BACKGROUND: High homology of GH with placental GH (pGH) and hPL frequently resulted in falsely high GH levels in competitive immunoassays during pregnancy. However, in immunometric assays, falsely high or low GH levels can result from GH-like molecules binding to both or only one monoclonal antibody. Since our GH-IFMA assay detected GH suppression in both normal and acromegalic pregnancies, we evaluated potential negative interference of pregnancy serum in the assay. METHODS: GH was measured in samples from acromegalic patients with and without the addition of normal pregnancy serum using a sensitive in-house two-step GH-IFMA (no crossreactivity with pGH, Prolactin or hPL). Standard GH assay curves were run with and without pGH (20 and 22 K). Pegvisomant, a GH-antagonist with high homology to GH, was also tested for cross-reactivity. RESULTS: Addition of pregnancy serum to acromegaly serum resulted in marked decrease in GH, but addition of pGH did not change GH measurements. Redesign of the routine assay by switching the positions of the antibodies ("inverted" assay) completely abrogated the interference of pregnancy serum. GH by both routine and inverted assays declined progressively throughout pregnancy in controls, with higher nadir levels in the "inverted" assay (median 0.03 μg/L vs 0.50 μg/L, P<0.05). GH suppression during acromegalic pregnancy previously found with the routine assay was not observed in the "inverted" assay. Pegvisomant does not cross-react with GH in the "inverted" assay. CONCLUSIONS: GH measurements in pregnancy by immunometric assays must be made after exclusion of pregnancy serum interference by dilutional tests. Redesigning a two-step immunometric GH assay by switching the positions of the antibodies can be a successful strategy to abrogate such interference.
BACKGROUND: High homology of GH with placental GH (pGH) and hPL frequently resulted in falsely high GH levels in competitive immunoassays during pregnancy. However, in immunometric assays, falsely high or low GH levels can result from GH-like molecules binding to both or only one monoclonal antibody. Since our GH-IFMA assay detected GH suppression in both normal and acromegalic pregnancies, we evaluated potential negative interference of pregnancy serum in the assay. METHODS: GH was measured in samples from acromegalicpatients with and without the addition of normal pregnancy serum using a sensitive in-house two-step GH-IFMA (no crossreactivity with pGH, Prolactin or hPL). Standard GH assay curves were run with and without pGH (20 and 22 K). Pegvisomant, a GH-antagonist with high homology to GH, was also tested for cross-reactivity. RESULTS: Addition of pregnancy serum to acromegaly serum resulted in marked decrease in GH, but addition of pGH did not change GH measurements. Redesign of the routine assay by switching the positions of the antibodies ("inverted" assay) completely abrogated the interference of pregnancy serum. GH by both routine and inverted assays declined progressively throughout pregnancy in controls, with higher nadir levels in the "inverted" assay (median 0.03 μg/L vs 0.50 μg/L, P<0.05). GH suppression during acromegalic pregnancy previously found with the routine assay was not observed in the "inverted" assay. Pegvisomant does not cross-react with GH in the "inverted" assay. CONCLUSIONS: GH measurements in pregnancy by immunometric assays must be made after exclusion of pregnancy serum interference by dilutional tests. Redesigning a two-step immunometric GH assay by switching the positions of the antibodies can be a successful strategy to abrogate such interference.