Polyethylene glycol precipitation as a screening method for macroprolactinemia.

To the Editor: Vieira et al. (1) used the Wallac Delfia immunofluorometric assay to demonstrate that macroprolactin is a common cause of apparent hyperprolactinemia, and this confirms our experience (2) and that of others (3). However, their data validating the polyethylene glycol (PEG) precipitation as a screening method for detecting macroprolactinemia are substantially different than ours (2) and appear to be inconsistent. They suggest that a recovery of .65% of serum prolactin (PRL) after PEG precipitation indicates the absence of macroprolactin; however, their Fig. 1 shows that samples giving such recoveries contained 10–40% high-molecular weight PRL as determined by gel filtration chromatography. Furthermore, the data shown in Fig. 1 of the recovery of PRL after PEG precipitation and the proportion of PRL present as the high-molecular weight forms determined by gel filtration show considerable scatter such that a sample showing 50% recovery after precipitation with PEG might contain 15–95% macroprolactin. Their reproducibility studies also showed considerable imprecision (CV, 7–28%) for the PEG precipitation technique, and this may be one contributing factor. In my experience, the PEG precipitation test has been more reproducible (CVs, 5.8% and 6.5%) and more definitive. An initial report (2) demonstrated that in 69 cases with PRL .19.4 mg/L (700 milliunits/L), the recovery of .40% of the PRL after precipitation with PEG identified all 52 samples containing only monomeric PRL but included one sample containing 10% of the immunoreactive PRL as macroprolactin. A recovery of ,40% identified all 16 samples containing substantial quantities of macroprolactin (34–90% of the immunoreactive PRL). I have now used precipitation with PEG to examine 195 samples with total PRL .19.4 mg/L (700 milliunits/L) over a 38-month period. Using a conservative cutoff of 50%,I found low recovery after PEG precipitation in 30 samples (15.4%); the presence of macroprolactin was confirmed by gel filtration chromatography in all but one sample. Similar results were obtained in a neighboring district, with macroprolactin identified in 25 of 145 (17.2%) samples. These data are summarized in Fig. 1 and demonstrate that when macroprolactin is present it is the predominant immunoreactive form of PRL present. When no macroprolactin is present, gel filtration chromatography shows no peak of higher molecular weight PRL. The reasons for the differences between my experience with PEG precipitation and that of Vieira et al. (1) are not immediately apparent. I have used a similar technique with identical reagent concentrations but use reagents at room temperature, whereas Vieira keep their PEG at 4 °C. Temperature affects the recovery of PRL after precipitation with PEG (2), and this may be a source of variation. However, most of the differences may be related to their definition of the presence or absence of macroprolactin on gel filtration chromatography. It is not clear whether a clearly defined peak of macroprolactin was demonstrated in all samples categorized as containing high-molecular weight forms of PRL, and it would be most helpful if representative chromatograms were published so that the efficiency of the separation can be seen. In my experience, macroprolactin is a common cause of apparent hyperprolactinemia and a cause of diagnostic confusion, which may lead to inappropriate treatment. I agree with Vieira et al. (1) and Lindstedt (4) that all samples showing apparent hyperprolactinemia should be examined for macroprolactin; however, it should be noted that although all commercial assays for PRL investigated thus far react with macroprolactin (albeit to a varying extent), not all of these assays can be used with the PEG precipitation technique (2). Centrifugal ultrafiltration may provide an alternative means of demonstrating the presence of macroprolactin (5), and additional work is in progress to validate this technique.