← Back to Previous Page

 

S046: RECONSIDERING DIAGNOSTIC CUTOFFS FOR PROLACTIN IN PITUITARY ADENOMAS: IMPACT OF SEX, AGE, AND TUMOR SIZE
Maya Harary; Mishek Thapa; Stuart Harper; Ines Donangelo; Anthony Heaney; Won Kim; Marvin Bergsneider; UCLA

Objective: To assess the relationship between prolactin levels (PRL), and patient sex, age and tumor size in a large surgical cohort of prolactinomas and clinically non-functioning pituitary tumors (CNFs), testing the adequacy of commonly referenced diagnostic PRL cutoffs.

Methods: Medical records from a quaternary referral center were reviewed for patients undergoing first-time resection of a pituitary neuroendocrine tumor (PitNET) with pathologic diagnosis of prolactinoma or CNF. Patients with ACTH- or GH-secreting tumors were excluded. Maximum preoperative PRL (ng/mL) was recorded. Preoperative MRI was reviewed for maximum diameter, and volume was estimated by the ABC/2 method. Statistical analyses were performed in R (v4.5.1).

Results: A total of 459 patients met inclusion criteria, including 119 prolactinomas (25.9%) and 340 CNFs (74.1%); 238 were female (51.9%) and 221 male (48.1%, Table 1).

• Prolactin levels: Median maximum PRL was higher in prolactinomas than CNFs—female: 100.9 ng/mL (IQR 63.4–181.8) vs 26.8 (10.9–56.2); male: 157.0 (48.3–1200.0) vs 16.3 (10.2–25.3); all p<0.001—yet with substantial overlap between groups.
• Sex differences: In CNFs, women had higher PRL than men (median 26.8 vs 16.3 ng/mL, p<0.001, Figure 1A). After normalizing PRL to tumor volume, females also had significantly higher values than males in both CNF and prolactinoma groups (both p<0.001, 1B).
• Tumor size: Prolactinomas were smaller in women than men at surgery (median diameter 9 mm vs 19 mm, p<0.001, 1C), consistent with earlier, symptom-driven detection (e.g., amenorrhea). CNFs presented as macroadenomas of comparable size in both sexes. There were distinct sex-specific patterns in PRL to tumor size relationship per pathology (1D).
• Age correlations: PRL negatively correlated with age in women for both prolactinomas (R≈–0.30, p=0.004) and CNFs (R≈–0.30, p<0.001) (Figure 2). In men, correlations were weaker and not significant (CNF R≈–0.15; prolactinoma R≈–0.30).
• Tumor size and age: When PRL was examined relative to tumor size across age quartiles, distinct sex-specific patterns emerged (Figure 3). Men and women demonstrated differing trends in both prolactinoma and CNF groups, highlighting the complex interplay of biology, tumor burden, and age in shaping PRL secretion.
• Outliers: Although categorized as CNFs for analytic clarity, mammosomatotrophs (PIT1-lineage) tumors sometimes secreted PRL at prolactinoma-like levels, representing notable outliers within the CNF group.

Conclusion: PRL secretion in pituitary adenomas is not determined by pathology alone; sex, age, and tumor size materially shape serum levels, creating overlap between prolactinomas and CNFs. A single diagnostic PRL cutoff is therefore inadequate. Integrating patient sex, age, and tumor size into PRL interpretation may improve diagnostic accuracy, minimize unnecessary dopamine-agonist trials in non-prolactinomas, expedite neurosurgical referral when appropriate, and enhance patient counseling.

 

Figure 1:

 

Figure 2:

Figure 3:

 

← Back to Previous Page