Predicting Risk Of Severe Radiation Pulmonary Fibrosis

Purpose/objective(s):To investigate prospectively whether clinical,dosimetric, functional,biological parameters are predictive for Grade≥2 pulmonary fibrosis induced by combined chemo-radiotherapy in patients with thoracic cancer.Materials/Methods:Eligible patients were previously untreated,histologically or cytologically proved thoracic cancer,18 to 70 years old with no distant metastasis. Neoadjuvant platinum-based chemotherapy was given,followed by 3D-CRT. Dosimetric parameters were generated as following:maximum lung dose,mean lung dose(MLD),and V₅,V₁₀,V₁₅,V₂₀,V₃₀,V₄₀.These physical metrics,together with clinical parameters of smoking history and pulmonary function tests(PFTs) were analyzed for associations with severe pulmonary fibrosis.Serial cytokine levels, including IL-1α,IL-1β,IL-6,IL-7,TNF-α,and TGF-β₁,were determined according to an IRB approved protocol by ELISA before and weekly during RT.To avoid interpatient divergence,post-RT/pre-RT ratios of serum markers were calculated as surrogates. Pulmonary fibrosis was evaluated according to the Common Terminology Criteria for Adverse Events version(CTCAE) 3.0.To perform univariate analyses,the log-rank test and student's T test were used.Fisher's exact test was applied to analyze the variation of cytokines between patients with severe and non-severe fibrosis.Cox regression was used for multivariate analysis.Results:104 patients(70%with lung cancer) were eligible for analyses with post-RT follow-up≥6 months or fibrosis occurred within 6 months.Median follow-up was 15.8 months(range,3.7-50.7).Chemotherapy regimens mainly included NP,EP,and CAP. The medium pre-RT cycle was one(range,1-6).Ten patients(9.90%) were found to have Grade≥2 pulmonary fibrosis and the actuarial incidence was 3.96%(4/101) at 6 month,6.93%(7/101) at 1 year.Median time to severe fibrosis was 7.4 months(range, 0.6-28.7).By univariate analysis,no dosimetric parameter(including MLD and V₂₀, p=0.511 and 0.460,respectively) was associated with Grade≥2 pulmonary fibrosis.The difference in pre-RT PFTs(FEV₁ and FEV₁/FVC) between patients with and without severe pulmonary fibrosis was statistically significant(p=0.004 and 0.0007, respectively).IL-1αw2/0 was found to be significantly associated with Grade≥2 fibrosis(p=0.023).Moreover,patients with a decreased IL-1αlevel at any week during RT were more likely to develop Grade≥2 fibrosis.(w1/0 p=0.048,OR=4;w2/O p=0.022,OR=6).Such trend was persistent throughout the whole RT course.By multivariate analysis,pre-RT FEV₁/FVC was found marginally associated with severe treatment-related pulmonary fibrosis(p=0.058).Conclusions:Our observations indicated that early variation of circulating IL-1αduring 3D-CRT was significantly associated with an elevated risk for developing severe pulmonary fibrosis.Decrease of IL-1αin week2,as well as Pre-RT PFTs(FEV₁ and FEV₁/FVC) were demonstrated independent risk factors for Grade≥2 pulmonary fibrosis in this group of patients.Neither of the dosimetric factors,including MLD and Vdose,nor TGF-β₁ was found to be significantly associated with severe pulmonary fibrosis.Further study in larger population is warranted.