Early postoperative growth hormone levels: high predictive value for long-term outcome after surgery for acromegaly

Citation
S. Valdemarsson et al., Early postoperative growth hormone levels: high predictive value for long-term outcome after surgery for acromegaly, J INTERN M, 247(6), 2000, pp. 640-650
Citations number
37
Categorie Soggetti
General & Internal Medicine","Medical Research General Topics
Journal title
JOURNAL OF INTERNAL MEDICINE
ISSN journal
09546820 → ACNP
Volume
247
Issue
6
Year of publication
2000
Pages
640 - 650
Database
ISI
SICI code
0954-6820(200006)247:6<640:EPGHLH>2.0.ZU;2-I
Abstract
Objectives. To explore the prognostic value of early - within 1 week - post operative growth hormone (GH) measurements with regard to outcome after sur gery for acromegaly in a short- and a long-term perspective. Design. Retrospective study of patients operated on between 1987 and 1998, including follow-up for up to 60 months. Setting. University hospital. Subjects. Sixty-eight patients with acromegaly. Intervention. Pituitary surgery aiming at adenomectomy with preservation of pituitary function. Main outcome measures. The effect of the operation was evaluated after 3 mo nths, mostly by means of an oral glucose load or by insulin-like growth fac tor 1 (IGF-1). The specificity, sensitivity and the predictive values of an early postoperative mean GH concentration less than or equal to 4.8 mU L-1 , as well as of the GH response to thyrotropin-releasing hormone (TRH) 3 mo nths after surgery, were calculated with regard to outcome of the operation in both 3-month and long-term perspectives. Results. Fifty patients (73.5%) showed a satisfactory effect at the evaluat ion 3 months postoperatively; 45 of these were followed between 12 and 60 m onths. Relapse was registered in five cases: 12, 12, 24, 24 and 48 months a fter surgery. In the long-term perspective, the predictive value of an early mean GH less than or equal to 4.8 mU L-1 was 93.6% with regard to a satisfactory effect of surgery, compared with 90.2% for a normalized somatomedin C (SmC)/IGF-1 and 90.0% for an absent GH response after TRH. An early mean GH > 4.8 mU L -1 had a 77.8% predictive value for persistent or recurrent disease, compar ed with 85.7% for persistently increased SmC/IGF-1 and 68.8% for an abnorma l GH release after TRH 3 months after surgery. In the short-term perspective, the specificity and the predictive value of an early GH less than or equal to 4.8 mU L-1 were 77.3 and 97.1%, respectiv ely. Early GH > 4.8 mU L-1 had a 94.4% sensitivity but a predicative value of only 63.0% for an unsatisfactory effect. Conclusion. Measurement of GH within 1 week after surgery is highly predict ive for outcome of surgery for acromegaly. Specifically, an early mean GH l ess than or equal to 4.8 mU L-1 is as predictive for a satisfactory effect of treatment as a normalized IGF-1 3 months after surgery. Early postoperat ive GH values > 4.8 mU L-1 have a high sensitivity for persistent or recurr ent disease in both the short- and long-term perspectives, but lower predic tive value. The usefulness of the TRH test can be questioned.