Abstract

Warfarin inhibits vitamin-K dependent proteins involved in bone mineralization and the prevention of vascular calcification (bone Gla protein BGP, matrix Gla protein MGP). In this multicenter, cross-sectional study with 3-year follow-up, data from 387 patients on hemodialysis for ≥1 year at 18 dialysis units were analyzed. Patients on warfarin treatment for > 1 year (11.9% of the population) were compared with the remaining cohort for vertebral fractures, vascular calcifications and mortality. Vertebral fractures and vascular calcifications were sought in L-L vertebral X-rays (D5 to L4).

Compared with controls, warfarin-treated male patients had more vertebral fractures (77.8 vs. 57.7%, p<0.04), but not females (42.1% vs. 48.4%, p=0.6); total BGP was significantly reduced (82.35 vs. 202 μg/L, p<0.0001), with lower levels in treated men (69.5 vs. women 117.0 μg/L, p=0.03). In multivariate logistic regression analyses, the use of warfarin was associated with increased odds of aortic (OR 2.58, p<0.001) and iliac calcifications (OR 2.86, p<0.001); identified confounders were age, atrial fibrillation, angina, PPI use and total BGP. Seventy-seven patients died during a 2.7±0.5 year follow-up. In univariate Cox regression analysis, patients on warfarin had a higher risk of all-cause mortality (HR 2.42, 95% CI 1.42-4.16, p=0.001) when compared with those untreated and data adjustment for confounders attenuated but confirmed the significant warfarin-mortality link (HR: 1.97, 95% CI: 1.02-3.84, P=0.046).

In hemodialysis patients, additional studies are warranted to verify the risk/benefit ratio of warfarin, which appears to be associated with significant morbidity and increased mortality.

Keywords: Hemodialysis, matrix Gla protein, mortality, osteocalcin, vascular calcifications, vertebral fractures, warfarin.