Radiation and Cancer Physics

SS 01 - Physics 1 -Best of Physics

9 - Modeling Patient-specific Risk Factors for Central Lung Tumor SBRT Planning

Sunday, October 21
1:45 PM - 1:55 PM
Location: Room 214 A/B

Modeling Patient-specific Risk Factors for Central Lung Tumor SBRT Planning
A. Modiri1, S. R. Rice2, C. A. Schonewolf3, A. T. Berman4, S. J. Feigenberg3, C. B. Simone II5, S. M. Bentzen6, and A. Sawant1; 1Department of Radiation Oncology, University of Maryland, School of Medicine, Baltimore, MD, 2Department of Radiation Oncology, University of Maryland Medical Center, Baltimore, MD, 3Department of Radiation Oncology, University of Pennsylvania, Philadelphia, PA, 4University of Pennsylvania, Department of Radiation Oncology, Philadelphia, PA, 5Department of Radiation Oncology, University of Maryland School of Medicine, Baltimore, MD, 6Greenebaum Comprehensive Cancer Center and Department of Epidemiology and Public Health, University of Maryland School of Medicine, Baltimore, MD

Purpose/Objective(s): Although lung SBRT has shown impressive outcomes in peripheral, inoperable, early-stage lung tumors, the use of similar dose schedules in central and ultra-central lung tumors has been associated with higher rates of severe and even fatal toxicities in multiple studies. In current clinical practice, physicians rely on aggregated, population-based dose constraints and fractionation schemas, modified on a case-by-case basis according to institutional or individual experience, and most commonly, the constraints from the recently reported RTOG 0813, based on expert opinion. However, there is a critical need for a data-driven framework that enables physicians to make objective choices in treatment planning and fractionation so as to optimize patient outcomes in terms of local control and treatment-related toxicities. In this study, we develop a mathematical model to balance these tradeoffs for predicting outcomes with morbidity.

Materials/Methods: Retrospective data from 58 consecutive central lung SBRT patients (46.5% females and 53.5% males) with median age of 70 years (range 48-94), median number of fractions of 5 (range 3-8), median dose per fraction of 10 Gy (range 7.5-20), and median tumor size of 2.0 cm (range 0.8-5.0), treated from 2009 to 2013, were analyzed in a multi-institutional study. At a median follow up of 36 months, 5%, 22% and 19% of patients experienced local failure (LF), nodal failure (NF), distant metastases (DM), respectively, and 29% died within 2 years from SBRT. Also, 3%, 9%, and 9% experienced rib fracture (RF), chest wall syndrome (CWS) and pneumonitis, respectively. Using multivariate logistic regression, predictive patient-specific outcome risk models were created and analyzed where 57 patients were used for training and 1 for validating in 58 iterations. Predictor variables included age, sex, diabetes, Eastern cooperative oncology group performance status (ECOG PS), smoking status, COPD, prior lung cancer (treated by either surgery, chemo or radiation), radiation dose and fractionation (as equivalent 4-fx dose) and tumor size. In addition, we performed survival and time-to-event analysis.

Results: Our model’s correct prediction rate for LF, NF, DM, dying within 2 years, RF, CWS and pneumonitis was 100%, 76%, 79%, 79%, 97%, 93% and 93%, respectively. In addition to tumor size, dose, and history of prior lung cancers, increasing age, lower ECOG PS and diabetes were significant predictors of LF (P<0.05). NF and RF were highly correlated with pneumonitis (P = 0.04), and CWS was highly correlated with the advent of DM (P = 0.01). The 4-year Kaplan Meier outcomes were DM: 33.1% ± 8.9%, NF: 26.4% ± 6.3%, and LF: 9.6% ± 4.6%. Overall survival was reduced by a factor of 2.3 for females (P<0.05), and, a factor of 6.1 for patients with COPD (P<0.05).

Conclusion: We generated a mathematical model toward a standardized adjustment of radiotherapy planning for an improved therapeutic ratio using patient-specific risk factors for central lung SBRT.

Author Disclosure: A. Modiri: None. C.A. Schonewolf: None. A.T. Berman: None. C.B. Simone: Employee; Nemours/Alfred I. duPont Hospital for Children. Chair, Executive Council; Chair, Lung Committee; Proton Collaborative Group (PCG). Editor-in-Chief; Annals of Palliative Medicine. Chair, Lung Resource Panel; American Society for Radiation Oncology. S.M. Bentzen: Travel Expenses; University of Copenhagen.

Arezoo Modiri, PhD

No relationships to disclose.


Send Email for Arezoo Modiri


9 - Modeling Patient-specific Risk Factors for Central Lung Tumor SBRT Planning

Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Modeling Patient-specific Risk Factors for Central Lung Tumor SBRT Planning