The differential diagnosis of the thrombotic microangiopathies includes disease processes presenting with thrombocytopenia and MAHA (Table 70-1). Most commonly, the clinician must distinguish between TTP/HUS and the DIC syndromes or distinguish TTP/HUS from a collagen vascular disease (CVD) such as systemic lupus erythematosus (SLE) with vasculitis (Table 70-2). DIC may be associated with many diseases (see Table 70-1) and results in widespread generation and deposition of intravascular fibrin in small blood vessels, accompanied by a consumptive coagulopathy with clotting factor activation as a primary process. This condition is to be distinguished from TTP/HUS, in which the primary process is platelet aggregation at the vascular endothelial surface.
Distinguishing Thrombotic Thrombocytopenic Purpura from Disseminated Intravascular Coagulation
DIC may be initiated by either endothelial injury (the intrinsic coagulation pathway) or release of tissue thromboplastin (the extrinsic coagulation pathway). Either process can cause platelets to undergo intravascular aggregation when exposed to collagen and thrombin. The clinical manifestations of DIC arise from injury related to intravascular thrombosis or from bleeding that results from consumption and depletion of clotting factors and platelets. Acute DIC syndromes usually present with bleeding and demonstrable hypo- or afibrinogenemia. In more chronic forms of DIC in which reticuloendothelial system clearance mechanisms, clotting factor production, and marrow cell production compensate for consumption, patients can present with thrombosis rather than hemorrhage. Typical thrombotic problems include deep vein or superficial thrombophlebitis, pulmonary embolus, cerebrovascular accidents, or nonbacterial (marantic) endocarditis.
DIC shares with TTP the features of MAHA and thrombocytopenia (see Table 70-2). In one series, the frequency of MAHA as judged by peripheral smear was 68% in patients with DIC. In the same series, thrombocytopenia (platelet count <150,000/μL) was seen in 96% of cases of DIC; however, significant thrombocytopenia (platelet count <50,000/μL) was much less common (57%).
A number of laboratory findings will help to distinguish TTP from DIC (see Table 70-2). As noted above, hypofibrinogenemia is typical of fulminant, consumptive DIC. In addition, as clotting factors are consumed, prolongation of the prothrombin time (PT) and partial thromboplastin time (PTT) is seen. Also, the systemic generation of fibrin can lead to activation of the fibrinolytic system, either by activation of plasminogen by tissue activators released from vascular endothelium or by Hageman factor–dependent activation. Fibrinolysis results in the appearance of fibrin degradation products (FDPs) in the serum that can be measured. While none one of these laboratory tests alone can distinguish TTP from DIC, in the aggregate they are sufficient to permit a specific diagnosis (see Table 70-2 and Chap. 69).
Finally, the clinical context will often help to determine if DIC is a viable explanation for the observed abnormalities. Most often, the disease or disorder that is the precipitant of DIC is apparent. Amniotic fluid embolus, retained products of conception, and abruptio placentae are obstetric complications often associated with DIC. Septicemia is the substrate for DIC in many patients encountered in the ICU. The presentation of these patients may be fulminant, as in the Waterhouse-Friderichsen syndrome (shock, bleeding diathesis, and adrenal insufficiency in association with meningococcemia), or more indolent, as in the complex postoperative patient with a “smoldering” intra-abdominal source of infection. Trauma is a well-recognized cause of DIC and fibrinolysis resulting from the exposure of tissue thromboplastins and plasminogen activators to plasma.37 In trauma patients, unexplained bleeding during or after a surgical procedure may be the first manifestation of DIC. In the setting of head injury, as many as 70% of patients have clinical evidence of DIC.38
In summary, TTP/HUS can usually be distinguished readily from DIC by the clinical context and laboratory parameters. It is important to apply this principle, as all of the defining features of TTP/HUS—thrombocytopenia, MAHA, fever, neurologic dysfunction, and renal abnormality—are so common in the critically ill that a clear approach to this possible diagnosis is warranted.
Approach to the Pregnant Patient
Hematologic aberrations associated with PIH (pre-eclampsia/eclampsia; see Chap. 105) include MAHA, thrombocytopenia, and alterations of the coagulation mechanism. Evidence of MAHA can be found in 2% to 15% of women with PIH, and thrombocytopenia has been reported in as many as 18% of patients. Thus a subset of patients with PIH will have two of the cardinal hematologic abnormalities of TTP/HUS.5 It has been suggested that PIH and TTP/HUS share pathophysiologic features, specifically aberration in prostaglandin metabolism at the platelet-endothelial cell interface. In this regard it is interesting that plasma exchange has been reported to be a successful therapy in PIH. A subset of patients with PIH and thrombocytopenia may have marked elevation of liver function tests (HELLP syndrome). Right upper quadrant pain is often present and may mimic cholecystitis or peritonitis.
Postpartum HUS is characterized by predominant renal involvement; neurologic signs and fever are usually absent. It has been suggested that this syndrome is a clinical counterpart of the generalized Schwarzman reaction. It is hypothesized that bacterial endotoxins or vasoactive amines are discharged into the maternal circulation and either stimulate the coagulation cascade or initiate thrombosis by damage to the vascular endothelium.
As mentioned above, DIC syndromes accompany major complications of pregnancy. Usually these processes are associated with a catastrophic process such as amniotic fluid embolus or abruptio placentae. Less apparent and more indolent DIC may be encountered with retained products of conception.
Finally, TTP/HUS itself is encountered in pregnancy but is rare, with approximately 70 cases reported. It is likely that most pregnant patients with thrombocytopenia and MAHA are best classified as having PIH with hematologic abnormalities.10 The key features distinguishing these patients from the group with HELLP syndrome are the emergence of disease in late pregnancy, relatively mild hematologic irregularities, and prompt resolution of thrombocytopenia, MAHA, neurologic symptoms, and liver function abnormalities following delivery. HUS begins postpartum, with a predominance of renal dysfunction. DIC differs from TTP as described above (see Table 70-2). The small remaining group of patients whose clinical presentation is best described as TTP/HUS are best managed in pregnancy with the usual treatments (see below), with expeditious delivery or termination of pregnancy.39,40
Distinguishing Thrombotic Thrombocytopenic Purpura from Collagen Vascular Disease
An association between TTP/HUS and SLE has been reported. In one review, evidence of SLE was found in 7 of 64 cases initially diagnosed as TTP/HUS.1 TTP/HUS has also been reported in association with other CVDs, including rheumatoid arthritis, ankylosing spondylitis, and polyarteritis nodosa. On the other hand, vasculitis in association with SLE or another CVD may mimic TTP/HUS with findings of renal failure, fever, neurologic disturbance, thrombocytopenia, and MAHA. Serum complement levels are usually low in patients with vasculitis and normal in those with TTP/HUS (see Table 70-2). Antinuclear antibodies are positive in the great majority of patients with SLE. These CVD screening tests are indicated in all patients with a tentative diagnosis of TTP/HUS. If a diagnosis of a specific CVD can be made, therapy should be directed at that disorder, rather than at the associated hematologic problems.