Anticoagulant-related nephropathy

INTRODUCTION — Anticoagulant-related nephropathy (ARN; also referred to as warfarin-related nephropathy) is a type of acute kidney injury (AKI) that may be caused by excessive anticoagulation with warfarin and other anticoagulants [1,2]. ARN has been associated with irreversible kidney injury in some patients and with increased risk of mortality.

While initially termed "warfarin-related nephropathy," we prefer the term "anticoagulant-related nephropathy" because the entity has been associated with anticoagulants other than warfarin [1].

This topic reviews the epidemiology, pathogenesis, clinical features, and treatment of ARN. The differential diagnosis of AKI, glomerulonephritis, and hematuria are discussed elsewhere. (See "Diagnostic approach to adult patients with subacute kidney injury in an outpatient setting" and "Glomerular disease: Evaluation and differential diagnosis in adults" and "Etiology and evaluation of hematuria in adults".)

BACKGROUND AND EPIDEMIOLOGY — Acute kidney injury (AKI) resulting from glomerular hemorrhage has been described in patients with glomerular lesions (ie, kidney disease) in the absence [3-5] and presence [6,7] of coagulopathy (International Normalized Ratio [INR] 6 to 9 range). More recently, AKI has been described among patients without underlying kidney disease and with more modest elevations of INR [8]. A detailed biopsy study of nine patients who developed otherwise unexplained AKI in association with warfarin overdose revealed the predominant lesion of tubular injury and obstruction with red blood cells (RBCs) and RBC casts [9]. The recognition of a characteristic histologic lesion that was associated with the clinical presentation of otherwise unexplained AKI in the setting of over-anticoagulation led to the term "anticoagulant-related nephropathy."

The incidence of ARN is difficult to determine. This is because studies that have examined incidence have relied upon a presumptive diagnosis of ARN (defined by an elevation in the serum creatinine within several days of an abnormally elevated INR) rather than a more definitive diagnosis (defined by biopsy) [8,10,11]. (See 'Diagnosis' below.)

Studies have suggested that an increase in serum creatinine above baseline is observed in approximately 17 percent of warfarin-treated patients who do not have underlying chronic kidney disease (CKD) and who have an elevated INR >3 [8,11].

However, it is not clear that the increase in creatinine was related to ARN in these studies [8,11]. Studies were based on review of charts or administrative data (such as diagnostic codes). Thus, other possible causes of AKI were not rigorously excluded, and biopsies were not performed. In addition, by contrast to the landmark biopsy study cited above, in which all nine cases had hematuria [9], hematuria was absent in many patients in the large, retrospective series [8]. (See 'Pathogenesis' below.)

Thus, in the absence of biopsy data, epidemiologic studies probably overestimated the incidence of ARN. Incidence was also overestimated by ascertainment bias since the detection of AKI was dependent upon having the serum creatinine measured within one week of the elevated INR; sicker patients are more likely to have creatinine measured and are also more likely to develop ARN. (See 'Risk factors' below.)

However, in practice, ARN may be underdiagnosed because of two major reasons:

●Generally, nephrologists are reluctant to perform kidney biopsy in patients who need therapeutic anticoagulation. This is because of concerns related to the risk of thrombosis/thromboembolism while systemic anticoagulation is held and to the risk of hemorrhage from the kidney biopsy site after the systemic anticoagulants are restarted.

Concerns regarding the risk of hemorrhage may be heightened in the patient whose AKI occurred in a setting of recent over-anticoagulation, which is the situation typical of ARN. In such cases, the biopsy is often delayed until the INR returns to a therapeutic range. Since, in ARN, the creatinine often improves or stabilizes when the INR returns to a therapeutic range, the clinician may conclude that a biopsy is no longer needed.

●The possibility of ARN may be overlooked because it is most common among patients who have multiple risk factors for AKI from any cause, and AKI may be attributed to other causes. For example, patients at highest risk for ARN are those with underlying CKD and heart disease. CKD and heart disease both predispose to AKI from any cause and are often treated with therapies that can cause AKI (such as angiotensin-converting enzyme [ACE] inhibitors, angiotensin receptor blockers [ARBs], diuretics, and low-salt diet). (See 'Risk factors' below and "Etiology and diagnosis of prerenal disease and acute tubular necrosis in acute kidney injury in adults", section on 'Etiology'.)

In such patients, episodes of AKI are often attributed to multifactorial causes, and extensive diagnostic evaluation is not performed. (See "Etiology and diagnosis of prerenal disease and acute tubular necrosis in acute kidney injury in adults", section on 'Etiology'.)

Also, the prevalence of ARN may be low even though the incidence is high. This may add to the perception that it is an uncommon problem. Most of the described episodes of potential ARN occurred in the first eight weeks after initiating anticoagulation [8,10]. In these studies, approximately 30 percent of patients with CKD and ARN died within approximately one month after the onset of ARN, thus reducing the prevalence [8].

RISK FACTORS — The major risk factor for ARN is moderate or severe coagulopathy induced by warfarin or other anticoagulants. The severity of the coagulopathy required to cause ARN is not known. Most commonly, the International Normalized Ratio (INR) is in the 4s [8,10,11].

However, factors other than the degree of coagulopathy contribute to the development of acute kidney injury (AKI). In the series cited above, there was no difference in the peak INR between those with AKI and those without AKI (4.91 versus 4.21) [10]. (See 'Pathogenesis' below.)

The risk of ARN may be higher among patients with underlying chronic kidney disease (CKD) [8,10-16]. In one study cited above, presumed ARN occurred in 34 percent of patients who had CKD at baseline [8]. In another retrospective analysis of 49 warfarin-treated CKD patients who had at least one episode of INR >3, 37 percent developed an increase in serum creatinine ≥0.3 mg/dL within one week of their first INR >3 [10]. This reported incidence is higher than the reported incidence of 17 percent among warfarin-treated patients who do not have underlying CKD [8,11]. (See 'Background and epidemiology' above.)

Other proposed risk factors include older age (ie, >80 years), male sex, diabetes mellitus, heart failure, hypertension, and glomerulonephritis, particularly with nephrotic syndrome [8,10,11,16].

PATHOGENESIS — Our understanding of pathogenesis is based upon histologic analysis of kidney tissue obtained by biopsy and on experiments in animal models [9,17,18]. The pathogenesis of ARN is likely multifactorial and includes:

●Glomerular hemorrhage

●Obstruction of renal tubules by red blood cell (RBC) casts

●Tubular epithelial cell injury

The initiating event appears to be glomerular hemorrhage, caused by excessive anticoagulation due to warfarin or other anticoagulants [1,19]. Structurally abnormal glomerular barriers, such as are present in thin basement membrane disease or thick basement membrane disease, are especially vulnerable to hemorrhage [6,7].

Glomerular hemorrhage results in the formation of obstructing RBC casts within renal tubules. Obstructing intraluminal RBC casts are the most conspicuous histologic feature of ARN in tissue obtained from patients who have undergone biopsy [9] and in animal models [17,18].

However, the number of obstructed tubules observed on microscopy does not appear sufficient to explain the decrease in the glomerular filtration rate (GFR). In both patients and animals with ARN, only a small percentage of tubules in any given histologic section show complete obstruction with RBC casts [6,7,20].

One explanation for this is that analysis by microscopy underestimates the total number of obstructed tubules since only a very small cross section of any given tubule can be viewed on a slide. Tubules that do not appear obstructed in one histologic section could be completely obstructed at a section of the tubule that is either proximal or distal to the section being examined.

Another explanation is that, in addition to obstruction, glomerular hemorrhage resulting from anticoagulation causes pervasive tubular injury. This hypothesis is supported by evidence that glomerular hemorrhage and/or hemoglobinuria are associated with tubular epithelial cell injury in patients with immunoglobulin A (IgA) nephropathy [3-5] or paroxysmal nocturnal hemoglobinuria [21,22]. Tubular injury may be mediated, at least in part, by oxidative activity of heme and iron [23-26]. Alternatively, anticoagulants may be directly toxic to renal tubules [20].

HISTOLOGIC FEATURES — The major histologic feature of ARN is obstruction of renal tubules (mainly distal) by red blood cell (RBC) casts (picture 1) [9]. The glomeruli show little or no abnormalities by light, immunofluorescence, or electron microscopy [9].

CLINICAL PRESENTATION — Observational studies suggest that most episodes of acute kidney injury (AKI) occur within eight weeks of starting warfarin [8,10,11]. We do not know with certainty the interval between the increased International Normalized Ratio (INR) and the onset of AKI. However, based on clinical observations [8] and on our animal model, only a few days of coagulopathy are needed to inflict ARN [8,17].

Other presenting features depend on the severity of AKI. Patients with severe AKI (ie, marked reduction in the glomerular filtration rate [GFR]) may present with hypertension, signs of volume overload, and reduced urine output [6,7]. (See "Diagnostic approach to adult patients with subacute kidney injury in an outpatient setting", section on 'Clinical manifestations'.)

Some patients may present with hematuria and are subsequently determined to have a serum creatinine that is elevated above baseline [6,7]. In initial case reports and in a detailed analysis of nine patients with biopsy-proven ARN, all patients had hematuria [6,9,27]. However, hematuria has not been universally described in all patients with presumed ARN [8,10].

We do not believe the absence of hematuria excludes a possible diagnosis of ARN. It is possible that hematuria is transient and not detected in all patients at the time of presentation with AKI. This may be due to the fact that hemorrhaging glomeruli are shut down by red blood cell (RBC) casts or acute tubular injury by the time patients came to clinical attention. This has been described in other settings of glomerular bleeding. (See "Loin pain-hematuria syndrome", section on 'Hematuria'.)

The absence of hematuria has also been observed in animal models of ARN [17,18].

Gross hematuria is less common than microscopic hematuria [8-10]. In the study cited above, in nine patients with ARN, only one had gross hematuria [9].

Analysis of the urinary sediment reveals dysmorphic RBCs and, occasionally, RBC casts. Leukocyturia is not commonly seen.

Proteinuria of varying severity has been reported, but the studies that have examined proteinuria included only patients with underlying chronic kidney disease (CKD) [6,9,27]. Thus, it is not clear whether ARN in the absence of underlying CKD causes significant abnormal proteinuria.

Among patients with ARN who are treated with warfarin, the INR is elevated. In various studies, the average INR among patients with ARN was in the mid to upper 4s [8-11].

However, as noted above, patients who are on nonwarfarin agents, such as dabigatran, may also be at risk for ARN [1,28]. There are no characteristic features on imaging, and the kidney ultrasound is typically normal.

DIAGNOSIS — The diagnosis of ARN should be suspected among patients who present with acute kidney injury (AKI) in the setting of excessive anticoagulation. A definitive diagnosis is made by kidney biopsy (see 'Histologic features' above). However, biopsies are usually NOT performed, at least initially, among patients who are anticoagulated, because the risk of bleeding is high.

Based upon our experience, among patients who develop AKI and are on chronic warfarin therapy, a presumptive diagnosis of ARN may be made if a severe warfarin coagulopathy is present and if other causes of AKI have been excluded by clinical features and serologic tests. (See "Diagnostic approach to adult patients with subacute kidney injury in an outpatient setting", section on 'Overview'.)

As noted above, the severity of the coagulopathy required to cause ARN is not known. In reported cases of possible anticoagulant-induced nephropathy, the International Normalized Ratio (INR) was most commonly in the 4s [8-11]. However, in some patients, the coagulopathy may have already been corrected by the time AKI comes to attention.

In addition, among patients on nonwarfarin anticoagulants, including direct thrombin inhibitors (dabigatran) or factor Xa inhibitors, ARN may be present in the setting of a normal INR. Among such patients, a diagnosis of ARN is reasonable if AKI is associated with obvious evidence of a coagulopathy. In the case reports describing AKI associated with dabigatran coagulopathy, patients had obvious hemorrhage from multiple sites, as well as histologic evidence of hemorrhage into renal tubules [1,19,28]. There is evidence that factor Xa inhibitors (apixaban) can induce AKI as well [29].

It is important that other causes of AKI be excluded to the degree that this is possible without a kidney biopsy (see "Diagnostic approach to adult patients with subacute kidney injury in an outpatient setting", section on 'Overview'). The evaluation should include a careful history (to exclude hypotensive episodes or contrast exposure to suggest acute tubular necrosis), review of medications, and physical examination (eg, signs of volume contraction).

In all patients, a kidney ultrasound should be obtained to exclude obstruction. (See "Clinical manifestations and diagnosis of urinary tract obstruction (UTO) and hydronephrosis".)

Laboratory testing should be performed to exclude causes of glomerulonephritis. Specific tests are discussed elsewhere. (See "Diagnostic approach to adult patients with subacute kidney injury in an outpatient setting", section on 'Serologic testing and role of kidney biopsy'.)

Among patients who have no other cause of AKI identified by the evaluation described above, a diagnosis of presumptive ARN is justified and precludes, at least initially, the need for a biopsy.

However, a kidney biopsy may be warranted if any of the following conditions apply:

●The serum creatinine continues to increase or remains stably elevated above the patient's baseline, even though the coagulopathy has been restored to a therapeutic range. Among most patients with ARN, the serum creatinine will improve slightly within the first few weeks after restoration of a therapeutic INR. Continued increase of the creatinine restoration of a therapeutic INR suggests that another cause of AKI is present.

●Hematuria persists or emerges after restoration of a therapeutic INR. Hematuria that persists despite this suggests that the patient may have an underlying glomerulonephritis [10,30,31] or abnormally thin [7] or thick glomerular basement membrane [6].

Such patients should undergo a kidney biopsy for diagnosis. The coagulopathy must be reversed prior to biopsy in order to minimize the risk of bleeding.

DIFFERENTIAL DIAGNOSIS — The differential diagnosis of acute kidney injury (AKI) is broad and depends on findings on the urinalysis and quantitation of protein excretion. Among patients who present with hematuria, and in particular with dysmorphic red blood cells (RBCs) and RBC casts, an important differential diagnosis is glomerulonephritis or vasculitis. There are a number of reasonably sensitive and specific tests for glomerulonephritides. These tests are discussed elsewhere. (See "Glomerular disease: Evaluation and differential diagnosis in adults".)

As noted above, some patients present without conspicuous hematuria (see 'Clinical presentation' above). Among such patients, one should consider prerenal and urinary tract obstruction.

Prerenal disease caused by volume depletion or treatment with angiotensin-converting enzyme (ACE)/angiotensinogen receptor blocker (ARB) may be excluded by history (absence of vomiting, diarrhea, decreased oral intake) and medication review and occasionally by a trial of volume repletion. (See "Etiology and diagnosis of prerenal disease and acute tubular necrosis in acute kidney injury in adults", section on 'Evaluation and diagnosis'.)

Urinary tract obstruction may be excluded with a kidney ultrasound. (See "Clinical manifestations and diagnosis of urinary tract obstruction (UTO) and hydronephrosis".)

Other causes of AKI with a benign urine sediment include hypercalcemia, acute phosphate nephropathy, acute high-dose use of nonsteroidal antiinflammatory drugs (NSAIDS) or acetaminophen, crystal-induced nephropathy (eg, urate from acute overproduction or oxalate from acute high intake), and myeloma cast nephropathy. (See "Diagnostic approach to adult patients with subacute kidney injury in an outpatient setting", section on 'Overview'.)

TREATMENT — There are no prospective studies to specifically inform the management of the patient who develops unexplained acute kidney injury (AKI) while receiving oral anticoagulant therapy. In particular, the role of kidney biopsy in patients with presumptive ARN remains to be determined.

Kidney biopsy may provide critically important information. For example, if the biopsy confirms ARN in a warfarin-treated patient, it may be appropriate to switch the patient to dabigatran or rivaroxaban, the novel oral anticoagulants (NOACs), which may have a lower risk of ARN. On the other hand, if the kidney biopsy confirms ARN and the patient is receiving dabigatran or rivaroxaban, an appropriate strategy might be to simply reduce the dose of the NOAC. Finally, if the kidney biopsy in the patient receiving apixaban shows ARN, the appropriate strategy might be to switch the patient to rivaroxaban or dabigatran [32].

The utility of biopsy for a diagnosis of ARN requires further study.

The optimal method for correcting excess anticoagulation after the use of warfarin (eg, returning an increased International Normalized Ratio [INR] to the desired range) generally depends upon the degree of INR elevation, the indication for anticoagulation, and whether clinically significant bleeding is present. (See "Management of warfarin-associated bleeding or supratherapeutic INR" and "Anticoagulation for prosthetic heart valves: Management of bleeding and invasive procedures".)

We believe that this approach (ie, selecting the method of restoration of a therapeutic INR based upon degree of INR elevation and absence or presence of clinically significant bleeding) is acceptable for patients with AKI related to ARN. Methods for correcting excess anticoagulation are described elsewhere. (See "Management of warfarin-associated bleeding or supratherapeutic INR".)

The reversal of coagulopathy induced by NOACs is discussed elsewhere. (See "Management of bleeding in patients receiving direct oral anticoagulants".)

In addition to reversing the coagulopathy, supportive care should be provided as for all patients with AKI. (See "Overview of the management of acute kidney injury (AKI) in adults".)

The use of glucocorticoids in the treatment of ARN has been suggested but is unproven [33].

PREVENTION — The most important measure to prevent ARN is proper adjustment of the anticoagulant dose. This is particularly important for patients with chronic kidney disease (CKD), who are more vulnerable to ARN. (See 'Background and epidemiology' above and 'Risk factors' above.)

OUTCOME — In most patients with ARN, the serum creatinine stabilizes or improves slightly within the first few weeks after correction of the warfarin coagulopathy. However, many patients may have little or no recovery of kidney function [9,34]. In the study cited above of nine patients with biopsy-proven ARN, five patients did not regain previous kidney function [9].

Presumptive ARN has been associated with an increase in mortality independent of age, sex, race, hemorrhage, atrial fibrillation, heart failure, and diabetes mellitus [8,11]. In the study of 4006 patients with an elevated International Normalized Ratio (INR) >3, patients who developed presumptive ARN had decreased five-year survival compared with those without (58 versus 73 percent, respectively). The sharpest reduction in survival occurred during the first few weeks after the INR increase in this study (figure 1).

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Acute kidney injury in adults".)

SUMMARY AND RECOMMENDATIONS

●Anticoagulant-related nephropathy (ARN) is a type of acute kidney injury (AKI) that is caused by excessive anticoagulation with warfarin or other classes of anticoagulants. (See 'Introduction' above.)

●The major risk factor for ARN is moderate or severe anticoagulation induced by warfarin or other anticoagulants. The risk appears to increase at International Normalized Ratio (INR) >4. Other risk factors include underlying chronic kidney disease (CKD), diabetes mellitus, heart failure, hypertension, and glomerulonephritis, particularly with nephrotic syndrome. (See 'Risk factors' above.)

●Our understanding of pathogenesis is based upon histologic analysis of kidney tissue obtained by biopsy and on studies in animal models. The initiating event appears to be glomerular hemorrhage caused by excessive anticoagulation. Glomerular hemorrhage results in the formation of obstructing red blood cell (RBC) casts within renal tubule and in tubular injury, the oxidation activity of heme and iron. (See 'Pathogenesis' above.)

●Patients present with AKI within several days (usually within one week) of an abnormally elevated INR. Microscopic and, less commonly, gross hematuria may be present. Hematuria may be transient; however, and we do not believe that the absence of hematuria at the time of presentation should exclude the possibility of ARN. There are no characteristic features on imaging, and the kidney ultrasound is typically normal. (See 'Clinical presentation' above.)

●The diagnosis should be suspected among patients who present with AKI in the setting of excessive anticoagulation (INR >4). A definitive diagnosis is made by kidney biopsy. However, biopsies are usually NOT performed, at least initially, among patients who are anticoagulated, because the risk of bleeding is high. A presumptive diagnosis of ARN may be made if a moderately severe coagulopathy is or has recently been present and if other causes of AKI have been excluded by clinical features and serologic tests. A biopsy may be required for definitive diagnosis if creatinine continues to increase or remains significantly elevated above baseline after coagulopathy is corrected. (See 'Diagnosis' above.)

●Among patients who present with hematuria, the differential diagnosis includes all causes of glomerulonephritis and vasculitis. Among those who present without hematuria, one should consider prerenal and postrenal causes, as well as acute tubular necrosis, and other, less common causes of AKI. (See 'Differential diagnosis' above.)

●Patients with ARN should have restoration of a therapeutic INR. The optimal method for correcting excess anticoagulation after the use of warfarin (eg, returning an increased INR to the desired range) generally depends upon the degree of INR elevation, the indication for anticoagulation, and whether clinically significant bleeding is present. The patient should also receive supportive care for AKI. (See 'Treatment' above and "Management of warfarin-associated bleeding or supratherapeutic INR" and "Overview of the management of acute kidney injury (AKI) in adults".)

●The most important measure to prevent ARN is proper adjustment of the anticoagulant dose. This is particularly important for patients with CKD, who are more vulnerable to ARN. (See 'Prevention' above.)

●In most patients with ARN, the serum creatinine stabilizes or improves slightly within the first few weeks after correction of the warfarin coagulopathy. However, some patients may have little or no recovery of kidney function. Presumptive ARN has been associated with increased mortality risk. (See 'Outcome' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Lee A Herbert, MD, who contributed to earlier versions of this topic review.