Virtual Library

Start Your Search

T. Zhuang



Author of

  • +

    P3.08 - Poster Session 3 - Radiotherapy (ID 199)

    • Event: WCLC 2013
    • Type: Poster Session
    • Track: Radiation Oncology + Radiotherapy
    • Presentations: 1
    • +

      P3.08-028 - Impact of dosevolume parameters and clinical factors on severe acute radiation pneumonitis for lung cancer patients treated with concurrent chemo-radiotherapy (ID 1634)

      09:30 - 09:30  |  Author(s): T. Zhuang

      • Abstract

      Background
      Acute radiation pneumonitis (SARP) is a limiting factor on treating lung cancer with radiotherapy. Radiation dose to the lung was used to predict the occurrence of SARP, but the data were from American or European people, and might not suitable for Chinese people. This study was to identify predictive valueof different dosimetric parameters for SARP based on Chinese people.

      Methods
      147 NSCLC patients treated with concurrent chemotherapy and 3D-CRT between 2006 and 2010 was collected. Radiation pneumonitis(RP) was diagnosed according to RTOG criteria. Grade 3 or even severe RP was defined as SARP. Logistic dose response model was established, and Lyman - Kutcher - Burman normal tissue complication probability(NTCP) model was fitted. The predictive value of model was explored.

      Results
      The incidence of SARP is 9.5% (14/147). MLD, V20, V30, V40, and V50 (P = 0.017, 0.025, 0.010, 0.009 and 0.027, respectively) are determining factors for SARP.Our datasets shows that for SARP < 5%, MLD, V20, V30 V40 and V50 should be ≤ 16.77 Gy, V20 ≤ 34.15%,V30 ≤ 23.62%,V40 ≤ 18.57%,V50 ≤ 13.02%. ROC analysis show that areas under MLD, V20, V30, V40 and V50 curves is corresponding to 0.678, 0.661, 0.667, 0.677, and 0.651, respectively. In addition, the sensitivity and specificity of each parameter at cutoff values are: 78.0% and 48.1% for MLD; 42.9% and 82.0% for V20; 78.6% and 52.9% for V30; 71.4% and 61.7% for V40, and 57.1% and 67.7% for V50.As predictive value of each parameter alone is relatively week, using two or more parameters to predict SARP is recommended. By logistic regression, tumor locations is a determined factor for SARP. (P = 0.020, B = 2.042 95%CI= 1.121- 3.374). The incidence of SARP is greater in patients with tumors in right lower lung than other locations (22.2% vs 6.7%, P = 0.023). The best fit parameter value for logistic dose response model is shown as follow: b0=-6.66, b1=0.2520, TD50=26.43Gy, γ50=1.67. The fit curve shows that when MLD<17Gy, it is similar to QUANTEC curve. When MLD amount to about 17-18Gy, the curve becomes sharper, which implies that probability for SARP increases. The best fit parameter value for LKB-NTCP model is volume factor: n = 0.87 ± 0.40, slope factor: m = 0.27 ± 0.10, and radiation dose cause more than 50% complication TD50 (1) = 29.5 ± 8.0 Gy. Logistic regression and ROC analysis show that NTCP value is a determined factor for SARP.(Logistic regression: P=0.013;ROC Area under curve: 0.707,P=0.019).

      Conclusion
      MLD, V20, V30, V40 and V50 are determining factors for SARP. As predictive value of each parameter alone is relatively week, using two or more parameters, or using NTCP to predict SARP is recommended. Patient with tumor in right lower lung were at higher risk to develop SARP. The predictive values of dosmetric parameters are better in patients with tumor in upper lobes. The models show that when MLD amounts to about more than 17Gy, the curve becomes sharper and SARP odd increases. Limit MLD under 17Gy if possible is recommended for chinese patients in the clinic.