Endocrine Abstracts

Available glucocorticoid (GC) replacement regimens in adrenal insufficiency (AI) only roughly correspond to physiological steroid profiles. Control of substitution quality is therefore difficult but significant, as even mild chronic over- or under-replacement may be clinically relevant. FKBP5 regulates GC receptor sensitivity by reducing its affinity to cortisol when bound to the receptor complex. FKBP5 methylation has been inversely correlated with cortisol levels both in healthy controls and in patients with endogenous hypercortisolism. We analyzed FKBP5 gene methylation (DNAm) within introns containing GC responsive elements as well as promoter and proximal enhancer regions by bisulfite pyrosequencing in a cohort of 86 patients with primary (PAI, n=57) and secondary (SAI, n=29) AI. Results were correlated with GC dose, salivary and 24-hour urinary cortisol, prevalence of adrenal crises (AC) per patient-year and 24-hour blood pressure (BP) levels. GC dose and DNAm were negatively correlated for the majority of the investigated regions (intron 1 r_s=-0.45, P<0.01, intron 5 r_s=-0.35 P<0.01, intron 7 r_s=-0.23 P=0.034, promoter A1 r_s=-0.35 P<0.01, proximal enhancer A2 r_s=-0.38 P<0.01). Intronic DNAm correlated negatively with 24-hour urinary cortisol (intron 2, r_s=-0.25, P=0.032) and positively with bedtime salivary cortisol (intron 7, r_s=0.3, P<0.01). We observed a positive correlation between the prevalence of AC and intronic DNAm (intron 2 and 5, r_s=-0.29 P<0.01 for each). Systolic 24-hour and day-time BP, systolic and diastolic night-time BP and noctural dipping correlated negatively with DNAm within several intronic, promoter and proximal enhancer regions. GC replacement was higher, whereas intronic DNAm was lower in PAI compared to SAI (GC: 22 (10-60) vs 20 (10-37,5) mg P=0.032, intron 5: 11% vs 15% P=0.028). FKBP5 methylation analysis may provide helpful further insight regarding the evaluation of GC replacement and might help improving assessment of GC load in AI, as it correlates with replacement doses, cortisol levels and 24-hour BP. Our observations warrant further analyses in larger cohorts.