Poor diagnostic value of the euglycemic-hyperinsulinemic clamp in the clinical assessment of insulin resistant women with PCOS.

Insulin resistance (IR) per se and/or the ensuing compensatory hyperinsulinism, have been always considered as a key point in the pathogenesis of PCOS. Aim of our study was to evaluate which of these two aspects plays a key role in the pathogenesis of PCOS. Using a cross-sectional study design, 15 PCOS (Rotterdam criteria) and 16 age/BMI-matched control women (with suspect of IR) underwent clinical measures of IR after a 3-month withdrawal of insulin sensitizers and oral contraceptive pills. In an academic clinic setting, average glucose infusion rate (M-value) (during an euglycaemic–hyperinsulinaemic clamp), AUCinsulin and AUCglucose (during an Oral Glucose tolerance Test), HOMA-IR, ISI-Matsuda, Oral Dispositional Index (DIo) and Visceral Adiposity Index (VAI) were investigated. The prevalence of IR (according the M-value cut-off of 4.7 mg/Kg/min proposed by Bergman et al.) in the two groups was comparable: 14/16 (87.5%) for control women and 15/15 (100%) for PCOS women (p=0.484). No significant differences were observed between the two groups for M-value (p=0.540), VAI (p=0.406) and DIo (p= 0.813). Women with PCOS showed significantly higher levels of fasting insulin [median (IQR): 22 (19-37) vs. 13.55 (10.25-18.67) mU/ml; p<0.001], 30’ after OGTT insulin [96 (66-230) vs. 48 (26.25-83.25) mU/ml; p=0.003] and consequently significant differences in derived indexes (HOMA-IR, ISI-Matsuda, AUCinsulin). Furthermore, by performing a Kruskal-Wallis test between the four Rotterdam-PCOS phenotypes, the complete phenotype showed significantly higher levels of VAI (p=0.007). Our study suggests that for the assessment of IR in PCOS, the gold standard euglycemic-hyperinsulinemic clamp, an expensive method requiring an experienced operator to manage the technical difficulties, has also the limitation that utilizes steady-state insulin levels that may be supraphysiological. This finding may be explained in the reversal of normal portal to peripheral insulin gradient. Thus, the glucose clamp may not accurately reflect the ovarian insulin action under physiological conditions, nor gives us information about the compensatory hyperinsulinism. In non-diabetic women with PCOS, a dynamic test, such as OGTT (in particular the initial phase of stimulated insulin secretion), provides more useful information than the gold standard glucose Clamp.