Roberto Salvatori

Pituitary stalk thickening (PST) in adults may arise from neoplastic, inflammatory, infectious, or congenital disease, yet the underlying cause frequently remains undetermined after initial evaluation. For patients without an established diagnosis, no adult-specific data define how long magnetic resonance imaging (MRI) surveillance should be maintained. Current expert reviews recommend early reassessment at 3-6 months, followed by longitudinal imaging for approximately 3-5 years. Among initially indeterminate lesions, spontaneous regression is more likely than progression, although late diagnosis of histiocytic disease can nevertheless occur. We suggest reducing MRI frequency after sustained stability and considering discontinuation only in carefully selected low-risk adults, while maintaining longer surveillance whenever endocrine, radiological, or systemic risk features persist.

Pituitary disorders often include complex radiology imaging, rare clinical presentations, and need for multidisciplinary decision-making and prolonged follow-up, all features that make them a natural target for artificial intelligence (AI). The resulting literature is expanding rapidly, but carries an unfamiliar vocabulary and set of methods, which can leave clinicians without a clear entry point. This short perspective offers such an entry point, succinctly outlining how AI learns from data, how a clinical AI application moves from question to deployment, providing examples, and addressing how to read the field critically.

Context: Macroprolactinemia is a well-recognized cause of hyperprolactinemia and an important diagnostic pitfall in endocrine practice. However, interpretation of published quantitative prolactin data remains sparse as studies vary in confirmation method, assay platform, polyethylene glycol (PEG) recovery cutoff, and reporting of prolactin measurement. Evidence Acquisition: PubMed, Embase, Scopus, Web of Science, the Cochrane Library, and Google Scholar were systematically searched. Eligible studies reported macroprolactinemia-specific quantitative prolactin data in patients with confirmed macroprolactinemia defined by PEG precipitation, gel filtration chromatography (GFC), or both. Two reviewers independently performed study selection, data extraction, and quality assessment. Findings were summarized using study-level descriptive synthesis. The review was prospectively registered in PROSPERO and conducted in accordance with PRISMA 2020 guidelines. Evidence Synthesis: Forty-five studies encompassing 2853 macroprolactinemia cases from 21 413 screened patients with hyperprolactinemia across 22 countries were included. Among 33 studies eligible for primary quantitative analysis, the median study-level central total prolactin attributed to macroprolactinemia was 61.4 ng/mL ([IQR] 42.0-80.0; range 28.1-137.6), and the median study-level post-PEG monomeric prolactin was 11.7 ng/mL (IQR 8.3-13.2; range 4.0-17.0)). The median study-level maximum total prolactin was 264.5 ng/mL (IQR 97.0-425.5; range 81.8-663.0); extreme elevations were attributable to coexisting prolactinomas. Conclusion: In confirmed macroprolactinemia, total prolactin elevation is typically moderate, and post-PEG monomeric prolactin is usually within or near the normal range. The post-PEG monomeric prolactin value, rather than percent recovery alone, is the most informative parameter for distinguishing isolated macroprolactinemia from coexisting true hyperprolactinemia. These quantitative benchmarks may help clinicians to avoid unnecessary investigation or treatment.

Purpose The immunoassay “hook effect” may cause spuriously low prolactin (PRL) measurements in patients with macroprolactinomas, leading to misdiagnosis and inappropriate management. This systematic review aimed to determine the tumor size threshold and distribution at which the hook effect has been documented, characterize the pre-dilution PRL pattern, quantify the magnitude of PRL underestimation, and assess the clinical consequences of unrecognized hook effect. Methods PubMed, Embase, Scopus, Google Scholar, and Web of Science were searched for articles published between January 2000 and August 2025. Studies reporting individual patient data with confirmed hook effect (pre- and post-dilution PRL, tumor size, and confirmation by serial dilution) were included. Quality was assessed using the JBI Critical Appraisal Checklist for Case Reports. Data were synthesized using individual patient data (IPD) quantitative methods. The review was registered on PROSPERO (CRD420261375863) and reported per PRISMA 2020 guidelines. Results Twenty-three patients from 14 studies across 9 countries were included, with a median maximum tumor diameter of 48.0 mm (IQR 35.5–55.0; range 22–100 mm); the smallest confirmed hook-effect tumor diameter measured 22 mm, reported in two independent cases, and 9 of 23 cases (39.1%) occurred in tumors below the conventional giant adenoma threshold (≥ 40 mm). Median pre-dilution PRL was 100.0 ng/mL (IQR 62.0–164.5), while median post-dilution PRL was 6,460.0 ng/mL (IQR 3,608.5–16,657.5), yielding a median fold increase of 64.7× (IQR 46.8–141.5). Pre-dilution PRL ranged from 26 to 200 ng/mL in 82.6% of cases (95% CI: 62.9–93.0%), values often misattributed to stalk effect. Unrecognized hook effect led to unnecessary surgery in 55.6% of cases with documented clinical outcomes. The hook effect was documented across multiple immunoassay platforms. Conclusions Pre-dilution PRL values often overlap with the range attributed to stalk effect, and serial dilution testing should be considered for pituitary macroadenoma with PRL disproportionately low relative to tumor size, regardless of whether the tumor meets the giant adenoma threshold to prevent misdiagnosis, unnecessary surgical intervention and delayed initiation of dopamine agonist therapy.

Retinoblastoma protein (Rb) is a tumor suppressor whose inactivation causes retinoblastomas, pineoblastomas, and possibly other central nervous system tumors. Although mice with Rb1 gene inactivation invariably develop pituitary adenomas, and some evidence suggests that a subset of pituitary adenomas have lack or reduced expression of Rb protein, to date no evidence has been reported that patients with germline mutations in the RB1 gene (that encodes for Rb protein) are at risk of developing pituitary adenomas. Here, we report the case of a patient with childhood (age 1 year) onset retinoblastoma because of a germline pathogenic variant of the RB1 gene who presented with a somatotropinoma diagnosed at a young age (20) and whose pituitary adenoma cells showed loss of expression of Rb protein by immunohistochemistry, suggesting a role of Rb in preventing the development of pituitary adenomas.

BACKGROUND: Inferior petrosal sinus sampling (IPSS) is an established diagnostic technique used to distinguish between ectopic adrenocorticotropic hormone (ACTH) production and Cushing disease. The utility of IPSS outside of the differential diagnosis of ACTH-dependent hypercortisolism, however, is rarely reported. In this study, the authors describe the use of IPSS to lateralize a prolactin (PRL)–secreting adenoma in a patient with hyperprolactinemia and two distinct pituitary lesions. OBSERVATIONS: A 37-year-old female presented with symptomatic hyperprolactinemia that was refractory to escalating dopaminergic therapy. Sellar MRI demonstrated an 8-mm right-sided pituitary lesion and a distinct 5-mm left-sided pituitary lesion. IPSS was performed with PRL and ACTH measurements. Lateralization was assessed using both absolute PRL levels and PRL levels normalized to ACTH, and these data indicated right-sided PRL hypersecretion. The patient underwent successful endonasal endoscopic resection of the right-sided lesion, which was immunohistochemically positive for PRL, without removal of the left-sided lesion. She remained in remission 2.5 years after surgery, without evidence of hypopituitarism. LESSONS: This case highlights the potential utility of IPSS in selected patients with hyperprolactinemia and bilateral pituitary tumors. This technique may be helpful in guiding surgical decision-making in prolactinoma patients with multiple sellar masses. https://thejns.org/doi/10.3171/CASE25511

Inferior petrosal sinus sampling (IPSS) is an established diagnostic technique used to distinguish between ectopic adrenocorticotropic hormone (ACTH) production and Cushing disease. The utility of IPSS outside of the differential diagnosis of ACTH-dependent hypercortisolism, however, is rarely reported. In this study, the authors describe the use of IPSS to lateralize a prolactin (PRL)–secreting adenoma in a patient with hyperprolactinemia and two distinct pituitary lesions. A 37-year-old female presented with symptomatic hyperprolactinemia that was refractory to escalating dopaminergic therapy. Sellar MRI demonstrated an 8-mm right-sided pituitary lesion and a distinct 5-mm left-sided pituitary lesion. IPSS was performed with PRL and ACTH measurements. Lateralization was assessed using both absolute PRL levels and PRL levels normalized to ACTH, and these data indicated right-sided PRL hypersecretion. The patient underwent successful endonasal endoscopic resection of the right-sided lesion, which was immunohistochemically positive for PRL, without removal of the left-sided lesion. She remained in remission 2.5 years after surgery, without evidence of hypopituitarism. This case highlights the potential utility of IPSS in selected patients with hyperprolactinemia and bilateral pituitary tumors. This technique may be helpful in guiding surgical decision-making in prolactinoma patients with multiple sellar masses. https://thejns.org/doi/10.3171/CASE25511

Context: There are limited head-to-head trials comparing pharmacological treatments for acromegaly. Objective: Systematically review the efficacy and safety of pharmacological treatments for acromegaly and conduct a network meta-analysis (NMA) enabling indirect comparisons. Methods: MEDLINE and Embase were searched to identify randomized controlled trials (RCTs) of acromegaly therapies. Screening and data extraction followed PRISMA guidelines. Feasibility assessment evaluated homogeneity and consistency assumptions required for NMA. Bayesian NMAs estimated relative treatment effects and ranking probabilities. Results: Twenty-two records covering 18 RCTs were included. Biochemical control rates were comparable among long-acting injectable somatostatin receptor ligands (SRLs), including lanreotide autogel (LAN-ATG), octreotide long-acting release (OCT-LAR), pasireotide, the GH receptor antagonist pegvisomant, oral octreotide (O-OCT), octreotide subcutaneous depot (SC-OCT-D), and the once-daily oral SRL paltusotine. Paltusotine demonstrated significantly higher biochemical control vs O-OCT and SC-OCT-D (odds ratios [ORs], 95% credible intervals [CrIs]: 7.34 [1.48-36.07] and 7.85 [1.72, 36.25]). Pasireotide showed significantly higher biochemical control vs OCT-LAR (OR: 2.03 [1.29-3.23]). Paltusotine had significantly lower discontinuations due to adverse events (AEs) vs O-OCT and SC-OCT-D, (ORs: 0.022 [0.001-0.424] and 0.022 [0.001-0.343]), with similar rates to other treatments. Treatment-emergent AEs (TEAEs) and serious TEAEs were comparable across treatments. Rankings suggested paltusotine as the treatment with the highest probability of ranking as the most effective (or tolerable) treatment across all endpoints studied. Conclusion: This systematic review and NMA consolidate recent high-quality RCT evidence for acromegaly treatments. Paltusotine emerges as a promising alternative to injectable SRLs, with favorable efficacy, safety, and AE-related discontinuation patterns. These findings may inform clinical decision-making and guideline development, if confirmed by clinical experience.