PURPOSE

To evaluate the utility of Insulin-like growth factor I (IGF-I) standard deviation score (SDS) as a surrogate marker of severity of hypopituitarism in adults with pituitary pathology.

METHODS

We performed a retrospective data analysis, including 269 consecutive patients with pituitary disease attending a tertiary endocrine clinic in 1990-2015. The medical files were reviewed for the complete pituitary hormone profile, including IGF-I, and clinical data. Age-adjusted assay reference ranges of IGF-I were used to calculate IGF-I SDS for each patient. The main outcome measures were positive and negative predictive values of low and high IGF-I SDS, respectively, for the various pituitary hormone deficiencies.

RESULTS

IGF-I SDS correlated negatively with the number of altered pituitary axes (p < 0.001). Gonadotropin was affected in 76.6 % of cases, followed by thyrotropin (58.4 %), corticotropin (49.1 %), and prolactin (22.7 %). Positive and negative predictive values yielded a clear trend for the probability of low/high IGF-I SDS for all affected pituitary axes. Rates of diabetes insipidus correlated with IGF-I SDS values both for the full study population, and specifically for patients with non-functioning pituitary adenomas.

CONCLUSIONS

IGF-I SDS can be used to evaluate the somatotroph function, as a valid substitute to absolute IGF-I levels. Moreover, IGF-I SDS predicted the extent of hypopituitarism in adults with pituitary disease, and thus can serve as a marker of hypopituitarism severity.