INTRODUCTION — The association between kidney dysfunction and bleeding was recognized more than 200 years ago [1]. However, there remains an incomplete understanding of the underlying pathophysiology. Impaired platelet function is one of the main determinants of uremic bleeding. This impairment is due largely to incompletely defined inhibitors of platelet function in the plasma of patients with markedly reduced kidney function. Abnormal platelet-endothelial interaction and anemia also play a role.

This topic reviews platelet dysfunction in uremia. It should be noted that the term "uremic" is generally not specifically defined in most studies of this topic review, which typically include patients undergoing regular maintenance dialysis treatments as well as those with chronic kidney disease (CKD) who are not on dialysis. Most of these patients are not overtly "uremic." Platelet dysfunction in other settings is discussed elsewhere. (See "Congenital and acquired disorders of platelet function".)

CLINICAL AND LABORATORY MANIFESTATIONS — Clinical bleeding in uremia may involve the skin, resulting in easy bruising; the oral and nasal mucosa; gingiva; gastrointestinal and urinary tracts; and respiratory system. Excessive bleeding may also occur in response to injury or invasive procedures [2-4].

Uremic patients may display increased bleeding sensitivity to aspirin as there is a transient, cyclooxygenase-independent prolongation of the bleeding time following the use of aspirin in uremic patients that is greater than that seen in normal subjects taking aspirin [5].

Although an association between bleeding time prolongation and uremia has long been suggested [6], there are no good studies that unequivocally demonstrate an increased risk of either spontaneous bleeding or bleeding with a procedure that is associated with a prolonged bleeding time among patients with chronic kidney disease (CKD). The bleeding time is rarely done any longer due to poor standardization, accuracy, reproducibility, and ability to predict bleeding risk.

In addition, there is not a specific level of blood urea nitrogen (BUN) or serum creatinine above which bleeding risk is elevated and below which it is not.

Generally, levels of circulating coagulation factors are normal (or even elevated), and there is no prolongation of the prothrombin or partial thromboplastin times, unless there is a coexisting coagulopathy [7]. Some patients may have a mild thrombocytopenia, but it is rarely severe enough to cause bleeding [7,8].

Other tests of platelet function including aggregation in response to adenosine diphosphate (ADP), collagen/epinephrine closure time, and adenosine triphosphate (ATP) release in response to arachidonic acid may be abnormal in uremic patients [9-13]. These tests are not used routinely to assess bleeding tendency among uremic patients. There are also abnormalities in protein expression in dysfunctional uremic platelets [14].

PATHOGENESIS — The cause of uremic bleeding is multifactorial and includes platelet dysfunction as well as abnormal platelet-endothelial interaction [15-18]. The most important factor is platelet dysfunction. Platelet dysfunction is due to both decreased platelet aggregation and impaired platelet adhesiveness. The impairment in platelet adhesiveness may result at least in part from intrinsic dysfunction of glycoprotein IIb/IIIa, a platelet membrane glycoprotein that normally plays a major role in both platelet aggregation and adhesion by its interaction with fibrinogen and von Willebrand factor [19-21]. There is not a deficiency in von Willebrand factor levels in uremic patients [22].

Other contributing factors that are intrinsic to platelets include abnormal expression of platelet glycoproteins, altered release of adenosine diphosphate (ADP) and serotonin from platelet alpha-granules, faulty arachidonic acid and depressed prostaglandin metabolism, decreased platelet thromboxane A2 generation, and abnormal platelet cytoskeletal assembly. Contributing factors that are extrinsic to platelets include the action of uremic toxins, anemia, increased nitric oxide (NO) and cyclic guanosine monophosphate (GMP) production, functional von Willebrand factor abnormalities, decreased platelet production, and abnormal interactions between the platelet and the endothelium of the vessel wall.

The best characterized of these factors (uremic toxins, anemia, and NO production) are discussed below.

Uremic toxins — The observation that uremic platelets mixed with normal plasma function normally has suggested a plasma factor as the culprit. Consistent with this postulate is the observation that mixing uremic plasma with normal platelets impairs platelet function. The importance of circulating toxins is suggested by the occasional beneficial effect of acute dialysis on platelet dysfunction, although abnormalities in the bleeding time and other in vitro tests of platelet function are rarely normalized [23], and some studies have indicated that dialysis, particularly hemodialysis, does not improve and may even transiently worsen platelet function in some patients [24]. Despite these observations, initiation or intensification of dialysis in severely uremic patients is often recommended if the patient is bleeding or in anticipation of invasive procedures.

Urea is not the major platelet toxin, and there is no predictable correlation between the blood urea nitrogen (BUN) and the bleeding time in patients with kidney failure [6]. The lack of effect of urea on uremic platelet dysfunction is supported by several observations. As an example, the adverse effect of uremic plasma on platelet function cannot be replicated by adding urea, guanidinoacetic acid, or creatinine to the plasma [9]. Additionally, a report describing three family members with familial azotemia, a rare, autosomal dominant condition characterized by high plasma urea levels, demonstrated normal in vitro platelet function [25].

Guanidinosuccinic acid and methylguanidine have been suggested as potential contributors to uremic platelet dysfunction. High levels of guanidinosuccinic acid and methylguanidine have resulted in abnormal platelet function [9], likely through stimulation of NO production [26]. (See 'Nitric oxide' below.)

Anemia — Anemia is a common finding among patients with chronic kidney disease (CKD) and is due primarily to decreased kidney erythropoietin production (see "Treatment of anemia in nondialysis chronic kidney disease" and "Hyporesponse to erythropoiesis-stimulating agents (ESAs) in chronic kidney disease"). Correction of anemia with blood transfusions or erythropoiesis-stimulating agents (ESAs) often improves platelet function. (See 'Correction of anemia' below.)

It has been proposed that rheologic factors play an important role in the overall relationship between anemia and platelet function. At a hematocrit above 30 percent, the red cells primarily occupy the center of the vessel, while the platelets are in a skimming layer at the endothelial surface. This close proximity allows the platelets to adhere to the endothelium and then form a platelet plug when there is endothelial injury. With anemia, on the other hand, the platelets are more dispersed, thereby impairing adherence to the endothelium. Anemia may also contribute to platelet dysfunction through effects on ADP and thromboxane release as well as circulating NO and cyclic GMP concentrations [27,28].

Nitric oxide — NO (endothelium-derived relaxing factor) is an inhibitor of platelet aggregation that is produced by endothelial cells and platelets. Studies in uremic patients have shown that platelet NO synthesis is increased and that uremic plasma stimulates NO production by cultured endothelial cells [29]. The increase in NO synthesis may be due to elevated levels of guanidinosuccinic acid, a uremic toxin that may be a precursor for NO in this setting [26]. Increased NO levels also lead to increased cyclic GMP levels in the blood, which lead to reductions in thromboxane A2 and ADP levels, further impairing platelet aggregation [27].

These findings may be clinically important since the administration of an NO synthesis inhibitor normalizes the bleeding time in uremic rats [30]. The beneficial effect of estrogens on platelet function in uremic patients may be due, in part, to a reduction of NO synthesis. (See 'Estrogen' below.)

TREATMENT — No specific therapy is required in patients without bleeding, even in the setting of severe azotemia. However, correction of platelet dysfunction is desirable in patients who are actively bleeding or who are about to undergo a surgical procedure (eg, kidney biopsy). Desmopressin, discussed below, is often recommended as initial therapy in these situations. Other options are discussed below. Platelet transfusion is generally not recommended for bleeding due to uremic platelet dysfunction, since transfused platelets would similarly acquire the uremic defect. As with any patient, it is important to identify any sources of bleeding and treat those, if possible. (See "Angiodysplasia of the gastrointestinal tract", section on 'Endoscopic treatment' and "Unique aspects of gastrointestinal disease in dialysis patients", section on 'Angiodysplasia'.)

A number of different modalities to improve platelet function and reduce bleeding can be used in this setting, which vary in their onset and duration of action.

Dialysis — Either hemodialysis or peritoneal dialysis can partially correct the bleeding time and other in vitro tests of platelet function in uremic patients [31-33]; whether either dialysis modality actually decreases active bleeding or the risk of major bleeding is unknown. However, dialysis is often recommended in uremic patients with active bleeding or who are in need of invasive procedures. Heparin-free hemodialysis or peritoneal dialysis (which does not require heparin administration) are preferable in patients with active bleeding. If hemodialysis is used, systemic anticoagulation should be avoided, with alternative therapies used, if necessary [7] (see "Anticoagulation for the hemodialysis procedure"). The comparative effectiveness of peritoneal dialysis compared with hemodialysis in improving uremic bleeding has not been studied, although one study suggested that peritoneal dialysis was superior to hemodialysis in correcting in vitro tests of platelet function [32].

Desmopressin (DDAVP) — The simplest, most rapidly acting, and probably least toxic acute treatment for platelet dysfunction in the uremic patient is the administration of desmopressin, an analog of antidiuretic hormone with little vasopressor activity [34-36]. Desmopressin is effective in improving the bleeding time or in vitro tests of platelet dysfunction in some patients and appears to act by increasing the release of large factor VIII:von Willebrand factor multimers from endothelial cells [37]. Other factors may include increases in platelet membrane glycoprotein expression [38]. Desmopressin also improves in vitro tests of platelet function in uremic patients treated with antiplatelet agents [39]. Unfortunately, whether desmopressin actually reduces active bleeding or the risk of clinically important bleeding associated with invasive procedures has not been conclusively demonstrated.

Desmopressin may be given intravenously or subcutaneously at a dose of 0.3 mcg/kg (in 50 mL of saline over 15 to 30 minutes if intravenously). When used for prevention of bleeding related to a procedure, it is administered 30 to 60 minutes before the procedure. The improvement in bleeding time typically begins within approximately one hour and lasts four to eight hours [34]. Tachyphylaxis typically develops after the second dose, perhaps due to depletion of endothelial stores of the factor VIII:von Willebrand factor multimers. Reduced urine volume and hyponatremia may occur in patients who have urine output. Rare thrombotic events have been described following desmopressin administration [40].

Correction of anemia — Raising the hemoglobin to approximately 10 g/dL or higher will reduce the bleeding time in many patients, occasionally to a normal level [41,42], but, as with desmopressin, there has not been demonstration that bleeding or risk of bleeding is ameliorated with correction of anemia. Nonetheless, we generally raise the hemoglobin level close to 10 g/dL in uremic patients prior to high-risk invasive procedures in anemic patients if there is no contraindication. The elevation in hemoglobin level can be achieved acutely by red cell transfusions [42] or chronically via the administration of recombinant erythropoiesis-stimulating agents (ESAs) [27,41]. The improvement in platelet function persists for as long as the hemoglobin remains elevated.

In addition to improving anemia, ESAs may have a direct beneficial effect on platelet function. Possible mechanisms include increasing the number of glycoprotein IIb/IIIa molecules on the platelet membrane, improving the defect in thrombin-induced phosphorylation of platelet proteins, reducing nitric oxide (NO) levels, and improving platelet calcium signaling [43-45]. However, the administration of ESAs has not been consistently associated with changes in platelet number, enhanced platelet aggregation, increased platelet adhesion to endothelial cells, or changes in concentration of large factor VIII:von Willebrand factor multimers or other factors. Thus the observed improvement in the bleeding time may be mostly due mostly to rheologic effects described above. (See 'Anemia' above.)

Estrogen — More chronic control of bleeding can be achieved in many uremic patients by the administration of conjugated estrogens (0.6 mg/kg intravenously daily for five days, 2.5 to 25 mg orally per day, or 50 to 100 mcg of transdermal estradiol twice weekly) [46-50]. Estrogen treatment is most commonly indicated in patients on dialysis who have chronic gastrointestinal tract bleeding not amenable to other treatment, such as angiodysplasia or multiple colonic polyps. These agents begin to act on the first day, with peak control reached over five to seven days; the duration of action is one week or more after therapy has been discontinued [46,48,49]. Intravenous estrogens have been best studied and appear to have the most reproducible effects, so are recommended as primary therapy rather than oral or transdermal therapy [27,46,48,49].

The effect of estrogens is dose dependent (ie, the beneficial effect is not seen at doses of 0.3 mg/kg) [49] and is primarily mediated by estradiol, acting via the estrogen receptors [51]. The long-term use of this regimen is limited by estrogen-related side effects. It is possible, however, that related compounds with weak estrogenic activity would also be beneficial [50].

The safety of estrogen use among men with uremic platelet dysfunction is not known with certainty. Most studies of estrogens in this setting have been of limited duration. Men have been treated with short-term estrogens, with hot flashes being the primary side effect [50,52]. Thus, it would appear to be safe to use estrogens in men, if needed, for up to five days.

The mechanism by which estrogens act is not well understood, but may be due to decreased generation of NO [53].

Cryoprecipitate — The infusion of cryoprecipitate (10 units intravenously every 12 to 24 hours) can shorten the bleeding time in many uremic patients [54]. The improvement in bleeding time begins within one hour and lasts 4 to 24 hours; it is presumably mediated by the presence in cryoprecipitate of a substance that enhances platelet aggregation, such as factor VIII:von Willebrand factor multimers or fibrinogen (see "Clinical use of Cryoprecipitate"). The potential risk of infectious complications from the use of cryoprecipitate has limited use of this modality to patients with life-threatening bleeding who are resistant to treatment with desmopressin and blood transfusions.

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: Chronic kidney disease in adults".)

SUMMARY AND RECOMMENDATIONS

●Impaired platelet function is a major determinant of uremic bleeding. Bleeding may involve the skin, resulting in easy bruising, or the oral and nasal mucosa and gingiva. Gastrointestinal bleeding and hematuria may also occur. (See 'Introduction' above and 'Clinical and laboratory manifestations' above.)

●Patients may display increased sensitivity to aspirin, although other coagulation parameters are generally intact. The platelet count is usually normal, and there is no prolongation of the prothrombin or partial thromboplastin times, unless there is a coexisting coagulopathy. (See 'Clinical and laboratory manifestations' above.)

●Causes of platelet impairment include intrinsic platelet defects, abnormal platelet-endothelial interaction, uremic toxins, and anemia. (See 'Introduction' above and 'Pathogenesis' above.)

●Patients who are actively bleeding or who are about to undergo a surgical procedure should have correction of platelet dysfunction. Treatment options include correction of anemia, desmopressin (DDAVP), dialysis, estrogens, or cryoprecipitate. Therapies vary in their onset and duration of action, and most have been shown only to reduce the bleeding time or in vitro tests of platelet function rather than to reduce active bleeding or the risk of bleeding with invasive procedures (see 'Treatment' above):

•Raising the hemoglobin to approximately 10 g/dL may reduce the bleeding time. The improvement in platelet function will persist for as long as the hemoglobin remains elevated. Erythropoiesis-stimulating agents (ESAs) may also have a direct beneficial effect on platelet function. (See 'Correction of anemia' above.)

•Desmopressin provides the simplest and most rapid acute treatment for platelet dysfunction in the uremic patient. Administration of desmopressin at a dose of 0.3 mcg/kg given in 50 mL of saline over 15 to 30 minutes intravenously or by subcutaneous injection is preferred. The improvement in bleeding time begins within one hour and lasts four to eight hours. The response to subsequent doses is generally diminished (tachyphylaxis). (See 'Desmopressin (DDAVP)' above.)

•Either hemodialysis or peritoneal dialysis can partially correct the bleeding time in approximately two-thirds of uremic patients. Hemodialysis should be performed without systemic anticoagulation. (See 'Dialysis' above.)

•Prolonged control of bleeding may be achieved by the administration of conjugated estrogens (0.6 mg/kg intravenously daily for five days, 2.5 to 25 mg orally per day, or 50 to 100 mcg of transdermal estradiol twice weekly). These agents begin to act on the first day, with peak control reached over five to seven days; the duration of action is one week or more after therapy has been discontinued. (See 'Estrogen' above.)

•The infusion of cryoprecipitate (10 units intravenously every 12 to 24 hours) can shorten the bleeding time in many uremic patients. The improvement in bleeding time begins within one hour and lasts 4 to 24 hours. Potential infectious complications limit the use of cryoprecipitate to patients with life-threatening bleeding who are resistant to treatment with desmopressin and blood transfusions. (See 'Cryoprecipitate' above.)

ACKNOWLEDGMENT — The editorial staff at UpToDate would like to acknowledge Steven Coutre, MD, who contributed to an earlier version of this topic review.