Urine phosphoethanolamine is a specific biomarker for hypophosphatasia in adults

Objectives: We investigated the utility of urine phosphoethanolamine (PEA) as a marker to aid in diagnosing and/or confirming hypophosphatasia (HPP) in adults and for monitoring patients on enzyme replacement therapy (ERT).

Methods: Data was collected from seventy-eight adults who were referred to the Vanderbilt Program for Metabolic Bone Disease for evaluation of a possible or confirmatory HPP diagnosis between July 2014 through December 2019. Fifty-nine patients were diagnosed with HPP and nineteen were excluded from a diagnosis of HPP. The urine PEA results of those patients with a confirmed diagnosis of HPP and those patients with a diagnosis of HPP excluded were captured and compared to other laboratory and clinical parameters consistent with HPP, including alkaline phosphatase (ALP) activity, plasma pyridoxal 5'-phosphate (PLP), the presence of musculoskeletal abnormalities, and genetic testing for pathogenic mutations in ALPL.

Results: Initial urine PEA values in patients in our HPP cohort and not on ERT were significantly higher (median = 150.0 nmol/mg creatinine, IQR = 82.0-202.0) compared patients in our HPP negative group (median 18.0 nmol/mg creatinine, IQR = 14.0-30.0, p < 0.0001) and higher than patients on ERT (median 65.0 nmol/mg creatinine, IQR = 45.3-79.8). Patients who began ERT had a decline in urine PEA levels after treatment with a mean decrease of 68.1 %. Plasma ALP levels were significantly lower in the group of patients with HPP and not on ERT group (median = 24.0 U/L, IQR = 15.0-29.50) compared to the patients without HPP (median = 45.50 U/L, IQR = 34.0-62.0;) and plasma PLP levels were significantly higher in the HPP non-ERT group (median = 284.0 nmol/L, IQR = 141.0-469.4) compared to the patients without HPP (median = 97.5 nmol/L, IQR = 43.7-206.0;). The area under the curve (AUC) of urine PEA, ALP, and PLP to distinguish between HPP and non-HPP patients is 0.968, 0.927 and 0.781, respectively, in our cohort. Urine PEA had 100 % specificity (95 % CI of 83.2 % to 100.0 %) for diagnosing HPP at a value >53.50 nmol/mg creatinine with a sensitivity of 88.4 %; 95%CI 75.5 to 94.9 %. ALP had a 100 % specificity (95 % CI of 82.4 % to 100.0 %) for diagnosing HPP at a value <30.5 U/L with a sensitivity of 77.2 %; (95%CI 64.8 to 86.2 %). PLP had a 100 % specificity (95 % CI of 81.6 % to 100.0 %) for diagnosing HPP at a value >436 nmol/L with a sensitivity of 26.9 %; (95%CI 16.8 to 40.3 %). The most common pathogenic or likely pathogenic mutations in our cohort were c.1250A>G (p.Asn417Ser), c.1133A>T (p.Asp378Val), c.881A>C (p.Asp294Ala), c.1171C>T (p.Arg391Cys), and c.571G>A, (p.Glu191Lys).

Conclusions: Urine PEA is a promising diagnostic and confirmatory marker for HPP in patients undergoing investigation for HPP. Urine PEA also has potential use as a marker to monitor ERT compliance. Future studies are necessary to evaluate the association between PEA levels and clinical outcomes.