Impact Of Dose Volume Parameters On Radiation Pneumonitis For Lung Cancer Patients Treated With Intensity Modulated Radiation Therapy

ObjectiveThoracic radiotherapy is the most important factor of radiation pneumonitis. Intensity Modulated Radiation Therapy(IMRT) is widely used in thoracic tumor. Studies have proven that IMRT can improve the dose of tumor target and reduce the dose of lung, lowering the incidence of radiation pneumonitis. The dosimetric parameters are important predictors of RP. But the data of dose limiting which is used in the IMRT plans is afforded by the empirical data of three-dimensional conformal radiation therapy(3D-CRT). Compare with 3D-CRT, the dose distribution of IMRT are usually smaller in a larger volume of the lung, namely "a little to a lot". Accordingly, we designed this study to observe the incidence of RP after IMRT,analysis the predictive value of different dosimetric parameters for RP, and establish the corresponding model.MethodsCollect the information of NSCLC patients who treated with IMRT between December 2013 and July 2015. Follow up with patients more than six months;.The medical record was reviewed and the dosimetric was analyzed.RP was diagnosed according to RTOG and CTCAE4.0; Logistic dose response model was established, and calculate the NTCP value of the whole group of patients. The predictive value of model was explored.T test was used to identify difference between dosimetric parameters from patients RP<3and RP≥3。 To find out predictive power of each parameter, ROC analysis was introduced. Correlations between parameters was explored by Pearson correlation analysis. Factor reduction was used to identify represented factors.Results(1) The incidence of RP≥3 is 12.4%(15/121).(2) ROC analysis indicated that MLD, V30 are determining factors for RP≥3.(3)The best fit parameter value for logistic dose response model is shown as follow:.The relationship between the MLD and RP≥3: b0=-4.54, b1=0.1776, TD50=25.56 Gy, γ50=1.12. When MLD amount to about 17 Gy, the curve becomes sharper. That implies that probability for RP≥3 increases..The relationship between V30 and RP≥3: b0=-3.46, b1=0.0919, TD50=25.56 Gy, γ50=1.12.when V30≤25%,the curve is relatively flat.(4).According to the parameter value about 3D-CRT,we calculate the NTCP value of the whole group of patients.ROC analysis shows that NTCP value has a poor predictive value for RP≥3(Area under curve: 0.559,P=0.257).ConclusionsThe total incidence of RP≥3 is 12.4%.In IMRT,V30,MLD are determining factors for RP≥3.Establishing effective model can make full use of DVH information and predict the risk of RP;it is very conveninent and quick. The predictive value of model is better than simple parameters.The canter can establish the corresponding modol for different technologies for clinical application.