Central diabetes insipidus and pituitary stalk thickening in adults: distinction of neoplastic from non-neoplastic lesions

Abstract

Context Association of central diabetes insipidus (CDI) and pituitary stalk thickening (PST) may have several etiologies (including malignancies) and differential diagnosis remains often difficult. Objective The purpose of this study was to identify which clinical, biochemical or radiological features could help clinicians to make an etiological diagnosis, especially distinguishing neoplastic from non-neoplastic pituitary stalk lesions. Design and methods We retrospectively analyzed clinical, biochemical, radiological and histological data of 38 adult patients diagnosed with CDI and PST of proven etiology. Results Of the 38 pituitary stalk lesions included, 11 (29%) were neoplastic. A histopathological diagnosis was obtained in 22/38 (58%) patients. The three most frequently observed etiologies of PST were neuroinfundibulitis (34%), germinoma (21%) and histiocytosis (18%). Pituitary stalk thickness was larger for neoplastic lesions, particularly germinomas. Male gender and a very young age were statistically associated with a risk of germinoma. At least one anterior pituitary deficit was observed in nearly 60% of patients. Patients with neoplastic PST were more affected by multiple anterior pituitary dysfunction than patients with benign PST. A high serum prolactin level was individually the best predictor of a neoplastic origin (90% sensitivity and 60% specificity for a serum prolactin level 1.27-fold above the normal upper limit (ULN)). Conclusion We confirm a relatively high risk of malignancy in adult patients presenting with the association of CDI and PST. Young age, male gender, a very large thickening of the stalk, multiple anterior pituitary deficits and prolactin above 1.3× ULN increase the likelihood of a neoplastic origin.

Pituitary stalk thickening: neoplastic or not? - author’s response to the letter by Wang et al.

“…In patients with PST and CDI, one of the main goals of the medical workup should be to exclude a neoplastic underlying pathology. Others and we showed that PS thickness was slightly larger in patients with malignant disease (12,13) but a large overlap exists with benign lesions, making this criterion of limited use in the differential diagnosis (1). When analyzing the weak correlation between the thickness of pituitary stalk and serum PRL concentration in our series (r = 0.363, P = 0.063), one can easily observe that prolactin was much higher in most malignant cases than in patients with benign disease, independent of the size of the stalk (Fig.…”

“…We thank Dr Shirui Wang and colleagues for their interest in our recent publication on the distinction of neoplastic from non-neoplastic lesions in adult patients presenting with central diabetes insipidus (CDI) and pituitary stalk thickening (PST) (1). We acknowledge the fact that they could not confirm in their population our previous observation that a higher prolactin (PRL) concentration at diagnosis increases the likelihood of a neoplastic origin (2).…”

“…First, the definition of PST was not exactly the same in the two studies. While they used a constant cut-off of PS thickness ≥ 3 mm along the stalk (3), we rather used differential cut-offs based on the previously reported normal dimensions of the hypothalamic-pituitary axis on MRI (1,4,5,6,7,8). More specifically, the pituitary stalk was considered enlarged when its largest diameter was greater than 4.0 mm in its proximal portion at the level of the optic chiasm, 2.8 mm in its medial portion and/or 2.5 mm in its distal portion at the level of pituitary insertion.…”

Pituitary Stalk Enlargement in Adults

“…Hormonal deficiencies such as DI, central hypocortisolism and central hypothyroidism need to be explored and treated. Data obtained in our cases of DI associated with enlargement of the PS may indicate that tumoral causes of PS enlargement are more prone to be associated with pituitary dysfunctions and increased PRL levels compared with nonneoplastic causes [64]. The potential mass effects (visual field defects, cranial nerve involvement) also need to be evaluated.…”

Reevaluation of the value of prolactin in differentiating malignant from benign pituitary stalk thickening: Letter to the Editor

“…It remained controversial about the classification of Langerhans cell histiocytosis. For comparison, we classified 'L' type histiocytosis as benign lesions as Devuyst did (1). Since the definition of PRL ULN was not reported in the Devuyst's study, we set the ULN as 25 μg/L for women, and 20 μg/L for men according to Guidelines of the Pituitary Society for the diagnosis and management of prolactinomas (3).…”