Objective: We evaluated the significance of single serum LH estimates (as assessed by radiometric assay (IRMA) and Leydig cell in-vitro bioassay (BIO)) for the diagnosis of polycystic ovary syndrome (PCOS) in women with infertility and cycle abnormalities.
Design: Hormonal and clinical comparisons between subgroups were made based on classification according to (a) rigid clinical and endocrine (excluding LH) characteristics of PCOS, (b) elevated IRMA-LH concentrations, (c) BIO-LH levels. In addition, androgen modulation of LH biopotency was studied in these patients.
Patients: Ninety-nine women presenting at our infertility Unit with oligo/amenorrhoea.
Measurements and results: Of the total study group, 35 women were diagnosed positive as PCOS and 42 showed elevated IRMA-LH levels. Only 51% (n = 18) of PCOS patients showed elevated IRMA-LH levels, and in PCOS significantly higher levels of BIO-LH, androstenedione, oestrone, and BIO/IRMA-LH ratios were found as compared to non-PCOS patients. In the group with elevated IRMA-LH only 43% (n = 18) of subjects were diagnosed as PCOS, and no difference in BIO/IRMA-LH ratios was found. With increasing BIO-LH levels the probability of PCOS rises sharply (P < 0.001), whereas this probability is of only marginal significance (P < 0.06) for IRMA-LH. In the total study group a correlation is observed between serum testosterone (T) levels and IRMA-LH (r = 0.47), and BIO-LH (r = 0.51) concentrations. This correlation is absent comparing serum T and BIO/IRMA-LH ratios (r = 0.15).
Conclusions: Results presented in this study indicate that (1) women with infertility and oligo/amenorrhoea classified based on signs of PCOS or IRMA-LH levels, exhibit different clinical and endocrine characteristics, (2) only 51% of PCOS women exhibit elevated IRMA-LH concentrations, and only 43% of women with elevated IRMA-LH were diagnosed as PCOS, (3) IRMA-LH levels are a poor predictor of PCOS, whereas the predictive value of BIO-LH is better, (4) elevated BIO/IRMA-LH ratios in PCOS are dependent on alterations in BIO-LH, rather than IRMA-LH concentrations, and (5) no correlation was observed between serum T levels and BIO/IRMA-LH ratios.