores perform better in patients treated with chemotherapy alone.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
Joan Maurel.
Funding
Has not received any funding.
Disclosure
C. Fernández-Martos: Honoraria (self): Merck; Honoraria (self), Research grant/Funding (self): Sanofi; Honoraria (self): Servier; Honoraria (self): Amgen; Leadership role: GEMCAD. A. Ruiz-Casado: Speaker Bureau/Expert testimony: Roche; Speaker Bureau/Expert testimony: Sanofi; Speaker Bureau/Expert testimony: Merck; Speaker Bureau/Expert testimony: Bayer; Speaker Bureau/Expert testimony: BTG; Speaker Bureau/Expert testimony: Servier; Speaker Bureau/Expert testimony: Celgene; Speaker Bureau/Expert testimony: Lilly. J. Gallego: Speaker Bureau/Expert testimony, Research grant/Funding (self): Lilly; Speaker Bureau/Expert testimony, Travel/Accommodation/Expenses: Amgen; Advisory/Consultancy, Research grant/Funding (self): BMS; Advisory/Consultancy: Ipsen; Advisory/Consultancy: Roche; Advisory/Consultancy: Servier; Advisory/Consultancy: Merck; Travel/Accommodation/Expenses: Novartis; Leadership role: Agamenon Spanish Registry Gastric Cancer. J. Aparicio: Advisory/Consultancy: Roche; Advisory/Consultancy: Amgen; Advisory/Consultancy: Merck; Advisory/Consultancy: Sanofi; Advisory/Consultancy: Bayer; Advisory/Consultancy: Servier; Advisory/Consultancy: Celgene. E. Gálvez: Speaker Bureau/Expert testimony: Hoffmann-La Roche; Speaker Bureau/Expert testimony: Bristol Myers Squibb; Speaker Bureau/Expert testimony: Merck; Speaker Bureau/Expert testimony: Boehringer Ingelheim. J. Maurel: Honoraria (self): Sirtex; Honoraria (self): Servier; Honoraria (self): Pierre-Fabre; Honoraria (self): Advance Medical; Honoraria (self): Shire; Honoraria (self): Astra-Zeneca; Honoraria (self): Bayer; Honoraria (self): Sanofi; Honoraria (self): Roche; Research grant/Funding (institution): Merck; Research grant/Funding (institution): Amgen; Research grant/Funding (institution): Biocartis; Research grant/Funding (institution): Nanostring; Research grant/Funding (institution): Incyte. All other authors have declared no conflicts of interest.