Dosimetric correlates for acute esophagitis in patients treated with radiotherapy for lung carcinoma.

PURPOSE: Acute esophagitis is a common complication of radiotherapy (RT) for non-small-cell carcinoma of the lung. Previous reports have related esophagitis to dosimetric parameters such as the length of the irradiated esophagus, maximal dose, or volume of the organ treated beyond a threshold dose. However, when using oblique beams, a portion of the esophageal circumference may be outside the treated field, resulting in partial esophageal irradiation. Therefore, our aim was to determine whether the irradiated esophageal surface area and/or esophageal volume are predictive of acute esophagitis in relation to other clinical and treatment-related factors. METHODS AND MATERIALS: Complete dose-volume information was gathered for 166 patients receiving definitive RT for Stage I-IIIB non-small-cell carcinoma of the lung at our institution. Seventy-eight patients received chemotherapy (37 before RT and 41 concurrently). All patients were treated to doses of 60-74 Gy (median, 70 Gy) delivered in single daily fractions of 1.8-2.1 Gy. The doses were prescribed to the isocenter without using heterogeneity corrections; however, the doses were corrected to account for lung heterogeneity in this report. Esophageal contrast was used to contour the esophagus from the cricoid to the gastroesophageal junction in each case. Esophagitis was scored according to the Radiation Therapy Oncology Group criteria with Grade 2 or worse considered clinically significant. To determine the importance of the irradiated surface area, the volumetric treatment plan for each patient was prepared for analysis by relating a surface area to each point of the esophagus contour. Spearman's rank correlation was used to correlate the esophagitis score with A(d), where A represents the surface area (in centimeters squared) receiving the dose, d, or greater (in Gray), or V(d), where V represents the volume (in centimeters cubed) receiving d (in Gray). The surface areas studied were A(5)-A(80) or V(5)-V(80) in 5-Gy increments. The clinical parameters studied in univariate analysis included patient age, stage, performance status, use of pretreatment chemotherapy, and use of concurrent chemotherapy. Step-wise regression analysis was then used to determine the statistically significant factors predicting acute esophagitis.
RESULTS: Forty-five patients (27%) developed Grade 2 or worse esophagitis, 37 developed Grade 2, 7 Grade 3, and 1 Grade 4. No deaths resulted from this complication. The most statistically significant single parameters for predicting acute esophagitis were A(55), V(60), and the use of concurrent chemotherapy. Age, stage, performance status, and pre-RT chemotherapy had no statistically significant influence on the incidence of acute esophagitis. On logistic regression analysis, A(55) (p < or =0.0005), V(60) (p < or =0.001), and the use of concurrent chemotherapy (p = 0.001) emerged as statistically significant correlates of acute esophagitis.
CONCLUSION: The esophageal surface area receiving > or =55 Gy, the esophageal volume receiving > or =60 Gy, and the use of concurrent chemotherapy were the most statistically significant predictive factors for early esophagitis. Adequate dosimetric coverage of the planning target volume remains the goal of RT planning. High values of A(55) and/or V(60) are indicative of the development of acute esophagitis and may indicate a need to explore alternative RT planning options.