LH levels in women with polycystic ovarian syndrome: have modern assays made them irrelevant?

Abstract

Objective To determine whether using newer monoclonal rather than polyclonal assays for measuring luteinising hormone (LH) alters the predictive value of LH and LH/follicle-stimulating hormone (FSH) ratios for polycystic ovarian syndrome. Design Prospective cohort study. Setting Fertility and Reproductive Endocrinology Clinic within a New Zealand Teaching hospital. Population Seventy-eight women presenting with oligomenorrhoea or hirsutism and polycystic ovaries on pelvic ultrasound and 59 volunteer controls with ultrasonically normal ovaries and a regular menstrual cycle. Methods Serum LH concentrations were measured using a polyclonal radio-immunoassay (Amerlex-M, Johnson & Johnson) and two monoclonal immunometric assays (Immulite, DPC; Cobas Core, Hoffman La Roche). The proportion of women with an elevated serum LH concentration in each group was calculated using both current local laboratory reference intervals and a new reference range derived from our control group. The LH/FSH ratios for women in both groups were also calculated using the three different LH assays. Main outcome measures LH concentrations and LH:FSH ratios measured using polyclonal and monoclonal immunoassays. Results Using the local laboratory normal range, a significantly higher proportion of women had an elevated LH when measured with a polyclonal assay (23.1%) than when measured with a monoclonal assay (12.8% Core, 6.4% Immulite) ( P < 0.05). LH/FSH ratios were significantly lower when monoclonal assay was used and receiver–operator characteristic curves suggest that LH/FSH ratios of 1 or lower provide the most reliable separation of women with polycystic ovarian syndrome from controls when these assays are used. Conclusions Clinicians should be aware that the use of monoclonal LH assays will result in significantly lower measured LH levels and LH/FSH ratios in women with polycystic ovarian syndrome than previously used polyclonal assays. Account should be taken of the assay type used, when using endocrinological parameters in the diagnosis of polycystic ovarian syndrome, or the identification of women who have LH hypersecretion.