Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Of the circulating blood cells, the platelet was the last to be fully described and its attributes determined. Although early studies by Osler, Hayam, and Bizzozero had identified small particles in the blood, these were thought to be bacteria, fragments of red blood cells (RBCs), or other hematopoietic elements. It was not until the development of a novel blood staining method by James Homer Wright that the true identity of these circulating blood cells and their relationship to hemostasis became apparent. By observing their common tinctorial properties, Wright demonstrated that blood platelets (initially called plates ) arose from bone marrow megakaryocytes. He found that these megakaryocytes extended a portion of their cytoplasm into the bone marrow sinusoids and shed platelets into the circulation ( Fig. 7.1 ).
These observations were carried one step further in 1910 by William Duke, who described three patients at the Massachusetts General Hospital who were bleeding and had low platelet counts, as determined by early cell counting procedures. He was able to demonstrate that “certain types of hemorrhagic disease may be attributed to an extreme reduction in the number of platelets.” Indeed, Duke showed that after an arteriovenous shunt was created between a healthy donor and a thrombocytopenic recipient, the platelet count would increase in the thrombocytopenic recipient and bleeding would cease. Although platelet transfusions had unknowingly been previously given in the form of whole blood transfusions, this was the first time it was shown that transfused platelets could ameliorate bleeding.
Since the time of these seminal observations, evaluation of thrombocytopenia has become a common hematology consultation. On a general inpatient hematology consultation service, approximately one-third of all consultations are called to assess thrombocytopenia (D. Kuter, personal observation, 2013). Some 5% to 10% of all hospitalized patients are thrombocytopenic, and for patients in medical and surgical intensive care units (ICUs), this figure rises to as high as 30% to 35%. Indeed, some data suggest that thrombocytopenic patients suffer a twofold higher mortality than those who are not thrombocytopenic.
The primary reason for evaluating any thrombocytopenic patient is to assess the risk of bleeding. In general, patients with platelet counts less than 20,000/µL are at increased risk of spontaneous bleeding and bleeding with procedures. These individuals are commonly the subjects of consultative hematology evaluations, since they may need treatment with transfusions or more specific therapies to ameliorate the bleeding risk.
Patients with milder degrees of thrombocytopenia with platelet counts from 20,000 to 50,000/µL rarely have any risk of spontaneous bleeding, but they may have increased bleeding risk with procedures. Such patients often require hematology consultation, especially if procedures are planned.
Patients with platelet counts ranging from 50,000 to 100,000/µL do not have an increased risk of spontaneous bleeding and can usually undergo most procedures without an increased risk of major bleeding complications. Nonetheless, this patient population is often strikingly limited in its access to medical care. Surgeons are frequently reluctant to operate on patients whose platelet counts are in this range, albeit for poorly documented reasons. Epidural anesthesia is often withheld for pregnant patients with platelet counts below 100,000/µL. Procedures as routine as colonoscopy, dental extraction, dental prophylaxis, prostate biopsy, and breast biopsy are often not undertaken in patients whose platelet counts fall in this range. Effective antiviral treatment for hepatitis C may be withheld from individuals with mild thrombocytopenia. Such patients often indirectly benefit from hematology consultation when practitioners are reassured of the extremely low risk of their actions.
A final group of thrombocytopenic patients, those with platelet counts between 100,000/µL and the usual lower limit of normal of 150,000/µL, deserves comment. Although platelet counts in this range are common, such patients rarely have bleeding symptoms and rarely present any bleeding risk with procedures. In the absence of other cytopenias, the mild thrombocytopenia may reflect mild autoimmune disorders, splenomegaly, early bone marrow conditions, medication effects, previous infections, or ethnic variations. Because recent data suggest that less than 5% of such patients experience a progression of their thrombocytopenia, the diagnosis of immune thrombocytopenia (ITP) now requires a platelet count under 100,000/µL. Such minimally thrombocytopenic patients and their providers may require reassurance, but hematology consultation and further evaluation is generally reserved for those who demonstrate a trend of declining platelet counts, have a count that falls below 100,000/µL, or have symptoms of other diseases.
Aside from the risk of bleeding, a second reason for consultation on the thrombocytopenic patient is to determine whether some other underlying medical condition is involved. Thrombocytopenia is often a common presenting sign of other diseases, such as systemic lupus erythematosus (SLE), primary bone marrow disorders, hematological malignancies (myeloid leukemia, lymphoproliferative disorders), chronic hepatitis, and splenomegaly. The etiology of the thrombocytopenia may be medically of more importance than the platelet count itself. The customary reluctance to evaluate patients with platelets between 100,000 and 150,000/µL may change; new data suggest that some patients with chronically low platelet counts (presumably from an inherited thrombocytopenia) have an increased risk of clonal hematological disorders later in life.
Of added interest for the hematology consultant is the thrombosis that paradoxically accompanies thrombocytopenic disorders of increased platelet turnover, such as thrombotic thrombocytopenic purpura (TTP), hemolytic-uremic syndrome (HUS), heparin-induced thrombocytopenia (HIT), antiphospholipid antibody syndrome (APLS), and ITP. Although ITP certainly confers a major risk of bleeding, it is also associated with an increased relative risk of thrombosis, ranging from 1.4 to 2.65 ; the thrombosis rate increases as the platelet count decreases.
The goals of the hematology consultant in evaluating the thrombocytopenic patient are:
To assess the risk of bleeding
To diagnose the underlying cause of the thrombocytopenia
To treat the thrombocytopenia as indicated
A platelet count above 100,000/µL is not associated with any significant bleeding risk and is rarely the subject of hematology consultation (unless it is part of the assessment for some other disease such as HIT, APLS, collagen vascular disorders, or hematologic malignancies).
For the purposes of this discussion, clinically significant thrombocytopenia is considered to occur when platelet counts are lower than 100,000/µL. As suggested earlier, such subjects can be roughly grouped into three categories. The first are those whose platelet counts are between 50,000 and 100,000/µL, in whom spontaneous bleeding does not occur and in whom the surgical bleeding risk is quite low. The second group includes those whose platelet counts are between 20,000 and 50,000/µL; for this group spontaneous bleeding rarely occurs, but the risk of bleeding with surgical procedures may be increased. Finally, the group that is of greatest concern is those whose platelet counts are lower than 20,000/µL, whose risk of spontaneous bleeding is increased, and for whom surgical bleeding risks are usually increased.
Although seemingly intuitive, the relation of the platelet count to bleeding risk is poorly defined. This is due not only to inadequate clinical studies but also to the inability of clinicians to accurately measure the second important platelet variable, platelet function. Platelet function is determined by many variables, including platelet size and age, intrinsic platelet function defects, the levels of plasma factors (e.g., von Willebrand factor [vWF]), exposure to medications (e.g., aspirin), and the presence of toxins (e.g., uremia). Reduced platelet function will increase the bleeding risk at any given platelet count. However, in some disorders of increased platelet destruction, such as ITP, as the platelet count declines, mean platelet volume rises, which tends to offset the decline in platelet count. This increase in mean platelet volume has been attributed to “phylogenetic canalization,” which suggests some feedback system in which the increased platelet volume (and hence increased function) tends to mitigate the decreased platelet numbers. In general, at equally low platelet counts, in disorders of increased platelet production, the platelets are larger, younger, and more functional; in disorders of reduced platelet production, the opposite is true and the bleeding risk is increased.
Although it is convenient to think of hemostatic risk solely as a function of the platelet count and platelet function, this is an oversimplification due to the many other variables (fever, infection, procedures, blood pressure, medications, coagulation factor abnormalities) that affect hemostatic risk. Rigid adherence to the customary threshold values of 50,000/µL for surgical hemostasis and 5000 to 10,000/µL for prophylactic platelet transfusion is inappropriate. Nonetheless, these platelet numbers are generally helpful and are based on the following evidence.
One example of the increased risk of bleeding with thrombocytopenia is seen in studies that showed the relation of the bleeding time to platelet count. Below a platelet count of 100,000/µL, a linear relation is observed between the decline in platelet count and the increase in bleeding time ( Fig. 7.2 ). Although the bleeding time is an unreliable predictor of clinical bleeding risk, this correlation is perhaps the clearest visual demonstration of the relation between a decline in platelet count and the increase in bleeding risk.
Early studies in leukemic children demonstrated a direct relation between the platelet count and the risk of spontaneous bleeding. As the logarithm of the platelet count fell below 100,000/µL, a linear increase was reported in the amount of hemorrhage that occurred. Most of this was accounted for by milder forms of hemorrhage, such as petechiae, ecchymoses, and epistaxis, which tended to occur when the count was below 50,000/µL. If only more major forms of hemorrhage were analyzed, an increase was evident as counts fell below 100,000/µL, but most bleeding events occurred below platelet counts of 10,000/µL; of this latter group, most bleeding events occurred below 5000/µL. For both minor and major bleeding episodes, the authors emphasized that no threshold platelet count existed; rather, a continuous increase in hemorrhagic risk was noted as the platelet count fell. The results of these often-quoted studies are confounded by the fact that many of these subjects were also treated with antipyretics that adversely affected platelet function and the use of capillary platelet counts, and there was a lack of adequate antibiotic treatment of these often febrile patients. Nonetheless, the findings of these studies support the relation of bleeding risk and platelet count and have been used (despite the authors' exhortations) to support the concept of the 50,000/µL threshold for surgical hemostasis and the 5000 to 10,000/µL threshold for prophylaxis.
Other platelet transfusion studies in leukemic patients who received chemotherapy have demonstrated that hemorrhage occurs to the same extent at platelet counts of 10,000/µL and at 20,000/µL.
A recent trial assessed the effect of the platelet count on bleeding (using a validated bleeding scale) in thrombocytopenic patients undergoing myeloablative chemotherapy for leukemia or stem cell transplantation ( Fig. 7.3 ). It clearly showed bleeding of grade 2 or higher on 25% of days with platelet counts of 5000/µL or less, on 17% of days with platelet counts from 6000 to 80,000/µL ( P < .001 for platelet counts of 5000/µL or less vs. counts of 6000 to 80,000/µL), on 13% of days with platelet counts of 81,000 to 100,000/µL ( P = .001 for platelet counts of 81,000 to 100,000/µL vs. counts of 6000 to 80,000/µL), and on 8% of days with platelet counts above 100,000/µL ( P < .001 for platelet counts of above 100,000/µL vs. counts of 6000 to 80,000/µL).
A biological estimate of the lowest effective platelet count comes from the work of Slichter and colleagues, who used RBCs labeled with chromium 51 to quantify fecal blood loss in thrombocytopenic aplastic patients in stable condition who were treated only with anabolic steroids. At platelet counts above 10,000/µL, patients had a normal blood loss of less than 5 mL/day. At platelet counts of 5000 to 10,000/µL, this loss rose slightly to 9 ± 7 mL/day; however, at platelet counts below 5000/µL, the loss was markedly elevated to 50 ± 20 mL/day.
To assess this apparent critical platelet threshold of 5000 to 10,000/µL further, Hanson and Slichter performed platelet kinetic studies in thrombocytopenic patients with platelet counts ranging from 12,000 to 70,000/µL. They found a fixed minimum requirement for 7100 platelets/µL per day to maintain vascular integrity; this was 18% of the normal daily turnover of 41,200 platelets/µL per day. These studies have provided the experimental justification for the current recommendations that prophylactic platelet transfusions be given only to those patients whose platelet counts are lower than 5000 to 10,000/µL.
One final pathophysiological basis for the above bleeding risk recommendations is data looking at the precise role that platelet surface activation plays in the coagulation cascade. When real-time measurements during clotting are used, thrombin generation appears to be maximal as long as the platelet count is above 10,000/µL; below that value thrombin generation declines in direct proportion to the platelet count.
A common hematology consultation is for the preoperative patient with thrombocytopenia. The potential risk of bleeding with surgery is difficult to assess for many patients, let alone those only with thrombocytopenia. As discussed earlier, there are many contributing factors that determine bleeding risk, of which thrombocytopenia is just one. The British Committee for Standards in Haematology has appropriately made recommendations for platelet transfusion for many procedures (see later) based on a consensus assessment of bleeding risk with various surgical procedures. These guidelines for transfusion have been widely applied, but they may be overly restrictive for patients with ITP where the reduction in platelet count may be accompanied by a rise in platelet size. Separate guidelines have been proposed for patients with ITP.
However, evidence to support these guidelines is hard to obtain. Few studies have correctly identified a platelet count threshold that is predictive of bleeding with procedures, and none have shown that prophylactic platelet transfusion is beneficial. In one study of 150 patients with platelet counts 100,000/µL or less undergoing bronchoalveolar lavage, prophylactic platelet transfusions were given to 65% of the 89 patients with platelets less than 50,000/µL and to only 8% of the 61 patients with platelet counts 50,000 to 100,000/µL. Bleeding at the time of the procedure was low and unrelated to the platelet count or platelet transfusion: 0% at less than 20,000/µL; 0% at 20,000 to 49,000/µL; and 1% at 50,000 to 100,000/µL. The authors recommended prophylactic platelet transfusion only for those 20,000 to 30,000/µL. In 50 patients with liver disease and platelets under 150,000/µL undergoing procedures prior to orthotopic liver transplantation, bleeding occurred in 0/18 (0%) patients with platelet counts 75,000 to 149,000/µL, but in 10/32 (31%) with platelets less than 75,000/µL ( P = .008). There was no relation of bleeding to underlying coagulation factor abnormalities. Finally, in 2654 patients with liver disease undergoing percutaneous liver biopsy, bleeding occurred in 3/1331 (0.2%) of those with platelet counts greater than 150,000/µL; in 5/738 (0.7%) of those with platelet counts 101,000 to 150,000/µL; in 3/509 (0.6%) of those with platelet counts 61,000 to 149,000/µL; but in 4/76 (5.3%) of those with platelet counts 60,000/µL or less.
It is unlikely that there will ever be clinical studies in thrombocytopenic patients to clarify the risk of bleeding, let alone the benefit of raising the platelet count, with most common surgical procedures. This is not an unfamiliar situation for the hematologist consulting on surgical patients, but it should not prevent the application of good clinical judgment and guidelines in the care of these patients.
To provide a pathophysiologic basis for the clinical evaluation of the thrombocytopenic patient, a brief review of the biology of platelet production is helpful. This approach allows the hematology consultant to relate the various causes of thrombocytopenia to relevant steps of platelet production ( Fig. 7.4 ).
The pluripotential stem cell gives rise through a stochastic differentiation process to precursor cells committed to megakaryocyte differentiation, called megakaryocyte colony-forming cells . Megakaryocyte colony-forming cells are mitotically active until some triggering event, as yet unidentified, causes them to stop their mitotic divisions and enter a process called endomitosis , in which DNA replication ensues, but neither the nucleus nor the cytoplasm undergoes division. This gives rise to polyploid megakaryocyte precursor cells that contain anywhere from 4 to 16 times the normal diploid complement of DNA—all contained within a single nuclear envelope. Initially, these cells are morphologically indistinct, but once they complete their endomitotic divisions, they grow into large, morphologically identifiable megakaryocytes.
Megakaryocytes occupy unique positions within the bone marrow. Early megakaryocyte precursor cells and stem cells occupy a niche close to the bone. As the megakaryocytes differentiate, they appear to follow a stromal cell-derived factor-1 (SDF-1) gradient and migrate close to the endothelial cells that line the bone marrow sinusoids. Cytoplasmic projections from the megakaryocytes then pass through the endothelial cell—not between its gap junctions—and appear in the bone marrow sinusoid, where they form long strands of megakaryocyte cytoplasm (called proplatelets ) ( Fig. 7.5 ) that are destined to become platelets. Whether individual platelets are then produced from proplatelets in the bone marrow sinusoids or occur in other tissues, such as the lungs, has been the subject of much speculation for decades. Mathematical models have suggested that most individual platelets are produced in the lung parenchyma, but cell biology studies have not supported this hypothesis until recently. Using intravital microscopy of green fluorescent protein-labeled megakaryocytes in mice, recent work suggests that approximately 50% of platelet production is accounted for by the migration of bone marrow megakaryocytes to the lungs where they release platelets.
Once in the circulation, the human platelet survives for 10 days; it then probably undergoes programmed cell death and is removed from the circulation. A companion hypothesis suggests that loss of surface sialic acids from platelets during aging produces asialyated platelets that are then removed from the circulation by the Ashwell-Morrell receptor on hepatocytes. No evidence suggests that platelet activation plays a major role in platelet clearance; indeed, most platelets that enter the circulation never undergo platelet activation before they undergo senescence and clearance. The tissue responsible for the clearance of senescent platelets has not been well determined. In animal models, splenectomy does not seem to alter the platelet lifespan; therefore one can assume that the clearance in humans also occurs by the reticuloendothelial system in throughout the body.
The key hematopoietic regulators of platelet production appear to be thrombopoietin (TPO) and SDF-1. TPO is necessary for megakaryocyte growth, and in its absence, platelet counts in animals and humans drop to about 10% of normal. Nonetheless, platelets continue to be made, albeit from a reduced number of low ploidy megakaryocytes. TPO appears to be made in a constitutive fashion by the liver, and its rate of synthesis is not affected by any known cytokine or disease. The only exception appears to be the reduction in TPO production that occurs in patients with liver dysfunction, such as chronic hepatitis and after partial hepatectomy ; in this setting, the platelet count declines in direct proportion to the reduction in functional liver volume. Once synthesized, TPO enters the circulation and is cleared by avid TPO receptors on platelets and probably bone marrow megakaryocytes. This results in a basal level of TPO that is necessary for maintaining the viability of stem cells, increasing the mitotic rate of megakaryocyte colony-forming cells, increasing megakaryocyte endomitosis and megakaryocyte maturation, thereby increasing platelet production.
The relation of circulating TPO levels to the platelet count depends on the nature of the thrombocytopenia. In situations in which the marrow has been damaged and platelet production is decreased, TPO clearance is decreased and serum TPO levels rise. For example, in patients with aplastic anemia with platelet counts of 10,000/µL, TPO levels rise from normal values of about 100 pg/mL to 2000 to 3000 pg/mL. However, when thrombocytopenia is due to peripheral destruction of platelets, and when the megakaryocyte mass is normal or increased, net clearance of TPO is normal and TPO levels are not significantly elevated. An example is ITP, in which the increased bone marrow megakaryocyte mass and the normal or slightly increased release of platelets into the blood result in normal TPO clearance and normal TPO levels. Measurement of TPO levels may be helpful in distinguishing between patients with decreased or increased platelet production ( Fig. 7.6 ). TPO assays are now clinically available.
Less well appreciated is the role SDF-1 plays in platelet production. When SDF-1 was administered to thrombocytopenic animals that lacked TPO, the platelet count rose to nearly normal. In normal healthy animals in which SDF-1 was transiently removed, the platelet count fell. SDF-1 appears to guide megakaryocyte progenitors to the bone marrow sinusoids and to trigger their shedding of proplatelets. The main unresolved issue in platelet biology is the mechanism of platelet formation from bone marrow megakaryocytes. This appears to be a very finely tuned mechanism that involves TPO, SDF-1, and endothelial cells.
As is discussed later, with this understanding of platelet biology, thrombocytopenia will occur by two general mechanisms:
Platelet destruction due to immune or nonimmune mechanisms overwhelms the compensatory ability of the bone marrow to increase platelet production up to sixfold.
Platelet production is reduced by disorders that inhibit specific steps in the production of platelets—stem cells, megakaryocyte colony-forming cells, megakaryocyte maturation, platelet shedding from megakaryocytes, platelet viability.
Disorders involving both mechanisms (e.g., ITP) are common.
Table 7.1 lists a general classification of the causes of thrombocytopenia.
|
Laboratory artifact (“pseudothrombocytopenia”) occurs in up to 0.2% of patients who have blood counts performed. Although this may be due to the use of therapeutic antiplatelet antibodies, such as abciximab, most cases of pseudothrombocytopenia result from the clumping of platelets that occurs ex vivo in anticoagulated blood samples. Several mechanisms have been proposed for this. Most involve conformational changes in the GPIIb–IIIa receptor that are due to the low divalent cation concentration and/or the lower temperature of anticoagulated blood; novel GPIIb–IIIa epitopes appear to be exposed, which react with preexisting antibodies in the patient's blood, causing aggregation. In both of these situations, collecting the blood in acid-citrate-dextrose (ACD; yellow top) or heparinized (green top) tubes, as well as keeping the samples at 37°C (98.6°F), usually prevents clumping and permits an accurate blood count. Although most modern cell counting devices will flag samples that contain platelet clumps, review of the peripheral blood smear may be the only way to detect this phenomenon ( Fig. 7.7 ). Certainly, patients who have a very low reported platelet count but who lack signs and symptoms of thrombocytopenia should be evaluated for pseudothrombocytopenia through review of the peripheral blood smear.
A second important, but easily diagnosed, cause of thrombocytopenia occurs primarily in ICU and post-surgical patients who have received multiple RBC transfusions and developed dilutional thrombocytopenia. Fresh whole blood is rarely used anymore, and transfusion with large amounts of packed RBCs, fresh frozen plasma (FFP), and intravenous fluids may result in dilution of the platelets. In massive trauma, it is important that adequate platelet transfusions be provided along with RBC and plasma transfusions (see Chapter 40 ).
A third cause of mild thrombocytopenia (with platelets usually in the 40,000 to 60,000/µL range) is the splenic sequestration commonly seen in patients with severe liver disease or other causes of splenomegaly. Because the body conserves the circulating platelet mass and not the platelet count, approximately one-third of the total platelet mass is normally sequestered in the spleen. With splenic enlargement, additional platelets become sequestered in the spleen. In patients with liver disease and splenomegaly, the situation may be more complicated; because the liver is the main source of TPO production, thrombocytopenia can be attributed to splenic sequestration and diminished TPO levels.
As with many other hematological conditions, the two remaining major categories of thrombocytopenia involve decreased production of platelets or increased destruction of platelets, or some combination of the two.
Decreased platelet production occurs in many situations, ranging from replacement of bone marrow by metastatic cancer or hematological malignancy to a lack of bone marrow due to bone marrow failure syndromes. Solitary thrombocytopenia may be present for years before a diagnosis of myelodysplasic syndrome (MDS) is made and is present in about two-thirds of MDS patients at diagnosis.
Toxins, ethanol ingestion, vitamin B12 deficiency, and use of certain medications can decrease megakaryocyte endomitosis and megakaryocyte maturation (see Fig. 7.4 ).
Cytotoxic chemotherapy can reduce platelet production in several ways. Many drugs (e.g., busulfan, carboplatin) affect stem cells but some (e.g., cyclophosphamide) spare stem cells. Other drugs (e.g., bortezomib) can actually decrease the shedding of platelets from existing megakaryocytes. With bortezomib, the number of megakaryocytes may be normal or elevated, although effective platelet production (thrombopoiesis) may be reduced, possibly due to the inhibition of NF-κB or the disruption of the SDF-1 gradient. Finally, a few drugs can directly induce platelet apoptosis.
It should not be forgotten that ITP is also a disease of inappropriately low platelet production in that megakaryocytes may be undergoing programmed cell death caused by antiplatelet antibodies or cytotoxic T cells. In thrombocytopenia associated with human immunodeficiency virus (HIV) infection, megakaryocyte mass, and megakaryocyte ploidy are markedly increased, but effective thrombopoiesis from these megakaryocytes is markedly diminished, presumably because of early programmed cell death of these megakaryocytes. Finally, liver resection or severe liver disease may decrease TPO production.
Disorders of increased platelet destruction are relatively common and include both nonimmune and immune disorders. Medications may be involved in both of these categories. Nonimmune thrombocytopenic disorders include disseminated intravascular coagulation (DIC), TTP, and HUS, as well as pulmonary hypertension, venoocclusive disease, and full-thickness burns. Mechanical devices such as continuous venovenous hemofiltration, intraaortic balloon pump counterpulsation, and extracorporeal membrane oxygenation can also increase platelet clearance. Platelets also may be triggered to undergo apoptosis with rapid clearance from the circulation in dengue fever and in patients exposed to novel chemotherapeutics (e.g., ABT-737) that inactivate Bcl-X L .
While thrombocytopenia due to bacterial sepsis is often due to increased platelet clearance by DIC, an additional mechanism may also occur; many bacteria, such as streptococcus, release neuraminidase that removes surface sialic acids from platelets. The desialylated platelets then bind to and are removed by the hepatic Ashwell-Morrel receptor, an asialoglycoprotein receptor that regulates the homeostasis of many plasma proteins, such as vWF. This receptor has also been thought to play a major role in the normal clearance of senescent platelets from the circulation.
Immune causes of thrombocytopenia can be divided into those that are related to antigen-antibody complex deposition onto platelet Fc receptors and those in which the antibody Fab region directly binds to the platelet. An example of the former is HIT, in which immune complexes cause platelet activation. An example of the latter is ITP, in which antibodies are directed against GPIIb–IIIa and GPIb–IX glycoproteins on the platelet surface; this produces opsonization of platelets and their removal by the FcγRIII receptors on macrophages in organs, such as the spleen.
ITP may be primary (not associated with any other known disease) or secondary to many other disorders, including autoimmune diseases (e.g., SLE, APLS) and lymphoproliferative diseases (see Chapter 8 ). ITP and other autoimmune cytopenias are commonly associated with lymphoproliferative disorders (chronic lymphocytic leukemia [CLL], Hodgkin lymphoma, non-Hodgkin lymphoma). ITP occurs in approximately 2% of patients with CLL and 1% of patients with other forms of non-Hodgkin lymphoma. ITP may be the first sign of lymphoproliferative disease and may occur many years before the diagnosis of malignancy is actually made. Importantly, monoclonal B-cell lymphocytosis is found in 5.1% of patients with a normal blood count and is associated with a 1.1% yearly rate of conversion to CLL. Lymphoproliferative diseases should always be considered in evaluating a patient for ITP.
For the hematology consultant, the urgency and pace of the evaluation are determined by the platelet count, the extent of bleeding or thrombosis, the need for procedures, the presence of antiplatelet agents, the extent of nonhematological symptoms, and the concurrent anemia and/or leukopenia. The following general approach to evaluating the patient is useful but should be individualized.
Pseudothrombocytopenia should always first be excluded by a careful review of the peripheral blood smear. This examination also helps assess for the presence of schistocytes, which are characteristics of TTP and HUS, disorders in which platelet transfusion are usually precluded. For patients receiving unfractionated or low-molecular-weight heparin, the anticoagulant should be stopped until HIT is excluded; the presence of HIT is another contraindication to platelet transfusion. Dilutional thrombocytopenia can be uncovered by a review of the transfusion record. The presence of splenic sequestration may be determined by physical examination or radiographic procedures, such as ultrasound or computed tomography. Finally, to determine whether platelet production is adequate, a bone marrow examination may be performed ( Fig. 7.8 ). Measurement of the serum TPO concentration may be of additional benefit in this evaluation.
As in the assessment of any other medical disorder, careful attention to the history and symptoms, physical examination findings, and appropriate laboratory investigations is essential. Although these suggestions are not meant to be exhaustive, the following approaches are often helpful in evaluating the thrombocytopenic patient.
Given the potential myriad causes of thrombocytopenia, a careful history taking is in order (see Chapters 1 and 2 ). Previous platelet counts are important for documenting the chronicity of the thrombocytopenia and for excluding cases of familial thrombocytopenia and macrothrombocytopenia. Familial thrombocytopenia is not rare and may be a predictor of increased risk for clonal disorder or platelet dysfunction; panels of genetic tests are becoming available to analyze such families. Recent viral infections and vaccinations may cause transient thrombocytopenia. It is mandatory for the clinician to inquire about exposure to new medications, such as antibiotics, heparin, herbal medications, quinine, illicit drugs, and antiplatelet agents. The former (e.g., linezolid, vancomycin, nafcillin) commonly cause thrombocytopenia, and the presence of the last (e.g., aspirin, nonsteroidal antiinflammatory drugs [NSAIDs], ketorolac) may explain ongoing bleeding. Current or recent exposure to unfractionated or low-molecular-weight heparin must be documented. Excessive ethanol ingestion may directly cause thrombocytopenia, or it may occur indirectly through hepatic cirrhosis. In pregnant patients, platelet counts from prior pregnancies may suggest gestational thrombocytopenia or recurrence of ITP; a history of hypertension and proteinuria may indicate HELLP syndrome ( h emolysis, e levated liver enzymes, and l ow p latelet count) (see Chapter 32 ). Recent headache, visual changes, confusion, or personality changes in patients with thrombocytopenia may suggest intracranial hemorrhage. A history of lymphoma or autoimmune disease (SLE, Hashimoto thyroiditis, APLS) may suggest a diagnosis of ITP.
In hospitalized patients, a history of recent RBC or platelet transfusion may be associated with post-transfusion purpura. The response to prior platelet transfusions may also be helpful in assessing whether platelet destruction is ongoing; if such destruction is present, the rise in platelet count will be transient and the corrected platelet count increment will be low. The use of therapies such as continuous venovenous hemofiltration, intraaortic balloon pumps, and extracorporeal membrane oxygenation is commonly associated with thrombocytopenia. The presence of renal failure may also predict an increased risk of hemorrhage due to uremic platelet dysfunction.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here