Adjustment of serum HE4 to reduced glomerular filtration and its use in biomarker-based prediction of deep myometrial invasion in endometrial cancer.

Background: We investigated the efficacy of circulating biomarkers together with histological grade and age to predict deep myometrial invasion (dMI) in endometrial cancer patients.

Methods: HE4ren was developed adjusting HE4 serum levels towards decreased glomerular filtration rate as quantified by the eGFR-EPI formula. Preoperative HE4, HE4ren, CA125, age, and grade were evaluated in the context of perioperative depth of myometrial invasion in endometrial cancer (EC) patients. Continuous and categorized models were developed by binary logistic regression for any-grade and for G1-or-G2 patients based on single-institution data from 120 EC patients and validated against multicentric data from 379 EC patients.

Results: In non-cancer individuals, serum HE4 levels increase log-linearly with reduced glomerular filtration of eGFR ≤ 90 ml/min/1.73 m. HE4ren, adjusting HE4 serum levels to decreased eGFR, was calculated as follows: HE4ren = exp[ln(HE4) + 2.182 × (eGFR-90) × 10]. Serum HE4 but not HE4ren is correlated with age. Model with continuous HE4ren, age, and grade predicted dMI in G1-or-G2 EC patients with AUC = 0.833 and AUC = 0.715, respectively, in two validation sets. In a simplified categorical model for G1-or-G2 patients, risk factors were determined as grade 2, HE4ren ≥ 45 pmol/l, CA125 ≥ 35 U/ml, and age ≥ 60. Cumulation of weighted risk factors enabled classification of EC patients to low-risk or high-risk for dMI.

Conclusions: We have introduced the HE4ren formula, adjusting serum HE4 levels to reduced eGFR that enables quantification of time-dependent changes in HE4 production and elimination irrespective of age and renal function in women. Utilizing HE4ren improves performance of biomarker-based models for prediction of dMI in endometrial cancer patients.