Pre-Treatment Predictors of Esophageal Stricture Formation Following Photodynamic Therapy for Barrett's Esophagus with High-Grade Dysplasia

Abstract

Introduction: The major local adverse effect of photodynamic therapy (PDT) with porfimer sodium is esophageal stricture formation. The exact cause of stricture formation is not known but might be influenced by treatment variables including treatment length, light dosimetry and use of a centering balloon. This study was performed to assess whether pre-treatment variables, including prior endoscopic therapy for Barrett's esophagus (BE), are associated with post-PDT stricture formation. Methods: Data were reviewed for all patients who have undergone PDT with porfimer sodium for BE with high-grade dysplasia at our institution since 1997. Exclusion criteria included palliative PDT for adenocarcinoma stage T2 or greater and PDT for squamous cell carcinoma. Results: 113 patients underwent 158 PDT courses (75 patients treated with a single PDT course, 32 patients with 2 courses, 5 patients with 3 courses, and 1 patient with 4 courses). 81% were male and 99% were Caucasian, with a mean age of 70.3 at the time of index PDT. 83% of patients received additional endoscopic treatment for BE, including 28 who underwent endoscopic mucosal resection (EMR) prior to PDT. The incidence of stricture formation following index PDT was 15% (17/113). When all PDT courses were considered, the overall incidence of stricture was 23% (36/158). Stricture rate was significantly higher following a second PDT course compared with index PDT (45% versus 15%, [P = 0.0002]). Of the 17 patients who developed stricture following index PDT, 6 underwent a second PDT course, with 3 developing recurrent stricture. There was no association between age, gender, or body-mass index and development of post-PDT stricture. There was no association between prior EMR and post-PDT stricture. Patients who developed stricture had longer length BE pre-treatment than those who did not develop stricture (8.3 cm versus 5.7 cm [P = 0.005] for index PDT only; 7.5 cm versus 5.7 cm [P = 0.004] for all PDT courses). Length of BE and multiple PDT courses were independent predictors of post-PDT stricture in a stepwise logistic regression analysis controlling for treatment variables, including treatment length. Conclusions: Endoscopic therapy for BE frequently requires multiple courses of PDT or PDT as one component in a multimodality approach. EMR has gained use as a neoadjuvant therapy, and does not appear to influence the likelihood of stricture development following porfimer sodium PDT. Our data demonstrate an increased risk of stricture following multiple PDT courses. Our data also suggest an association between post-PDT stricture and length of BE, but not treatment length.