Objective: To summarize the characteristics of patients with pituitary stalk thickening, analyze the association between pituitary stalk width and hypopituitarism, and develop a diagnostic model to differentiate neoplastic and inflammatory origins. Methods: A total of 325 patients with pituitary stalk thickening in a tertiary teaching hospital between January 2012 and February 2018 were enrolled. Basic characteristics and hormonal status were evaluated. Indicators to predict etiology in patients with histologic diagnoses were analyzed. Results: Of the 325 patients, 62.5% were female. Deficiency in gonadotropin was most common, followed by corticotropin, growth hormone, and thyrotropin. The increase in pituitary stalk width was associated with a risk of central diabetes insipidus (odds ratio [OR], 3.57; P<.001) and with a combination of central diabetes insipidus and anterior pituitary deficiency (OR, 2.28; P = .029). The cut-off pituitary stalk width of 4.75 mm had a sensitivity of 69.2% and a specificity of 71.4% for the presence of central diabetes insipidus together with anterior pituitary deficiency. Six indicators (central diabetes insipidus, pattern of pituitary stalk thickening, pituitary stalk width, neutrophilic granulocyte percentage, serum sodium level, and gender) were used to develop a model having an accuracy of 95.7% to differentiate neoplastic from inflammatory causes. Conclusion: Pituitary stalk width could indicate the presence of anterior pituitary dysfunction, especially in central diabetes insipidus patients. With the use of a diagnostic model, the neoplastic and inflammatory causes of pituitary stalk thickening could be preliminarily differentiated. Abbreviations: APD = anterior pituitary dysfunction; AUC = area under the curve; CDI = central diabetes insipidus; GH = growth hormone; MRI = magnetic resonance imaging; OR = odd ratio; PHBS = posterior hypophyseal bright spots; PST = pituitary stalk thickening; PSW = pituitary stalk width.
Objective: To summarize the characteristics of patients with pituitary stalk thickening, analyze the association between pituitary stalk width and hypopituitarism, and develop a diagnostic model to differentiate neoplastic and inflammatory origins. Methods: A total of 325 patients with pituitary stalk thickening in a tertiary teaching hospital between January 2012 and February 2018 were enrolled. Basic characteristics and hormonal status were evaluated. Indicators to predict etiology in patients with histologic diagnoses were analyzed. Results: Of the 325 patients, 62.5% were female. Deficiency in gonadotropin was most common, followed by corticotropin, growth hormone, and thyrotropin. The increase in pituitary stalk width was associated with a risk of central diabetes insipidus (odds ratio [OR], 3.57; P<.001) and with a combination of central diabetes insipidus and anterior pituitary deficiency (OR, 2.28; P = .029). The cut-off pituitary stalk width of 4.75 mm had a sensitivity of 69.2% and a specificity of 71.4% for the presence of central diabetes insipidus together with anterior pituitary deficiency. Six indicators (central diabetes insipidus, pattern of pituitary stalk thickening, pituitary stalk width, neutrophilic granulocyte percentage, serum sodium level, and gender) were used to develop a model having an accuracy of 95.7% to differentiate neoplastic from inflammatory causes. Conclusion: Pituitary stalk width could indicate the presence of anterior pituitary dysfunction, especially in central diabetes insipiduspatients. With the use of a diagnostic model, the neoplastic and inflammatory causes of pituitary stalk thickening could be preliminarily differentiated. Abbreviations: APD = anterior pituitary dysfunction; AUC = area under the curve; CDI = central diabetes insipidus; GH = growth hormone; MRI = magnetic resonance imaging; OR = odd ratio; PHBS = posterior hypophyseal bright spots; PST = pituitary stalk thickening; PSW = pituitary stalk width.