Measurement of human growth hormone in urine: development and validation of a sensitive and specific assay

Abstract

ABSTRACT A specific solid-phase immunoradiometric assay (IRMA), optimized for maximum sensitivity, has been developed for measurement of human GH (hGH) in urine. The sensitivity varied with sample size, giving a range of 0·001 to 0·003 mU/l for a sample volume of 2 ml. Recovery and dilution experiments, together with chromatography of urine samples, indicate that the method is specific for hGH. Added exogenous hGH was measured with a mean recovery of 101 ± 10% (s.d.) for 1 ml samples and 87 ± 8% for 2 ml samples. Measurements of samples diluted at 1:2 and 1:4 gave values of 97·4 and 96·6% respectively of those expected. Cross-reactions of human placental lactogen and prolactin were less than 0·008 and 0·04% respectively on a mol/mol basis. The assay was insensitive to the presence of NaCl (50–500 mmol/l), urea (50–1000 mmol/l), creatinine (1–20 mmol/l), Ca2+ ions (1–20 mmol/l), SO42− ions (1–1000 mmol/l), Mg2+ ions (0·05–50 mmol/l), 0·5–5% (w/v) glucose and a pH range of 6–9. Chromatography of unextracted samples showed that the immunoreactive material in urine eluted in a single homogenous peak with a similar position to monomeric pituitary hGH (22 kDa). Administered hGH (0·002%) was recovered in urine collected over a 2-h period following an intravenous injection. The urine output of hGH showed a good correlation with serum hGH in 18 patients following routine insulin tolerance tests and in 25 patients following an oral glucose tolerance test. Urine from normal subjects showed a mean night-time excretion of 498 ± 150 nU/h (s.d.) and a mean day-time excretion of 132 ± 93 nU/h. Preliminary results suggest that the measurement of hGH using this assay reflects the variation in blood and could be a useful diagnostic procedure. Journal of Endocrinology (1989) 121, 167–175