Background and study aims: This retrospective study reports 12 years' experience with pneumatic dilation treatment in patients with achalasia and attempts to define factors capable of predicting failure of endoscopic dilation.
Patients and methods: Consecutive patients with achalasia who received endoscopic balloon dilation were studied retrospectively. Repeat dilation was carried out if dysphagia persisted or recurred. A structured symptom score questionnaire (the Eckardt score) was conducted by phone with patients who had received dilation and had been followed up for more than 2 years. Failure was defined as the presence of significant dysphagic symptoms after more than two repeat dilations. Data for the first 2 years (short-term) and for the subsequent follow-up (long-term) were analyzed.
Results: From 1989 to 2001, 66 patients underwent endoscopic balloon dilation for achalasia; three perforations (4.5 %) occurred, with no mortalities. Dysphagic symptoms significantly improved 12 weeks after the procedure ( P < 0.05). Fourteen patients (20 %) required a second dilation procedure within a median of 7 months (range 1 - 52 months), and 13 of them underwent repeat dilations within the first 2 years. Five patients (7.5 %) required further surgical or endoscopic therapy. Fifty-eight patients received pneumatic dilation for more than 2 years; 32 (55 %) responded to the questionnaire. The mean dysphagia score was 1.7 (SD 1.2), with only five patients (16 %) having significant dysphagic symptoms during a median follow-up period of 55 months (range 26 - 130 months). The cumulative success rates for pneumatic dilation after 5 and 19 years were 74 % and 62 %, respectively. Cox regression analysis identified small balloon size (30 mm) as the only significant factor capable of predicting failure of endoscopic dilation ( P = 0.009; relative risk 5.3; 95 % confidence interval, 1.7 to 40.9).
Conclusions: Endoscopic balloon dilation is an effective treatment for achalasia, with minimal morbidity (60 % experience long-term benefit).