Determinants of survival in treated acromegaly in a single center: predictive value of serial insulin-like growth factor I measurements

We evaluated survival after optimal treatment for acromegaly and assessed the predictive effects of different remission criteria for survival in 164 consecutive acromegalic patients, treated by transsphenoidal surgery and adjuvant therapy between 1977 and 2002. The goal of treatment was a mean GH less than 5 mU/liter, a normal glucose-suppressed GH, and a normal IGF-I for age in all patients. Surgery initially cured 108 patients (66%). Adjuvant therapy for persistent disease was given to 49 patients. At the end of follow-up (mean, 12.3 yr), remission rates for surgery and multimodality treatment were 54% and 90%, respectively. In 2033 person-years of follow-up, 28 of 164 patients died, resulting in an observed:expected mortality ratio of 1.3 (confidence interval, 0.87-1.87). Significant predictors for survival were the duration of disease and the postoperative glucose-suppressed GH. The effects of these predictors became less significant with increasing follow-up duration. A time-dependent effect on survival was observed for serial IGF-I concentrations, but not for serial GH concentrations. Of the three remission criteria, IGF-I was the only one to be significantly associated with survival in this study, with a relative risk of 4.78 for an elevated as opposed to a normal IGF-I concentration