Predicting the risk of symptomatic radiation-induced lung injury using both the physical and biologic parameters...

To correlate the volume of lung irradiated with changes in plasma levels of the fibrogenic cytokine transforming growth factor beta (TGFbeta) during radiotherapy (RT), such that this information might be used to predict the development of symptomatic radiation-induced lung injury (SRILI).

The records of all patients with lung cancer treated with RT with curative intent from 1991 to 1997 on a series of prospective normal tissue injury studies were reviewed. A total of 103 patients were identified who met the following inclusion criteria: (1) newly diagnosed lung cancer of any histology treated with RT +/- chemotherapy with curative intent; (2) no evidence of distant metastases or malignant pleural effusion; (3) no thoracic surgery after lung RT; (4) no endobronchial brachytherapy; (5) follow-up time more than 6 months; (6) plasma TGFbeta1 measurements obtained before and at the end of RT. The concentration of plasma TGFbeta1 was measured by an enzyme-linked immunosorbent assay. Seventy-eight of the 103 patients were treated with computed tomography based 3-dimensional planning and had dose-volume histogram data available. The endpoint of the study was the development of SRILI (modified NCI [National Cancer Institute] common toxicity criteria).

The 1-year and 2-year actuarial incidence of SRILI for all 103 patients was 17% and 21%, respectively. In those patients whose TGFbeta level at the end of RT was higher than the pre-RT baseline, SRILI occurred more frequently (2-year incidence = 39%) than in patients whose TGFbeta1 level at the end of RT was less than the baseline value (2-year incidence = 11%, p = 0.007). On multivariate analysis, a persistent elevation of plasma TGFbeta1 above the baseline concentration at the end of RT was an independent risk factor for the occurrence of SRILI (p = 0.004). The subgroup of 78 patients treated with 3-dimensional conformal radiotherapy, who consequently had dose-volume histogram data, were divided into groups according to their TGFbeta1 kinetics and whether their V(30) level was above or below the median of 30%. Group I (n = 29), with both a TGFbeta1 level at the end of RT that was below the pre-RT baseline and V(30) < 30%; Group II (n = 35), with a TGFbeta1 level at the end of irradiation that was below the baseline but a V(30) > or = 30% or with a TGFbeta1 level at the end of RT that was above the pre-RT baseline but V(30) < 30%; Group III (n = 14), with both a TGFbeta1 level at the end of RT that was above the baseline and V(30) > or = 30%. A significant difference was found in the incidence of SRILI among these three groups (6.9%, 22.8%, 42.9%, respectively, p = 0.02).

(1) An elevated plasma TGFbeta1 level at the end of RT is an independent risk factor for SRILI; (2) The combination of plasma TGFbeta1 level and V(30) appears to facilitate stratification of patients into low, intermediate, and high risk groups. Thus, combining both physical and biologic risk factors may allow for better identification of patients at risk for the development of symptomatic radiation-induced lung injury.