Prospective assessment of incomplete ovarian suppression in breast cancer using LC-MS/MS and immunoassays for estradiol

BACKGROUND: During gonadotrophin-releasing hormone agonist (GnRHa) therapy some patients do not adequately suppress estradiol, termed "ovarian escape" (OE). We prospectively measured estradiol using liquid chromatography-tandem mass spectrometry (LC-MS/MS), the gold standard to determine the prevalence of OE and assess discordance with estradiol immunoassays. METHOD: Patients receiving endocrine therapy plus GnRHa for ≥3 months were enrolled at two Dublin hospitals. Estradiol was measured by LC-MS/MS, Abbott Alinity i (Alinity) and Roche Cobas e 801 (Cobas) immunoassays. OE was defined as estradiol >10 pmol/L during aromatase inhibitor (AI) treatment or >77 pmol/L during tamoxifen. RESULTS: 94 patients enrolled, 10 had OE (11%; 95% CI 5.2-18.7%) and were younger than those without OE (40 years, IQR 38-43 versus 46 years, IQR 42-49; P < 0.005), all OE cases were receiving an AI (14%, 10/71). Exploratory multivariable logistic regression suggested increasing age was associated with lower odds of OE (OR 0.85 95% CI 0.73-0.99), BMI was not (OR 1.00, 95% CI 0.82-1.23). None of the 10 cases of OE were detected by immunoassay (sensitivity 0% both immunoassays). For Alinity and Cobas specificities were 86.5% and 100%. Seven false-positive (FP) results occurred with Alinity, including 6 patients receiving CDK4/6 inhibitors. Retesting of the 7 Alinity FP samples using Cobas demonstrated persistent FP results in 3 cases, all were receiving fulvestrant. CONCLUSION: OE occurred in 11% of women receiving GnRHa and was more common in younger women. Standard immunoassays demonstrated clinically relevant discordance compared to LC-MS/MS. These findings support LC-MS/MS for estradiol monitoring in this setting.