Predictors of outcome in thymectomy for myasthenia gravis

Background. Factors determining predictability of response to thymectomy fo r myasthenia gravis (MG) vary in the literature. Methods. A 25-year retrospective review (1974 to 1999) of all thymectomies performed at a single institution was undertaken. Results. In 113 consecutive thymectomies for MG, women comprised 79% (89 of 113 patients), and mean age was 40 +/- 15 years. Complications occurred in 14% of patients (16 of 113). In-hospital mortality was 0, but 90-day hospi tal mortality was 0.88% (16 of 113 patients). Follow-up was obtained in 81% (92 of 113 patients) at a mean of 51 +/- 59 months postoperatively. Comple te remission was achieved in 21% of patients (19 of 92), and marked improve ment of MG in 54% (50 of 92), for a total benefit rate of 75%. Fourteen per cent (13 of 92) were unchanged, and 11% (10 of 92) were worse. Using univar iate analysis, sex, age, and pathology correlated significantly with outcom e (p < 0.05): 80% of women (57 of 70) benefited from the procedure, versus 57% of men (12 of 21). Eighty percent (57 of 70) of patients less than 51 y ears of age were improved or in remission, versus 57% (12 of 22) older than 50. Twenty-three percent (5 of 22) of patients with thymoma deteriorated, versus 7.1% (5 of 70) without thymoma. Sex did not significantly correlate in the multivariate model. Conclusions. Sex, age, and thymic pathology are potential predictors of out come in thymectomy for MG, and may shape treatment decisions and target hig her-risk patients. (C) 2001 by The Society of Thoracic Surgeons.