Start Your Search
P1.06 - Poster Session 1 - Prognostic and Predictive Biomarkers (ID 161)
- Event: WCLC 2013
- Type: Poster Session
- Track: Biology
- Presentations: 1
- Coordinates: 10/28/2013, 09:30 - 16:30, Exhibit Hall, Ground Level
P1.06-041 - Prognostic impact of cytoskeleton regulatory protein human Mena (hMena) isoforms in resected, node-negative, non-small-cell lung cancer: validation of a clinic-molecular prognostic model. (ID 2609)
09:30 - 16:30 | Author(s): A. Ceribelli
Human Mena and the isoform hMena[+11a] are cytoskeleton regulatory proteins involved in adhesion, motility, regulated in the epithelio-mesenchimal transition. Here, we investigated their potential prognostic value in node-negative non-small-cell lung cancer (NSCLC) patients.
Pan-hMena, hMena[+11a], E-cadherin, vimentin, ER-beta, EGFR, HER-2, pAKT, detected immunohystochemically on duplicate TMA and clinical factors (sex, age, histology, grading, T-size, number of resected nodes, RN) were correlated to 3-yr disease-free (DFS), cancer-specific (CSS), and overall survival (OS) using a Cox model. ROC analysis provided optimal cut-off values and model validation. A logistic equation including regression analysis coefficients was constructed to estimate individual patients’ probability (IPP) of relapse. Internal cross-validation (100 simulations with 80% of the dataset) and external validation was accomplished.
In a training set of 248 patients (median follow-up: 36 months, range 1-96), Pan-hMmena and hMena+11a were the only biological variables displaying significant correlation with outcome(s), confirmed by the cross-validation (replication rate: 78%, 83%), with a prognostic model accuracy of 61% (standard error 0.04, p=0.0001). Patients with high pan-hMENA expression had a non-significant trend towards a worse outcome, while patients with high hMena+11a expression had a significant and borderline significant advantage in DFS (p=0.03) and OS (p=0.056), respectively, and a non-significant trend towards a better CSS. Univariate and multivariate 3-yr median individual patient probabilities of recurrence were 70.9 (range 40.3-94.4) and 41.2 (range 13.6-86.5), respectively (data not shown). The subgroup of patients with High Pan-hMena/Low hMena11a relative expression fared significantly better than any of the other 3 groups (p≤0.002 for all outcomes). On the basis of the combination between this molecular hybrid variable and T-size and RN, a 3-class risk stratification model was generated; the derived 3-risk class survival model strikingly discriminated between patients at different risk of relapse, cancer-related death, and death for any cause, with a prognostic accuracy of 61% (standard error 0.03, p=0.01), according to ROC analysis. The 3-risk class survival model was externally validated in an independent dataset of 133 patients, and significantly discriminated between patients at Intermediate- and High-Risk of relapse and cancer-related death.
The expression of the hMena and its isoform may represent a powerful prognostic factor in early NSCLC and usefully complements clinical parameters to accurately predict individual patient risk..