Background: Prolactin levels are affected by stress, and in patients with moderate hyperprolactinaemia, a repeat test and/or a resting prolactin has been recommended, but there are very few data addressing the utility of these additional measurements.
Aim: To study the value of: A) Repeat measurement and B) Resting measurement of serum prolactin in mild to moderate prolactin excess (5107500 IU/L).
Methods and subjects: Case note review of 75 patients referred for evaluation of high prolactin (referral prolactin, P0). All patients had a repeat sample done at our investigation unit immediately following insertion of a forearm cannula (repeat prolactin, P1) and a second sample drawn after 120 min bed rest (resting prolactin, P2). P1 and P2 values were corrected for macroprolactin to achieve an estimate of monomeric prolactin. 14/75 patients who were on dopamine receptor antagonists were excluded from the analysis.
Results: P0 median was 1229, range 561 to 5949 IU/L. P1 was normal in 17/61 patients (27%), one due to macroprolactin; amongst these 17, P0 median was 892 IU/L (range 6161800). Of the remaining 44, eight showed normalisation of resting prolactin, P2; amongst these the P0 median was 1445 IU/L (range 5612412). However, the highest P1 in those in whom hyperprolactinaemia was excluded was 1062 IU/L.
|P0||P1||P2||No. (%)||P0 median (range) IU/L|
|Repeat normal||Hi||Lo||Lo||17 (28%)||892 (6161800)|
|Resting normal||Hi||Hi||Lo||8 (13%)||1445 (5612412)|
|Confirmed||Hi||Hi||Hi||36 (59%)||1484 (5775949)|
Conclusions: 41% of patients with elevated prolactin normalised on further testing. Most of these false positive results would have been corrected by a simple repeat sample. A resting prolactin is probably unnecessary if a simple repeat value is twice the upper limit of normal.
01 - 05 Apr 2006
European Society of Endocrinology