Virtual Library

Start Your Search

Daniel J. Mullen

Author of

  • +

    MA01 - Early Stage Lung Cancer: Questions and Controversies (ID 894)

    • Event: WCLC 2018
    • Type: Mini Oral Abstract Session
    • Track: Treatment of Early Stage/Localized Disease
    • Presentations: 1
    • Moderators:
    • Coordinates: 9/24/2018, 10:30 - 12:00, Room 202 BD
    • +

      MA01.03 - An Externally Validated Nomogram for Predicting Distant Metastasis After SBRT for Early Stage Non-Small Cell Lung Cancer (ID 11182)

      10:40 - 10:45  |  Author(s): Daniel J. Mullen

      • Abstract
      • Presentation
      • Slides


      SBRT is a standard option for patients with early stage NSCLC who are medically inoperable. While SBRT is associated with excellent local control, distant metastases (DM) represent the primary pattern of failure. Adjuvant systemic therapy has not traditionally been used in this patient population due to medical comorbidities. With the advent of immunotherapy that may be better tolerated, there has been a renewed interest in identifying patients that may derive benefit. We developed and internally validated a nomogram to predict the likelihood of DM after SBRT for early stage NSCLC which was then externally validated.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Our lung SBRT registry was queried for patients with early stage NSCLC treated with definitive intent from 2003-2017 and 1002 patients were identified for analysis to develop the model. A dataset from an external institution was used to similarly identify patients and 737 were used for the validation cohort. Random Survival Forest was used to assess importance, interactivity, and overall predictive ability with respect to DM for 14 variables. A Fine-Gray competing-risks regression model was formulated where apparent interactions were examined with likelihood-ratio tests. Backward variable selection was implemented to reduce to a parsimonious model. The concordance probability (C-index) of the model was internally validated with 10-fold cross validation.

      4c3880bb027f159e801041b1021e88e8 Result

      The median overall survival was 1.71 years internally and 1.92 years externally. Median follow-up was 18.3 months and 21.1 months. 1-year incidence of DM was 16% and 12.1% in the internal and external cohorts, respectively. Random Forest analysis suggested that tumor size and PET SUV are the most important predictors of distant failure. The 1-year cumulative incidence (CI) of DM was 18.5% for PET SUV ≥4.1 vs 8.4% for <4.1. 1-year CI for tumor size >3 cm was 26% vs 12.6% for ≤3 cm. The median time to DM was 0.86 years internally and 1.1 years externally. The final nomogram included tumor size, histology, PET SUV, age, KPS, and active smoking status, and had a cross-validated C-index of 0.62. The nomogram provides predictive value for probability of DM at 1-year between 10 and 70%.

      8eea62084ca7e541d918e823422bd82e Conclusion

      This novel nomogram with external validation can be used to predict the 1-yr DM risk after SBRT for patients with early-stage NSCLC, accounting for the competing risk of death. This nomogram may help define patient subsets for stratification in future clinical trials to help identify who may benefit from adjuvant systemic therapy after SBRT to reduce the incidence of DM and disease-related death.


      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.