74PD - Radiation-induced lung toxicity prediction modeling in NSCLC: Importance of baseline toxicity scoring

Date 06 May 2017
Event ELCC 2017
Session Imaging and locally advanced NSCLC
Topics Non-Small-Cell Lung Cancer, Locally Advanced
Lung and other Thoracic Tumours
Surgery and/or Radiotherapy of Cancer
Presenter Gilles Defraene
Citation Annals of Oncology (2017) 28 (suppl_2): ii24-ii27. 10.1093/annonc/mdx092
Authors G. Defraene1, W. Van Elmpt2, D. De Ruysscher2
  • 1Department Of Oncology, KU Leuven - University of Leuven, 3000 - Leuven/BE
  • 2Department Of Radiation Oncology (maastro), Grow – School For Oncology And Developmental Biology, Maastricht University Medical Center, Maastricht/NL

Abstract

Background

Validated outcome prediction models with high discriminative power are important for a cost-effective deployment of proton therapy for locally-advanced non-small cell lung cancer (NSCLC). We validated a model predicting lung toxicity 6 months after radiotherapy (RT) treatment.

Methods

The model published by Appelt et al. (Acta Oncol 2014) was selected from a literature search. It relies on a review of radiation pneumonitis reports and retained the most important predictors, Mean Lung Dose (MLD) as dosimetric factor and 6 factors influencing the patient’s susceptibility: pre-existing pulmonary comorbidity, age>63 years, mid/inferior tumor location and sequential chemotherapy as risk factors, and current smoking and smoking history as protective factors. A dataset of 109 NSCLC patients treated at MAASTRO Clinic using 1.8 Gy fraction doses (two fractions per day) up to 79.2 Gy was studied. The required parameters were retrospectively collected together with the dyspnea endpoint (CTC 3.0 scoring) at baseline and at 6 months after RT. All treatments were performed using 3D-conformal RT techniques as it was the case in the study of Appelt et al. Odds ratios were calculated using logistic regression modelling.

Results

19.3% of patients presented post RT dyspnea≥2. Our dataset confirmed current smoking and pulmonary comorbidity as prognostic factors for this endpoint, with similar odds ratios (OR = 0.28 (p = 0.02) and OR = 2.95 (p = 0.02), respectively). Tumor location OR was outside of the reported 95% CI. When predicting the change in dyspnea with respect to the baseline score (delta dyspnea≥1, prevalence of 18.6%), the two prognostic factors were not significant anymore (OR = 0.56 (p = 0.27) and OR = 0.47 (p = 0.21), respectively). Both factors exhibited strong correlation with the baseline patient status: worse baseline dyspnea is often a manifestation of existing comorbidities and it determines the probability to stop smoking. MLD was not associated with outcome in any of our models.

Conclusions

It is crucial to consider delta toxicity endpoints in prediction modeling in order to obtain meaningful models reflecting radiotherapy outcome. Acknowledgement: This project has received funding from the EU under grant agreement no 601826 (REQUITE).

Clinical trial identification

Legal entity responsible for the study

KU Leuven

Funding

EU

Disclosure

All authors have declared no conflicts of interest.