What is your approach to systemic anticoagulation for patients with hypoalbuminemia and nephrotic syndrome secondary to a non-membranous nephropathy condition?

Patients with nephrotic syndrome (NS) and hypoalbuminemia have a several-fold higher risk of venous thromboembolism (VTE) than the general population and also a somewhat higher risk of arterial thromboembolism (ATE), such as MI and stroke. This risk seems to be higher in membranous nephropathy (MN) than with other etiologies of NS, even after adjustment for the degree of hypoalbuminemia and severity of proteinuria (Barbour et al., PMID 21918501). Patients with minimal change disease, membranoproliferative glomerulonephritis, and focal and segmental glomerulosclerosis, if severely nephrotic, also have a high risk. Thrombosis occurs most commonly during the first year after diagnosis.

Prophylactic anticoagulation in NS is controversial. The likelihood of benefit in a given patient depends on the risk of thromboembolic events, the efficacy of the anticoagulant to prevent such events, and the risk of anticoagulation. There are no randomized controlled trials to guide management. In MN, the more severe the hypoalbuminemia, the higher the risk of VTE, which may also be true with other etiologies of NS. In a systematic review of 51 articles, Lin et al. (Lin et al., PMID 32274450) concluded that prophylactic anticoagulation should be considered in non-MN patients as well as MN patients with severe NS. They suggest anticoagulation with either warfarin or enoxaparin in MN patients with serum albumin < 3 g/dL and low bleeding risk (HAS-BLED 0-1) or serum albumin < 2.5 g/dL (< 3 g/dL with 1 or more other thrombotic risk factors) and moderate bleeding risk (HAS-BLED 2). In non-MN patients, anticoagulation is suggested with serum albumin < 2.5 g/dL and HAS-BLED of 0-1 or serum albumin < 2.0 g/dL (< 2.5 g/dL with 1 or more other thrombotic risk factors) and HAS-BLED of 2. They recommend against prophylactic anticoagulation with high bleeding risk (HAS-BLED >=3). The evidence for direct-acting oral anticoagulants (DOACs) is limited, which, when coupled with the inability of monitor efficacy, makes them a less favorable alternative, although in a recently published retrospective analysis, DOACs appeared to have similar efficacy and safety to warfarin (Cubilier et al., PMID 40485718). If used, prophylactic anticoagulation is generally continued as long as the patient remains nephrotic (this is also true of therapeutic anticoagulation). There is likely no role for antiplatelet drugs for VTE prophylaxis.

Although I generally have not used prophylactic anticoagulation in NS unless there is a history of previous thrombosis or multiple risk factors for thrombosis, I think the approach outlined by Lin et al. is reasonable.

All nephrotic proteinuria is associated with increased risk of thrombosis, but membranous nephropathy in particular has the highest associated risk (see Barbour et al., PMID 21918501). In adults with non-membranous nephrotic syndrome, I reserve anticoagulation for patients with a history of venous thromboembolism. Per KDIGO, "Prophylactic anticoagulation should be employed in patients with nephrotic syndrome when the risk of thromboembolism exceeds the estimated patient-specific risks of an anticoagulation-induced serious bleeding event." How to calculate this remains a matter of conjecture and extrapolation (though KDIGO does have some suggestions to consider). While I do not routinely give prophylactic anticoagulants in adults with non-membranous NS, in sufficiently high-risk individuals, it is not unreasonable (e.g., with very low serum albumin, immobility, or other risks). In most patients, I will at least use an aspirin if not contraindicated. There is no good trial-based evidence to guide prescribing practices, however.

Generally, I prefer AC for patients with serially declining albumin (usually <2.5 g/dL) and proteinuria >10 g/g, as well as other risk factors for thrombosis (NYHA 3/4, immobilization, factor 5 Leiden, AF, history of VTE, etc.). The question is how quickly I can get the patient in remission, and how long I think the patient will remain nephrotic. The rate for VTE is highest for MN, but it is also high for other conditions. The data is observational primarily and is with VKA, but lately, people have started using non-VKA drugs like apixaban. The volume of distribution and pharmacokinetics may be altered with drugs like apixaban, but not having to measure therapeutic levels can be appealing.

Generally I prefer AC for patients with serially dropping albumin (usually below 2.5 g/dl) and proteinuria>10 g/g and other reasons fo