Predictors of severe esophagitis include use of concurrent chemotherapy, but not the length of irradiated esophagus: A multivariate analysis of patients with lung cancer treated with nonoperative therapy

Purpose: To identify in a multivariate analysis treatment-related factors p redisposing patients (pts) with lung cancer to acute esophagitis, expressed as a severity grade or Esophagitis Index (EI). Methods and Materials: Acute esophagitis is prospectively scored as an RTOG Grade in our institution during and after thoracic radiotherapy. Charts, t oxicity forms and digitally reconstructed radiographs (DRRs) of all pts wit h lung cancer who received thoracic radiotherapy (RT) between 11/95 and 1/9 9 were reviewed. Esophagitis grades for each time point were verified by re view of weekly physician and nursing treatment notes, hospital discharge su mmaries and referring physician notes and then plotted on graph against tim e. The area under the curve was calculated for each patient's graph and was defined as an Esophagitis Index. The length of esophagus was measured on e ach anterior DRR while assuming that esophagus overlies the vertebral bodie s on the anterior films and projects over the edge of the vertebral body on the oblique DRRs. This assumption was confirmed in 10 pts by digitizing es ophagus on CT simulator-derived slices and visualizing its position on DRRs . To compare RT doses delivered with different fractionation schemes to sta ndard fractionated doses, the equivalent RT doses were calculated using the linear-quadratic formula and alpha/beta ratio of 10. Univariate and multiv ariate analyses of several factors potentially influencing the maximum esop hagitis grade, as well as EI, were performed. Results: A total of 277 pts were identified. Pts were included in the analy sis (n = 105) if they fulfilled the following criteria: chart, toxicity for m and DRRs were all available; parallel opposed fields (no multiple fields) were used for both the initial and off cord/cone down fields; and an equiv alent dose of 45.0 Gy or more was delivered. Seventy-eight pts had Stage II I; 32, Stage IV, and the remainder, Stages I, II, or recurrent lung cancer (85 non-small cell and 18, small cell). Seventy-four pts were treated with definitive intent. Chemotherapy was given concurrently with RT in 58 pts (i n 7 pts, with twice daily, or h.i.d., RT) and as induction treatment, in 11 . Only 2 pts required a treatment break of more than 1 week. Median total a nd equivalent RT doses, fraction size, and anterior esophageal length mere as follows: 59.9 Gy, 59.9 Gy, 2.0 Gy, and 14 cm (range, 4.2-21). The follow ing maximum grades of esophagitis were recorded: 1, in 54 pts; 2, in 17 pts ; 3, in 13 pts, and 4, in 1 pt. The mean EI for all pts; pts treated with s tandard RT alone; induction chemotherapy and standard RT; concurrent chemot herapy and standard (QD) RT; and b.i.d. RT with concurrent chemotherapy, wa s 41.5 (range, 0-317); 13.6; 24.5; 52.4; and 132.1, respectively (p < 0.001 ). Three pts developed an esophageal structure,within 3 months beginning RT . In multivariate analysis, the following factors were significantly associ ated with increasing EI: concurrent chemotherapy with QD RT and concurrent chemotherapy with b.i.d. RT (p < 0.001, considered jointly). Both factors w ere also associated with increasing maximum esophagitis grade (p = 0.011). Esophageal length was not associated with increasing EI or esophagitis grad e in either univariate or multivariate analyses. Conclusion: Concurrent chemotherapy and twice daily radiotherapy, especiall y if combined together, were associated with the highest acute maximum esop hagitis grade and esophagitis index in pts with lung cancer. The duration o f acute esophagitis was also longest in the concurrent chemotherapy/twice d aily radiotherapy group. Esophagitis Index appeared to be a more sensitive measure of acute esophagitis than the maximum esophagitis grade. The increa sing length of esophagus in the radiation field did not predict for the sev erity of acute esophagitis. (C) 2000 Elsevier Science Inc.