A Multicenter Study of Clinical Outcomes and Volumetric Trends in Microprolactinomas

Introduction: The diagnosis of pituitary microprolactinomas is often obscured by relatively low levels of elevated prolactin compared to macroprolactinomas. This may lead to varying patterns of medical therapy versus observation. We sought to correlate prolactin levels in suspected microprolactinomas with tumor volumes and clinical outcomes.

Methods: This was a multicenter retrospective study of patients with pituitary microadenomas with baseline prolactin levels >18 ng/mL for males and >30 ng/mL for females. A linear-mixed model was used to depict changes in tumor volume over time.

Results: There were 65 patients (18 males, 47 females; mean age: 44 years, range: 15–74). At the time of diagnosis, mean tumor volume was 95.9 mm³ (SD: 114.2 mm³, range: 4–500 mm³), and mean prolactin level was 59.4 ng/mL (SD: 34.4 ng/mL, range: 19.1–170 ng/mL). When analyzing the whole cohort, there was no significant correlation between prolactin levels and tumor volume at diagnosis (Pearson R = 0.09, p = 0.47). Upon sub-group analysis of patients with tumor volumes above (n = 19) and below (n = 46) the mean, there were significantly higher prolactin levels in patients with tumors above the mean volume of 95.9 mm³ vs. below (74.0 vs. 53.4 ng/mL, 95% CI: 2.432–38.8, p = 0.027).

Twenty-six patients were observed, 31 were treated with anti-dopaminergic therapy, and 8 had surgery (4 had surgery after failing medical therapy). There were significantly greater baseline prolactin levels for patients who treated surgically (mean: 86.4 ng/mL, SD: 39.6) than those treated medically (mean: 61.7 g/mL, SD: 38.2), or were observed (mean: 48.5 ng/mL, SD: 22.0; p = 0.02; F = 4.2; R ² = 0.12). However, there were no significant differences in baseline volumes for patients treated surgically (mean: 120.0 mm³, SD: 122.4), medically (mean: 111.2 mm³, SD: 118.2), or observed (mean: 70.24 mm³, SD: 106.1) (p = 0.33; F = 1.1; R ² = 0.04).

Among the 26 patients who were surveilled, 13 patients demonstrated spontaneous tumor shrinkage, 12 remained stable, and 1 patient’s tumor grew but was lost to follow-up. Linear mixed modeling demonstrated a statistically significant rate of tumor shrinkage over time of 3.67 mm³/year (p = 0.03; Fig. 1).

Fig. 1

There were no significant differences in baseline prolactin in patients whose tumors spontaneously shrunk (mean: 46.1 ng/mL, SD: 22.3) versus remained stable (mean: 52.3 ng/mL, SD: 22.7) (95% CI: −12.4 to 24.9, p = 0.50). Likewise, there were no significant differences in baseline volumes (mean: 56.2 mm³, SD: 65.1 vs. 91.0 mm³, SD: 140.9, respectively; 95% CI: −54.8 to 124.4, p = 0.43).

A baseline PRL/volume ratio metric was then explored to predict clinical outcomes. When analyzing patients who were observed versus those requiring surgery after initially being surveilled, there were significantly greater baseline PRL/volume ratios in surgical patients versus those observed (8.1 vs. 2.4, 95% CI: 0.80–10.7, p = 0.025). However, PRL/volume ratio did not predict patterns of medical treatment failure for patients successfully managed with medical treatment versus those who failed and required surgery (2.25 vs. 0.91, 95% CI: −7.7 to 5.03, p = 0.67).

Conclusions: Suspected microprolactinomas may demonstrate more convincingly elevated prolactin levels when measuring over 95.9 mm³. Tumors with baseline prolactin levels over 50 ng/mL may be more inclined to undergo medical treatment. In tumors with levels below 50 ng/mL, it may be reasonable to undergo surveillance as these tumors tend to spontaneously shrink over time. In tumors that are surveilled, an elevated baseline PRL/volume ratio of >8 may be indicate serial tumor growth that may necessitate medical and/or surgical intervention.

Publication History

Article published online:
05 February 2024