1203P - Prognostic model for progression free survival in patients with locally advanced non-small cell lung cancer treated with concurrent chemoradiation

Date 29 September 2012
Event ESMO Congress 2012
Session Poster presentation I
Topics Cytotoxic agents
Surgical oncology
Non-small-cell lung cancer
Biological therapy
Radiation oncology
Presenter wilma Uyterlinde
Authors W. Uyterlinde1, A.D. Vincent2, J. Belderbos3, T. Korse4, P. Baas5, M. van den Heuvel6
  • 1Thoracic Oncology, nki-avl, 1066cx - amsterdam/NL
  • 2Biometrics, Antoni van Leeuwenhoek Hospital, Amsterdam/NL
  • 3Radiation Oncology, Netherlands Cancer Institute, Amsterdam/NL
  • 4Clinical Chemistry, nki-avl, 1066cx - amsterdam/NL
  • 5nki-avl, amsterdam/NL
  • 6Thoracic Oncology, nki-avl, amsterdam/NL



A prediction model for progression free survival (PFS) in patients with non-small cell lung cancer (NSCLC) and treated with concurrent chemoradiation (CCRT), would help optimizing customized treatment planning. We analysed several patient factors related to PFS in patients with medically inoperable stage II and III treated with CCRT.

Patients & methods

Consecutive patients with locally advanced NSCLC treated with CCRT between 2008 and 2011 were included. Treatment consisted of intensity modulated radiotherapy up to 66 Gy (24 x 2.75Gy), and concurrent daily low dose cisplatin (6mg/m2). Age, gender, performance status, weight loss, TNM, tumor volume, involved lymph nodes, histology, metabolic activity of the primary (SUVmax), and a series of lab values including CYFRA, CEA, CA125, SCC, NSE, MT, Leukocytes, Neutrophils, HB, LD and platelets, were assessed before the start of treatment. All patients had a response evaluation by PET-CT scan six weeks after treatment followed by three-monthly X-rays or CT-scans. PFS was defined as the time from start of treatment until progression or death of any cause. Kruskal-Wallis and spearman correlation tests were used to assess associations between factors. Univariate, multivariate Cox proportional hazard models were used to identify high and low risk groups [Fig], defined using linear predictor tertiles.


Out of the 230 patients, 227 were eligible for analysis. With a median FU of 23 months the median PFS was 15 months (95% CI 12-21 months), and a 2 years overall survival of 55% (95% CI 48%-64%). At the time of analysis 134 PFS events had occurred. The primary factors associated with PFS were tumor volume and SCC, with the pairwise interactions between SCC and both gender and histology also being potentially important. A full Cox regression model including the interactions and a reduced model only including only SCC and tumor volume were constructed. In both models a difference in PFS at 24 months of roughly 40% was observed between the high-and low risk groups.


A model prognostic for PFS using SCC and tumor volume, was developed that identified stage II-III patients at high risk for progressive disease after CCRT. Validation of this model is ongoing.


All authors have declared no conflicts of interest.