BACKGROUND: Early detection of residual disease may benefit management strategies in patients undergoing transsphenoidal surgery for acromegaly. This requires establishing objective thresholds for early postoperative growth hormone (GH) assays, and incorporating these parameters into a scale for outcome prediction. METHOD: We analyzed a database containing the records of 86 patients who had undergone gross total transsphenoidal resection of GH-secreting pituitary adenomas. Early postoperative biochemical testing included a morning fasting basal GH assay on the first postoperative day (POD1) and a second GH assay following suppression with 100 g of oral glucose on the seventh postoperative day (POD7). Remission was defined as a normal IGF-1 with either a GH nadir <0.4 ng/ml following suppression with oral glucose or a basal fasting GH <1 ng/ml on follow-up dated >3 months after surgery. Receiver operator characteristic (ROC) curves identified optimal thresholds for all biochemical parameters. Logistic regression analysis assessed the statistical significance of factors associated with cure. A point system was developed, employing regression coefficients obtained from the multivariate statistical model to quantify the impact of each predictor on cure. RESULTS: Remission was achieved in 34.6 % of patients and was associated with smaller, non-invasive tumors with lower preoperative, POD1 and POD7 GH levels. Optimal thresholds obtained from the ROC analysis suggested that lower POD1 and POD7 GH values provided good sensitivity and specificity for cure, despite modest predictive values. The model with the best ability to predict outcome included size, POD1 GH and POD7 GH levels, with a score of ≥95 demonstrating high specificity for prediction of remission. CONCLUSION: Early postoperative GH assays are highly sensitivity and specific. The scoring system that we propose provided excellent predictive value and requires further validation in larger cohorts and in different populations. The model may help guide the intensity of follow-up and enable early identification of residual disease.
BACKGROUND: Early detection of residual disease may benefit management strategies in patients undergoing transsphenoidal surgery for acromegaly. This requires establishing objective thresholds for early postoperative growth hormone (GH) assays, and incorporating these parameters into a scale for outcome prediction. METHOD: We analyzed a database containing the records of 86 patients who had undergone gross total transsphenoidal resection of GH-secreting pituitary adenomas. Early postoperative biochemical testing included a morning fasting basal GH assay on the first postoperative day (POD1) and a second GH assay following suppression with 100 g of oral glucose on the seventh postoperative day (POD7). Remission was defined as a normal IGF-1 with either a GH nadir <0.4 ng/ml following suppression with oral glucose or a basal fasting GH <1 ng/ml on follow-up dated >3 months after surgery. Receiver operator characteristic (ROC) curves identified optimal thresholds for all biochemical parameters. Logistic regression analysis assessed the statistical significance of factors associated with cure. A point system was developed, employing regression coefficients obtained from the multivariate statistical model to quantify the impact of each predictor on cure. RESULTS: Remission was achieved in 34.6 % of patients and was associated with smaller, non-invasive tumors with lower preoperative, POD1 and POD7 GH levels. Optimal thresholds obtained from the ROC analysis suggested that lower POD1 and POD7 GH values provided good sensitivity and specificity for cure, despite modest predictive values. The model with the best ability to predict outcome included size, POD1 GH and POD7 GH levels, with a score of ≥95 demonstrating high specificity for prediction of remission. CONCLUSION: Early postoperative GH assays are highly sensitivity and specific. The scoring system that we propose provided excellent predictive value and requires further validation in larger cohorts and in different populations. The model may help guide the intensity of follow-up and enable early identification of residual disease.
Authors: Adriana G Ioachimescu; Talin Handa; Neevi Goswami; Adlai L Pappy; Emir Veledar; Nelson M Oyesiku Journal: Endocrine Date: 2019-11-01 Impact factor: 3.633
Authors: Tyler Cardinal; Casey Collet; Michelle Wedemeyer; Peter A Singer; Martin Weiss; Gabriel Zada; John D Carmichael Journal: Front Endocrinol (Lausanne) Date: 2021-11-18 Impact factor: 5.555
Authors: V Cambria; G Beccuti; N Prencipe; F Penner; V Gasco; F Gatti; M Romanisio; M Caputo; E Ghigo; F Zenga; S Grottoli Journal: J Endocrinol Invest Date: 2021-04-10 Impact factor: 4.256