Strategy for dose escalation using 3-dimensional conformal radiation therapy for lung cancer.

PURPOSE: Local failure is a major obstacle to the cure of locally advanced non-small-cell lung cancer. 3-Dimensional conformal radiation therapy (3-DCRT) selects optimal treatment parameters to increase dose to tumor and reduce normal tissue dose, potentially permitting dose escalation. There are several ongoing trials of dose escalation using 3-Dimensional conformal radiation therapy for non-small-cell lung cancer. We performed this analysis to determine if data derived from dose volume histograms could be used as the basis for designing the method of dose escalation in these trials. METHODS AND MATERIALS: Between 1990 and 1993, 31 patients were treated with 3-DCRT and had complete normal tissue dose volume histograms created as part of the planning process. The stage distribution was stage I/II 13%, stage IIIa in 45%, and stage IIIb in 42%. The median radiation dose to gross disease was 70.2 Gy (52.2-72 Gy). Elective mediastinal irradiation (50.4 Gy) was administered to 52% (16/31) of patients.
RESULTS: The major toxicity encountered in this experience was pulmonary. Dose-volume-histogram data were used to analyze the predictors of toxicity and showed a correlation between risk of pulmonary toxicity and indices of dose to lung parenchyma. Grade 3 or higher pulmonary toxicity occurred in 38% (3/8) of pts with >30%of lung volume receiving > or =25 Gy, versus 4% (1/23) of pts. with < or = 30% lung receiving > or = 25 Gy (p=0.04). Grade 3 or higher pulmonary toxicity occurred in 29% (4/14) of patients with a predicted pulmonary normal tissue complication probability of 12% or higher versus 0% (0/17) in patients with a predicted probability of less than 12% (p=0.03). The single fatality occurred in a patient with a calculated pneumonitis probability of 85% and a high percent (49%) lung volume receiving >= 25 GY.
CONCLUSION: This preliminary experience demonstrates a correlation between lung dose-volume-histogram data and the risk of severe pulmonary toxicity. This provides an opportunity to modify the method of radiation dose escalation. Dose-volume-histogram data can allow escalation according to the risk to the lung parenchyma (which is the major organ of concern) rather than escalation according to tumor dose levels. Because of teh major inter-patient variability of intrathoracic tumor bulk and anatomic distribution, this strategy is intuitively appropriate. This approach may facilitate completion of dose escalation studies and identification of maximum tolerable pulmonary dose levels.