Endocrine Abstracts

Background: Hyperprolactinemia is a biochemical finding with a broad differential diagnosis and is commonly measured during the work up for amenorrhea, galactorrhea, or hypogonadism. Diagnostic workup of hyperprolactinemia should be performed in a stepwise fashion to avoid incorrect diagnoses and/or unnecessary investigations. As mild hyperprolactinemia (i.e. <100 mg/l) can be physiologic or spurious, repeating the prolactin level is a crucial step in the initial workup. Unnecessary or premature imaging can lead to incidental findings which generate further investigations and surveillance, causing undue burden on patients and the healthcare system. Our primary objective was to quantify the frequency and proportion of MRI sella ordered for isolated mild hyperprolactinemia when not confirmed by repeat measurement. Our secondary objective was to assess the frequency and type of incidental findings resulting from inappropriate or premature neuroimaging.

Methods: A retrospective chart review was performed between 2012 and 2022 in the province of Alberta, Canada. Potentially eligible cases were identified by a data analyst from Alberta Health Services. Inclusion criteria were 1) patients >18 years of age; 2) elevated serum prolactin <100 mg/l; 3) MRI sella performed following the detection of hyperprolactinemia. Cases with neuroimaging performed for alternative indications were excluded. We classified ordering of MRIs as ‘appropriate’ or ‘inappropriate’ based on the clinical indication and the presence or absence of repeat prolactin measurement.

Results: Initial screening identified 3,768 cases. Of these, 1,967 (52%) had two or more prolactin measurements prior to the MRI while 1,801 (48%) had a single prolactin measurement. In 823/1801 cases (46%) the indication for imaging was categorized as appropriate on the basis of clinical suspicion of pituitary or hypothalamic disease (e.g.: amenorrhea, mass effect symptoms). For the remaining cases (n=978), the ordering of neuroimaging was categorized as inappropriate. Of the MRIs performed for these cases, 61% (n=592) reported normal findings and 39% (n=386) reported one or more abnormal findings. The abnormal findings were as follows: 21% (n= 201) reported a microadenoma, 2% (n=21) reported a macroadenoma, and 17% (n=164) reported other incidental findings (e.g.: white matter changes, cysts).

Conclusion: In 26% (n=978) of mild hyperprolactinemia cases, neuroimaging was performed prematurely without repeating a prolactin measurement or without an appropriate indication. Furthermore, these scans generated a large number of incidental findings which require additional investigation and follow-up. We plan to apply our results to future systems level quality improvement initiatives.