Abstract 63P
Background
ABC are anatomically distinct and genetically heterogeneous tumours, characterized by poor response to treatments and overall survival (OS) <12 months. However, an outlier subset of ABC patients (pts) experiencing long-term survival has been identified by an updated analysis of the ABC-02 trial. We aimed at characterizing pattern of care and clinicopathologic features of this population in a real-world setting.
Methods
Pts surviving >18 months from ABC diagnosis (group A) were retrospectively identified at three Cancer Centres from 2002 to 2017 and variables of potential interest retrieved. A control cohort of ABC with a mOS <18 months (group B) was matched for age, gender, ECOG PS, primary tumor site, disease status, prior surgery, and treatment modality. The impact of covariates on survival was evaluated by logistic regression analysis, using a binary endpoint (mOS > vs < 18 months).
Results
A total of 78 pts fulfilling the inclusion criteria was included in the analysis, 39 in each group. Pts in both group A and B had superimposable clinicopathologic features: 33.3% and 43.5% aged >70 years (p=0.48), 66.7% and 61.5% females (p=0.81), ECOG PS 0-1 in all cases (p=1), iCCA in 66.7% and 64.1% (p=1), metastastic disease in 59% and 64% (p=0.81), prior surgery in 25.6% and 25.6% (p=1), first-line doublet chemotherapy in 77% and 77% (p=1). The mOS was 29 (95%CI 24.6-33.5) and 9 months (95%CI 6.6-12.9) and median progression-free survival was 9 (95%CI 7.1-11.6) and 4 months (2.5-6.7), in the group A and B, respectively. At the logistic regression analysis, low neutrophil-to-lymphocyte ratio (NLR) (OR 0.38; p=0.04), achievement of objective response to treatment (OR 0.16; p<0.001) and number of treatment lines (OR 0.29; p<0.001) were significantly associated with long-term survival.
Conclusions
In this study, we reported a considerable subset of ABC experiencing long-term survival (18%) with conventional chemotherapy in clinical practice. In addition to clinical factors, we identified low NLR as a prognostic determinant that may allow for a more accurate selection of ABC long-term survivors. While awaiting for an in-depth molecular characterization of this subgroup, we propose NLR as a stratification factor for daily practice and clinical trials.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.