Isolated thrombocytopenia and thrombocytosis are part of the normal laboratory routine of complete blood count (CBC). However, platelet diagnostics, compared to that of leukocytes and the clarification of anemia, is an area often neglected, with only the platelet concentration being noted. Certainly, and not least, this has to do with the manual effort of further investigation, the limited routine diagnostic means, the complex functional diagnostics and the ignorance of differential diagnoses. Platelets are unjustly reduced to their function in the context of hemostasis, but, in fact, platelets are released in increased numbers in the case of inflammation, are involved in inflammatory reactions via their Fc receptors, and represent one of the first signs of an immune response (acute phase reaction). Thrombopoietin, which is responsible for their concentration, is produced in the liver, which is involved in most inflammatory reactions. This review article is to shine a light on differential diagnoses and the necessary knowledge in the field of cytometric measurement.
The concentration fluctuations of platelets within an individual over time are significantly lower than is suggested by the broad reference range from 150,000 to 350,000 thrombocytes per microliter (Figure 1). Laboratories and textbooks provide various reference ranges; in particular, the upper limit fluctuates between 350,000 and 450,000/μL. The lower limit is mostly stated as 150,000/μL. The reference range study of the German Working Group on Laboratory Hematology on healthy persons according to strict criteria has shown, however, that the reference ranges are narrower than usually assumed, and are even gender-dependent (Figure 2 ).
Hematologic laboratory diagnosis is, of course, a synoptic view of all parameters that allow for a classification of changes that are directly or indirectly related to hematological diseases. The blood count is the most common laboratory test in general, and anemia is the most common pathological finding. It is by far the most common symptom of impaired hematopoiesis, affects the patient’s well-being, and always requires further explanation and investigation. The same is true, however, for the platelets that remain constant within narrow limits. As a rule, an increase or decrease alone does not cause clinical symptoms, apart from petechial hemorrhage in connection with immune thrombocytopenia (ITP) or thrombosis in the case of essential thrombocytosis (ET). Therefore, in most cases, there is the question of what disease triggered the symptom thrombocytopenia or thrombocytosis. It is important to determine whether it is a primary hematologic disease, such as thrombocytosis with myeloproliferative neoplasia (MPN), or a secondary change in the context of another disease (e.g., inflammation, infection…), or a reduction by displacement of malignant processes or immunological mechanisms or just hypersplenism. Analogous considerations also apply, of course, to the changes in leukocytes and erythrocytes.
The laboratory program of hematological diagnostics to clarify thrombocytopenia or thrombocytosis is a stepwise diagnostic process, and it is influenced by supply and demand. The supply is the parameter spectrum that the laboratory physician believes he must make available to his clientele of sample submitters. Demand in this case is what the treating physician expects of her laboratory, whether it be an in-house or a subcontracted laboratory (see Table 1). Figure 3 shows a possible scenario of a complete blood count including a differential blood count for the different medical specialties. All these scenarios are subject to one form or another of stepwise diagnostics. Contrary to what some authors may have suggested, even today, the manual, microscopic blood count still plays a crucial role. One can apply pragmatic aspects to define a so called “diagnostic sieve” to ensure an economical use of laboratory resources and cost-efficient diagnostics. What this means on the cellular side is shown in Figure 4. Hematology devices based on the Coulter principle, in connection with thrombocytopenia, must be controlled by laser-optical devices (even more so, the lower the count), and these, in turn, must be controlled by means of optical fluorescence techniques with nucleic acid dyes, and these yet again by immunophenotypic methods. Aggregate formation should be excluded microscopically as early as possible, because the warnings (“flagging”) are relatively insensitive on most devices, including more recent models. The aggregates are usually placed at the rim and at the end of the feathered margin of the conventional blood smear, which is not examined by automated microscopes (DM systems by CellaVision, Lund, Sweden; HemaCAM by Horn Imaging, Aalen, Germany, HemoFAXS by Tissue issueGnostics GmbH, Vienna, Austria, etc.), in contrast to the capillary smear done on a device just being introduced to the market (Bloodhound or Cobas m511 by Roche Diagnostics, Mannheim, Germany).
Megakaryopoiesis and thrombopoietin
Megakaryopoiesis with formation of platelets is markedly different from erythropoiesis: the megakaryocytes remain stationary in the bone marrow, and their cell processes projecting into the sinusoids of the postcapillary venules continously release platelets. The older and more mature that megakaryocytes become, the larger they get, and the more platelets per megakaryocyte they can release (Figure 5). In immune thrombocytopenia (IgG-mediated Werlhof’s disease), this is particularly visible in a bone marrow aspirate smear: due to the increased demand, the megakaryocytes release platelets early on, when they are still small, while the preformed platelets are significantly more visible within the megakaryocyte (ineffective megakaryopoiesis, see Figure 6). The number of platelets released per megakaryocyte is much lower in this case.
In contrast to other hematopoietic growth factors, thrombopoietin (TPO) is produced in the hepatocytes in a constant amount, regardless of the demand. After release into the blood, the TPO is bound to platelets. In this form, it is inactive. Stimulation of megakaryocytes (MK) occurs only via free, unbound TPO. In case of a thrombocytosis, much TPO is bound to the platelets; the free fraction is small, and the stimulus on the MK low. The consequence: the platelet count drops. Conversely, the proportion of free TPO is increased in connection with thrombocytopenia not caused by antibodies, and the proliferation of megakaryocytes and platelet release are accordingly more pronounced. The consequence: the platelet count increases (Figure 7).
In contrast to a non-immunological thrombocytopenia, in the case of ITP there is too little free thrombopoietin to stimulate thrombopoiesis. At the same time, platelets are increasingly destroyed by antibodies. The outcome is thrombocytopenia (Figure 8). This may also be accompanied by an inhibition of megakaryocytes due to cytotoxic, CD8-positive T-lymphocytes.
The platelets released prematurely in the case of ITP are on average bigger in size and contain significantly more RNA, which is why Ken Ault, in analogy to reticulocytes, called them “reticulated” platelets [2, 3] after Kienast had first demonstrated the relationship between the RNA content and megakaryopoiesis in the bone marrow . In contrast to the reticulocytes of erythropoiesis, a reticular structure is morphologically not verifiable (no reticulated precipitation of RNA caused by dyes). In terms of morphology, platelets, unlike erythrocytes, contain a so-called granulomere, i.e., preformed vesicles and secretory pathways that cause a non-specific staining with the RNA-dyes (Figure 9). This can only be countered by very high-affinity RNA-dyes and partly by degranulation, e.g., using peptide TRAP6. The mean platelet volume (MPV) is an even more unreliable parameter for young platelets than it is for reticulocytes enlarged relative to erythrocytes, as far as pathophysiology AND metrology are concerned (high sensitivity, poor specificity for ITP and/or increased regeneration) . In the Pappenheim stain, the granulomere is stained in a somewhat deeper purple, but the differences are not as clear as in the bone marrow (Figure 4) because of a certain degree of maturation in the blood. The staining duration for the representation of the “reticulated” platelets must be at least 15 min (e.g., 2 h with thiazole orange), which hematology analysers cannot achieve. This is why the devices currently available, despite claims to the contrary, have a poor specificity for “reticulated” platelets even if the addition of RNAse is used as evidence and somewhat reduces the signal. Therefore, the expectations in the follow-up studies to the original publications with manual measurement on the flow cytometer [6–12] have not been met, or only partially, on routine equipment with a reticulocyte channel [13–18], since the Spearman correlation coefficients are only at about 0.6. This clinically valuable parameter of the increased premature platelet release has therefore rarely been incorporated into clinical decisions in the differential diagnosis of thrombocytopenia. Under proper staining and measurement conditions, a good separation between healthy and sick patients then becomes evident (Figure 10). This leaves currently only manual staining and manual measurement on a flow cytometer or on a hematology machine after it has been set up and prepared manually (Figure 11). Manufacturers of hematology instruments do not see a need for action at this time, because the parameter is already officially offered and the validation data have been accepted by the regulatory authorities (FDA).
When a case is referred by a general practitioner, hematologists are often confronted with the question of clarifying thrombocytopenias (without a coagulation analysis), especially when they do not occur in isolation, but are accompanied by anemia or neutropenia. Conversely, hemostaseologists may have to deal with hematological patients where despite all complex and expensive platelet tests having been done, no blood smear has been examined. The essential diagnosic methods for thrombocytopenia remain easy to grasps manageable (see Table 1) but, as far as ITP is concerned, are not without controversy .
The first, reflex-like, question an experienced physician will ask himself or herself during an initial diagnosis of thrombocytopenia is whether the platelet count is at all true, i.e., whether it was measured correctly or corrupted by aggregate formation. Before any, and sometimes, complex differential diagnostic tests of thrombocytopenia, are started, one must first rule out pseudothrombocytopenia, which occurs at a higher frequency than immunothrombocytopenia and is not attributable only to EDTA! The display of the aggregates and/or warnings is more or less unreliable, or absent entirely, depending on the hematology analyzer used. In the blood smear, the large aggregates in the case of EDTA are mostly located at the margin or end of the feathered edge, and can be overlooked during leukocyte differentiation (Figure 12). Clotted blood samples tend to exhibit many small, homogeneously distributed aggregates. The aggregates in the case of essential thrombocythemia (ET), a chronic myeloproliferative neoplasm, often contain very different sized platelets, and elevated levels may be incorrectly reported as normal. The collection tube itself has to be tested for clotting by means of a small wooden stick. Particularly in the venipuncture of infants, the colored caps of the tubes may be mistaken, and the anticoagulant may be missing.
Aggregates in the form of platelet clouds, thus, have many causes: In vitro: clotted blood samples, an EDTA-induced, IgM-mediated autoaggregation, heparin-induced either in the collection tube through its endotoxin contamination or in vivo through effective anti-heparin antibodies from the patient (heparin-induced thrombocytopenia type I and type II after immunization) or cold agglutinins [common in certain infections (e.g., mycoplasma), autoimmune diseases or hepatitis C]. A new hematology unit with image analysis instead of cytometry (Cobas M511, from Roche) is said to be capable of actually counting the platelets individually within aggregates.
A satellite phenomenon is the circular attachment of platelets especially to neutrophils, and less also to monocytes. This phenomenon is found mainly in heparinized blood, whether by administration of heparin in vivo (high dose with acute thrombosis) or in vitro for the purpose of anticoagulation (Figure 12).
In the case of EDTA-associated pseudothrombocytopenia, low-titer IgM antibodies are the cause, which are usually cold antibodies. In vivo, they do NOT cause thrombocytopenia in patients, as opposed to the aggregates occurring in vivo in connection with heparin-induced thrombocytopenia after a therapeutic dose, especially of high-molecular-weight heparin or real cold antibodies against platelets! Pseudothrombocytopenia evolves or strengthens in the time following the blood collection (>2 h). Accordingly, the sample must be reanalyzed after 24 h storage in the refrigerator if there is any suspicion from the history of the patient. In connection with direct measurement after the blood collection, it usually does not occur, leading to discrepant readings between the clinic with its own equipment and the laboratory center. Citrated blood is mostly used a substitute for EDTA and to determine the true value. In about 10% of cases of EDTA-induced pseudothrombocytopenia, however, these antibodies are also active in citrate. This is why a separate collection tube with magnesium sulfate instead of EDTA was developed (Thrombexakt® from Sarstedt, Nürmbrecht, Germany). Such a tube must be used, at least, under such circumstances. However, smears from this new collection system regularly exhibit a more or less strong aggregation of leukocytes (Nebe, unpublished). Capillary blood collection by lancing the fingertip with a direct smear on a slide is an alternative that is often overlooked.
Since the phenomenon of EDTA pseudothrombocytopenia is usually permanent, the patient and his/her treating physicians should be aware of it so as not to draw the wrong conclusions and, for example, postpone or reschedule a necessary operation or read into it illnesses the patient does not have. For this reason, the initial diagnosis should be made clearly on the basis of the above comparison of EDTA with citrate or Thrombexakt®. In doubt, a time kinetic analysis should be prepared to distinguish pseudothrombocytopenia from a clotted blood sample or cold antibodies, active in vivo, against platelets. Also, a separate report should be written to point out this finding.
Even when the counts are in the normal concentration range, false-low values may be measured in connection with essential thrombocythemia due to the inadequate dissolution of platelet clusters after release into the bone marrow. Since the former in vitro artifacts are more common than the pathogenetically relevant causes in vivo, in vivo agglutinates have been given little consideration in practice.
Vice versa, cryoglobulins, which often appear as small spherical precipitates, may simulate platelets (Figure 13). In the smear, they are sometimes difficult to detect without contrast enhancement, and they can also dissolve after exposure to immersion oil. Another interference comes in the form of cytoplasmic fragments that may occur primarily in AML and lymphoblastic lymphomas due to apoptotic pinching of leukemic blasts and that conceal some of the severe thrombocytopenias occurring there.
Since the platelets are defined by their particle size in flow cytometry being less than 20 fL, erythrocyte fragments, too, can masquerade as platelets, as happens in the case of severe iron deficiency or erythrocyte fragmentation syndromes where they also hide severe thrombocytopenieas. The immunological concentration determination using fluorescent monoclonal antibodies is helpful at this point. This error is in particularly clear view in connection with erythrocyte fragmentation syndromes (TTP, HUS, HELLP syndrome), where fragmentocytes and thrombocytopenia occur at the same time. The impedance change in the electrolyte at the passage of a cell through a circular measuring aperture (Coulter principle, named after the inventor and company founder Wallace Coulter) is indeed proportional to the cell volume, but cannot distinguish the cell type. The measurement result of a device using the Coulter principle (still widespread among many manufacturers) should be viewed with caution at values below 30,000/μL and be controlled by other methods. Laser optical light scatter measurements are advantageous and the use of dyes, or better yet fluorescence-labeled monoclonal antibodies against platelet antigens (CD41, CD42, CD61), may yield the correct result (Figure 14), especially in the range of under 20,000 platelets per microliter, a range which is omitted in control reagents and interlaboratory tests, which obscures the problem. Plausibility can also be verified using more simple methods such as the Neubauer counting chamber (hemocytometer) under the microscope or be estimated in a stained blood smear from the ratio of red blood cells to platelets. The precision of both methods does not allow for a reliable differentiation of, e.g., 15,000 and 18,000 PLT/μL. Here, one counts 15 or 18 platelets in the Neubauer chamber, which they do not look as perfect as normal platelets in ITP patients as in healthy people. Despite all efforts to obtain a correct platelet count in a sample, one should not overlook its cause, for example, any blasts of a causal, acute T-cell leukemia, which is why an examination of an MGG-stained blood smear is always much more informative than chamber counting. Also, contamination of equipment or washing solutions as well as electrical interference from the power grid may be mistakenly counted as platelets particularly with the simple Coulter method and conceal the true thrombocytopenia or feign thrombocytosis.
When a blood sample is obtained from infusion systems, especially in ICU patients, further considerations must be borne in mind, which are discussed below.
Once all potential interferences have been excluded, the correctly identified thrombocytopenia can have various causes. Consequently, the correct determination of reticulated platelets is an important decision tool in the differential diagnosis. The formation and maturation of platelets from megakaryocytes is mainly regulated by thrombopoietin, which comes exclusively from the liver. Therefore, distinct liver diseases regularly cause thrombocytopenia if they are accompanied by abnormal synthesis, which is why function parameters such as prothrombin time, cholinesterase, albumin levels… must be taken into account. It should be recalled that the platelet level itself ensures a relatively constant platelet count via free thrombopoietin. Conversely, platelet fragments use their TPO receptors in connection with ITP to make it look as if the level was normal, thus causing an inadequately low supplementary production. As described above, megakaryocytes can, if necessary, release platelets prematurely at any time of their maturation before reaching their maximum size; the earlier, the fewer, i.e., megakaryopoiesis becomes dramatically more inefficient. The microscopic images in the bone marrow aspirate smear demonstrate this impressively in the case of ITP (Figure 6).
Immunologically caused thrombocytopenia
The platelets are involved in the transport of immune complexes through their Fc receptors and can themselves become a target of the immune system (Figure 15). The immunogenic neoantigen may be a combined epitope of an infectious foreign molecule or drug and platelet protein, which causes the thrombocytopenia to disappear after the infection subsides or the drug is discontinued, as is generally the case with acute ITP esp. in children following viral infections or, for example, heparin induced thrombocytopenia (HIT type II). However, if the epitope recognized by the antibody consists of a pure platelet structure, as a “collateral damage” of an immune response, the result is chronic immune thrombocytopenic purpura (ITP, Werlhof’s disease), wherein the platelets are trapped in the spleen or are T-cell-mediated and directed against megakaryocytes, are not formed at all, or undergo apoptosis before they can release platelets (Figure 16). Furthermore, hypersplenism, regardless of the origin, causes thrombocytopenia and sometimes lymphopenia and neutropenia as well. The standard in ITP diagnostics is the MAIPA test, as first described by Prof. Kiefel . However, the platelet antibody determination by immunofluorescence on autologous platelets (PIFT) is quite successful (analogous to the direct Coombs test), if it is performed correctly. In the MAIPA test, only the molecules of the gpIIb/IIIa complex from the platelet lysate from the patient’s EDTA blood are bound to the ELISA microtiter plate via antibodies, and the autoantibodies are detected with an anti-human IgG using the sandwich technique. Therefore, only autoantibodies against the gpIIb/IIIa complex can be found, which is a frequent target in ITP, and only the antibody classes that are recognized by the secondary antibody. A modification of the test, devloped by Dr. Nguyen at the German Red Cross Blood Bank in Mannheim, checks this on microparticles in a flow cytometer (SASPA test), and he added more capture antibodies that also detect anti-HLA antibodies . Otherwise it corresponds well to the MAIPA test . Here it is important to detect and distinguish the IgG, IgM AND IgA antibodies: IgA is not measured under the original approach, but is positive in 10%–20% of ITP cases. IgM indicates the initial stage and, when it occurs alone, as in CLL, one must also consider paraneoplastic AB in the case of IgM persistence. The advantage of selective MAIPA and SASPA tests, with successful dissociation of membrane proteins, is the removal of the interfering Fc receptors (CD32) from the test, which are still present during the incomplete dissolution of the platelet membrane, a critical issueduring sample preparation, and, like with the PIFT, cause a background signal that deteriorates the detection limit. A fixation of the thrombocytes, a step in many PIFT protocols, is also counterproductive, both in direct and indirect assays, because this step also fixes the FcR-bound immunoglobulins and additionally increases the autofluorescence of cells through formalin. The platelets, washed, unfixed and left overnight or longer, release the interfering FcR-bound immunoglobulins again, while the autoantibodies binding via their Fab side remain. This achieves a much better detection limit, which makes the flow cytometric PIFT clinically useful as a screening test, because it compensates for the limited sensitivity of the MAIPA test. In this context, the umbrella term of immune thrombocytopenia (ITP) must be differentiated from the classic Werlhof’s disease, idiopathic thrombocytopenic purpura (also, and previously, abbreviated as ITP), for which the MAIPA or SASPA has a better specificity. In clinical practice (anticoagulation clinic), however, there are many concomitant thrombocytopenias, particularly in connection with hypersplenism, microangiopathy, chronic inflammation, infection or lymphoma (especially Waldenstrom’s disease), which are not to be diagnosed as Werlhof’s disease.
The platelet fragments (also known as “platelet dust” after complement lysis) are still coagulation-active in Werlhof’s disease where a few antibody molecules of high-affinity IgG per cell are sufficient for lysis, which is why these patients often tolerate the low platelet concentrations without bleeding complications. The immune complexes from membrane fragments of platelets and antibodies can still be measured in a well cleaned and alligned flow cytometer (trigger on fluorescently labeled platelet antigens such as CD61), but place high demands on testing (particle-free solutions) and metrology (optics). Remarkably, free antibodies against platelet antigens exist only in the rarest of ITP cases. They would be detectable in an indirect test (similar to the indirect Coombs test for erythrocytes) on test platelets, which is why most immunohematology laboratories have stopped doing indirect tests. This phenomenon has not been explained up to now.
An unpopular subject among ITP experts is the question of the cellular cytotoxic immune response against megakaryopoiesis and its share in the pathogenesis of ITP in the respective patient, since it is not visible in PIFT, MAIPA or SASPA tests and because the opinion today is that one can do the diagnosis of ITP without the antibody diagnostics and bone marrow puncture [19, 23]. That this non-antibody-mediated mechanism exists is quite easy to see in the bone marrow cytology, since the increased megakaryopoiesis is reduced despite a clinical diagnosis of ITP, with PIFT, MAIPA and SASPA being negative, the reticulated platelets not elevated, and withoutapoptotic megakaryocytes being present. The role of T-cells in ITP has been limited to the regulatory T-cells so far [24–28] and one publication has actually placed the antigen-presenting monocytes ahead of the T-cells . Conversely, in a study with a drug that induces LGL, the positive correlation was shown . Many colleagues and thus also the consensus recommendations, therefore, diagnose ITP solely on the clinical picture, given the problematic laboratory diagnostics, i.e., patient history and platelet count, as the above tests are not sufficiently sensitive and specific (PIFT) or are incomplete (MAIPA). The required number of cytotoxic T-cells, like that of the IgG molecules, is low and so is the number of targets. This is why looking for lymphocytic infiltrates morphologically in bone marrow cytology is futile. As is known from in vitro studies, individual cytotoxic T-cells may attack their target cells one by one (as an example of tumor cells has demonstrated), which is why few cells are sufficient. In cases in which ITP is mediated by clonal T-cells, one can detect cytotoxic cells in the bone marrow, however in low numbers. Therefore, it should not surprise that in such cases, which are not rare, the conventional treatments for ITP, which are all based on the autoantibody hypothesis, fail entirely or partially (anti-D, intravenous immunoglobulin, rituximab, splenectomy). Cortisone is used successfully in the T-cells at the beginning, but this is then followed by an increasing resistance of the T-cells (apoptosis resistance to cortisone). The expensive agonists of thrombopoiesis can still achieve a partial platelet increase .
The surface proteins of the platelets are sometimes polymorphic, as far as the HPA system and HLA molecules are concerned. Both can lead to immunization. The former cause alloimmune thrombocytopenia (analogous to alloimmune neutropenia) in newborns by immunization during pregnancy and the latter are responsible for the platelet decrease after transfusion of HLA-unmatched platelet preparations, as they are administered in most cases.
Thrombocytopenia and thrombocytosis in hematological diseases
Thrombocytopenia in chronic lymphocytic leukemia (CLL) is a good textbook example of the different causes of paraneoplastic thrombocytopenia: there may be a displacement of the megakaryopoiesis in the bone marrow or antiplatelet antibodies or hypersplenism as a result of infestation of the spleen by CLL or a combination thereof. This applies of course also to other B-NHL, and in case of immunocytoma (LPL, M. Waldenstroem), thrombocytopenia is almost always present.
The B-NHL often go hand in hand with a reactive proliferation of LGL cells (NK-cells and cytotoxic T-cells), which may be directed against the lymphoma, but may in some cases, due to the cytotoxic mechanism, lead to thrombocytopenia, with or without neutropenia, both of which occur in large granular lymphocyte leukemia (mostly T-LGL) more frequently.
In myeloproliferative diseases, thrombocytosis occurs initially, but over the course it more or less turns into progressive fibrosis: this may be a primary myelofibrosis (according to the new WHO classification, PMF, formerly known as idiopathic myelofibrosis (IMF) or osteomyelofibrosis (OMF)), which often starts with a hypercellular stage, or myelofibrosis secondary to polycythemia vera (PV), essential thrombocythemia (ET), or hairy cell leukemia (HCL).
If a myelodysplastic syndrome (MDS) affects megakaryopoiesis, thrombocytopenia ensues as well due to a disturbance in the maturation. In the case of PMF with myelodysplasia (MDS/MPN), the two effects myeloproliferation and dysmaturity cancel each other out, despite increased, yet ineffective, megakaryopoiesis in the bone marrow. Since dysplasia occurs also in ITP with increased megakaryopoiesis, the purely morphological analysis for the diagnosis of MDS with sole dysplasia of the megakaryopoiesis is a difficult undertaking. Because of their size and fragility megakaryocytes are not yet investigated by flow cytometry for dysplasia and aberrant marker expression.
A special case in the group of MDS is the 5q-syndrome, which is associated with thrombocytosis and actually represents a separate disease, which has now also been reflected in the treatment. Women are more often affected than men, and it is important to watch for the specific morphological signs of dysplasia.
The bone-marrow metastases of solid tumors as well as acute leukemias usually do not cause isolated thrombocytopenia, but anemia as well.
Thrombocytopenia may also occur as a concomitant phenomenon and, manifesting as a platelet decrease, is the earliest and leading sign within hours in the case of disseminated intravascular coagulation (DIC), where each hour of a more early intervention can be life-saving. Therefore, an intelligent measure is to introduce a so-called delta check for platelets for at-risk patients in the laboratory computer system (intensive care, sepsis patients…), which raises the alarm if the platelets drop suddenly, i.e., already at a time when the platelet concentration is still within the normal range and not only when the platelets have dropped below the lower limit of the reference range.
Platelet reductions also occur regularly in connection with burns, sepsis and even with malaria, in each case with a poor prognosis and often associated with DIC.
How common are the above situations as a cause of thrombocytopenia? This really depends on the circumstances: DIC and sepsis are seen more frequent in intensive care; lymphoma more frequently at the hematologist’s clinic; ITP at the anticoagulation clinic; malaria at the tropical disease specialist’s practice, etc. It is crucial that each of these specialists think openly, in all directions, and not focus only on their respective clientele.
Transfusion of platelets
After the foregoing, it is understandable why it is so difficult to establish a fixed limit for platelet transfusion: with ITP without a plasmatic clotting disorder, one is more tolerant, just like with acute ITP in connection with viral infections; even levels under 10,000/μL do not require immediate action, but the indication has to be more generous for patients with a coagulation disorder. Just as important as the plasmatic aspect is the platelet function, whose correct analysis depends on an adequate platelet count, which is precisely not available here. The otherwise-useful test of the PFA-100 function analyzer is of little use with thrombocytopenia, and often results in wrong diagnoses on account of misinterpretations in everyday practice. The longer a doctor knows the patient and/or his/her history, the better an assessment can be given, taking the onus off the above tests. A Willebrand patient with bleeding tendency has a different requirement compared to a tumor patient with a procoagulant situation caused by tumor-associated proteases. An ITP patient whose autoantibody blocks the receptors necessary for platelet function also has a different need than a patient with common ITP.
Within two weeks after the initial heparin dose, antibodies can develop in vivo and lead to thrombocytopenia incl. tissue necrosis at the injection site (heparin-induced thrombocytopenia (HIT type 2). Upon the second treatment, antibody formation and thrombocytopenia occur immediately. Diagnostically, there are a number of approaches for the detection of HIT type II: the HIPA test, serotonin release and PF4-ELISA. The simplest test is a platelet count comparing two samples from a patient +/− heparin preincubation in vitro in the hematology analyzer. The standard of the HIPA test originally described by Prof. Greinacher takes 2 days and some training, while the clinical decision (stopping or modifying heparin therapy) must be made quickly. HIT type 1 happens to appear from the first time on through the above mechanisms, and is less dramatic. Morphologically a pletelet sattelism may be seen in the blood film.
On the other hand, there are in vivo also other antibody-mediated thrombocytopenias apart from the one induced by heparin. We have seen such reaction also after the therapeutic IV administration of monoclonal antibodies against clotting factors or receptors to platelets.
Autoantibodies against platelets or immune complexes on their surface do not lead only or always to thrombocytopenia, either by complement-mediated lysis, phagocytosis or sequestration in the spleen: they can also cause a disturbance of the platelet function by blocking the receptor function, either through direct binding of the (auto-) antibody to the receptor or sterically (immune complexes).
Autoimmune diseases are a complex field (drug-induced, autoimmune disease, antiphospholipid syndrome, SLE, rheumatoid arthritis, Felty’s syndrome, Still’s disease, etc.), and so is hypersplenism incl. infectiological causes where parasitic or intracellular pathogens tend to cause splenomegaly with little or no increase in CRP, which is why the infectious DD is often not considered. An LGL proliferation in the blood or the analysis of resected lymph nodes may provide clues about an infectiological cause, if this is considered and the question is directed to the diagnostician.
Thus, the clarification of thrombocytopenia is a complex area, and patients with thrombocytopenia are seen and treated by hematologists, hemostaseologists and transfusion specialists, because there is not one single and comprehensive representation of all the causes of thrombocytopenia. Figure 17 summarizes once again the foregoing.
Platelet microparticles and exosomes
The measurement of small particles poses a special challenge for flow cytometry including hematology instruments, which is often omitted out of ignorance. A “small particle interest group” has formed as a result that deals intensively with the subject matter. It is crucial that all solutions be free of particles (sheath fluid, washing solutions and antibody reagents). There must also be an inline sterile filter that eliminates dirt particles from the tubings and valves. In addition, the diameter of the laser beam and the noise signal of the laser have to be adjusted. As the particle’s volume decreases, so does the intensity of its fluorescence (to the third power), which means that the light must be collected effectively [numerical aperture of the 90° collecting lens, immersion of the 90° lens (coupled to the cuvette), number of filters in the light path of the fluorescent light, concave mirror in front of the cuvette opposite the 90° collecting lens, introduction of a slot on the image plane to block out ambient fluorescence around the cell in the laser beam (reduction of the background from the “illuminated volume”)]. Less suitable trigger parameters for small particles are forward scattered light (FSC), axial light loss (ALL) and 90° side scattered light (SSC); instead, the fluorescence of nucleic acid dyes and immunofluorescence are better suited, in particular. Here too logical boolean gating and multi-color fluorescence have to be used to ensure that the analysis does capture what is meant to be measured. Especially after erythrocyte lysis, blood contains red cell fragments or ghosts (depending on the type of reagent), nuclear fragments of granulocytes and other interfering factors. The paramedical field of dark field microscopy is all about interpreting the variety of microparticles in the blood, in addition to the actual blood cells.
The lower the concentration of the target population and/or its portion of the measured results, the greater the role of a multiparameter strategy. First, an alternative trigger strategy may be helpful (e.g., CD61 as a measurement threshold), and rising measurement times may be countered via pre-enrichment by means of an immunomagnetic column. Given the demanding nature of the analysis, microparticles will not find their way into clinical diagnostics just yet.
Similar to pseudothrombocytopenia, there is, conversely, a pseudopropagation in the form of cryoglobulins, which can form spherical precipitates (Figure 13). Reactive thrombocytosis can reach up to 1.2 million/μL in intensive care patients, which is why the level of the platelet count is not a reliable criterion for distinguishing between a reactive and a myeloproliferative disease such as ET or PMF. The platelet increase is often the first sign of infection (TPO as acute-phase protein). The anisocytosis and size of platelets is also not a sure sign to distinguish between reactive and autonomous platelet proliferation. On the other hand, in the case of thrombocytosis (!), smaller to medium aggregates in the smear can be an indication of ET, which often already occur in the bone marrow of these patients, where the platelet clusters do not separate anymore after release or discharge from the megakaryocytes. The suspicion is particularly relevant when the platelets in the aggregate are of very different sizes. The relationship between these in vivo aggregates and the risk of thrombosis in not just a few patients with ET is not yet clear. Consequently, the platelet count must be checked, also in the case of thrombocytosis, in the smear at least at the first diagnosis. Fluctuating platelet counts over the course are relatively typical in patients with ET. Otherwise, the linear increase in platelets over longer periods (months) is a good criterion for CMPE, while reactive changes cause an increase and decrease in the platelet concentration, which is otherwise pretty constant within an individual (±5%, see Figure 1).
The degree of thrombocytosis in myeloproliferative disorders is less pronounced in other CMPE like IMF, CML and PV. Fifty percent of ET casestest positive for a mutation of the JAK2 gene (in PV nearly 100%), which is why these, and the BCR-ABL fusion gene, must be examined in such patients via PCR. Here, peripheral blood may be used initially in order to spare the patient a bone marrow biopsy. In the prefibrotic hyperproliferative phase, the primary myelofibrosis (PMF, previously named IMF or OMF) also exhibits thrombocytosis, which usually remains below 750,000/μL. The separation between ET with secondary myelofibrosis and the prefibrotic stage of primary myelofibrosis with its regularly occurring thrombocytosis is a difficult endeavor, and new molecular studies have shown a close relationship between the two diseases .
Thrombocytosis in connection with the 5q syndrome, a subtype of MDS, also is usually less than 750,000/μL, the arbitrary limit introduced by Wintrobe between reactive and autonomous proliferation in ET, which was lowered to 500,000 by Kvasnicka and Thiele .
The refractory anemia with ringed sideroblasts (MDS-RARS), another special form of MDS, may also be associated with thrombocytosis (RARS-T), which is why in addition to the SF3B1 mutation of RARS, the JAK2, calreticulin and MPL mutations must also be tracked down. However, those are not positive in all cases, and iron staining of a marrow particle-rich bone marrow smear remains the desirable option .
A pronounced iron deficiency is usually associated with mild thrombocytosis, which is why iron deficiency aggravates thrombocytosis in connection with inflammation (e.g., Crohn’s disease) or CMPE. Platelet counts in this case are usually below 500,000/μL. Technically, this can be caused or aggravated by the then-pronounced microcytic anisopoikilocytosis with schistocytes. Therefore, in these cases, the separation of platelets and erythrocytes in the histogram or the measuring point clouds must be checked, and any warnings coming from the device about deficient separation must be heeded.
The analysis of platelet function is a domain of hemostaseologists, and there is little standardization in this field, beside the PFA-100 analysis (Siemens), where after thrombocyte stimulation, the time to clog an aperture is measured. Otherwis, each of the experts in the field has developed his or her own method or modification, which is why different devices are available and why there is no consensus . Quite often, the fact that these tests depend on an adequate platelet concentration is ignored and that those results cannot be used in case thrombocytopenia, and therefore are wrong.
The extensive analysis of the neoexpression of activation-dependent molecules on platelets, inter alia, in the context of atherosclerosis research in recent decades, has not been established as routine. This has to do with the shear stress conditions for the platelets during the collection of blood (narrow needle, connecting hose between sampling needle and collection tube), although there are now various, i.e., non-standardized stabilization solutions, some of which have been commercialized and kept as a “secret recipe”. The most sensitive activation parameter is the platelets’ shape change even before any increased or neoexpression of surface antigens appear . Even though the shape change of platelets from discoid to stretched is an early and sensitive marker of activation, it does not produce a good signal differentiation in light scatter analysis (FSC vs. SSC) by flow cytometry, i.e., there is no good method of measurement for it. An interesting alternative is the already mentioned test of cell-cell interactions ex vivo, which are quite stable and reproducible. The adhesion of platelets to monocytes is particularly interesting in this context, and was identified in atherosclerosis research already many years ago as a prethrombotic state by detecting simultaneously monocyte and platelet markers on circulating monocytes in whole blood . The adhesion to cells such as monocytes is thus a stable correlate, of their activation in vivo easily measurable ex vivo, and therefore still a subject of clinical research [38, 39].
The preactivation of platelets during the collection of blood is a delicate issue, of which there are different views among experts. In any case, one thing that they agree on, is, that a large-bore cannula should be used in drawing blood slowly with little shear forces and little contact with a foreign surface (no tubes, i.e., butterfly). A detailed analysis of platelet function is always linked to a particular laboratory and the interpretation of the relevant expert. It always requires a fresh blood sample, which is why the patient has to come to the laboratory, not least because the special anticoagulants are only available there.
Pre-analysis and anticoagulants
A correct platelet count requires strict and fast anticoagulation. Under routine conditions, currently only the calcium chelator EDTA (ethylenediaminetetraacetic acid) is used in everyday practice, although citrate and ACD can be used too. The blood cells react differently to it, and the neutrophils are the most sensitive to it. The EDTA-based aging of the blood begins no later than after 6–8 h, varying from person to person. Platelets, however, are stable for more than 3 days for counting purposes. After that time period interferences esp. from neutrophil degradation start to interfere significantly. The conformation and expression of their membrane-bound adhesion molecules and receptors, however, depend very much on the anticoagulant and transport conditions [40, 41]. The temperature speeds up this process, however the cooling of the sample does not prevent it either. This demonstrates the difficulty of transport, which necessitates the production of an unstained blood smear immediately after the blood is collected and transport media containing fixatives to preserve any expression profile. There are two forms, K2-EDTA and K3-EDTA. The two different salts actually lead to slightly different reference ranges for the blood count, but the changes are not without controversy [42, 43]. One of the reasons is that EDTA requires a certain reaction time, as far as the hydration of the cells is concerned, which is why one should wait for at least 15 min before analyzing a freshly drawn blood sample. For platelet analyses, EDTA plus citrate-theophylline-adenosine-dipyridamole (EDTA-CTAD) has been proposed as a substitute .
When drawing blood via heparinized tubes, an auto-aggregation of platelets and neutrophils or a so-called satellite phenomenon may occur with relative frequency, where the platelets stick to neutrophils in a rosette like formation. The phenomenon is variably pronounced, depending on the donor, and, in addition to the polyanion effect of the high molecular weight polymer heparin and its non-specific binding is caused especially by the endotoxin content of the unpurified heparin sample tubes and activation of neutrophils in particular. Heparin for blood collection systems is obtained as a mass product from the non-sterile intestinal mucous membrane of pigs from slaughterhouses. Heparin blood is often used in leukemia typing as an anticoagulant, since cytogenetic testing with growth in stimulation media prohibits calcium withdrawal, and because the longer durability is beneficial in case of transport. Besides the interfering aggregation, the attachment of platelets to monocytes or myeloid blasts results in a misinterpretation of the expression of megakaryocytic antigens in immunophenotyping, which therefore always requires a visual fluorescence microscopic check. A blood count determination from heparin blood is therefore obsolete; for immunophenotyping, calcium chelators are still required. It is time that purified heparin or analogs like hirudin were used for anticoagulation .
Citrate must be used in liquid form, and the 10% dilution of the blood with a completely filled tube must be taken into account in the determination of concentration (also with respect to recalcification in coagulation tests). Therefore, the level of platelet counts in citrate tubes is usually, wrongly, below the EDTA level in comparative measurements. Citrate improves the durability of neutrophils only to a small extent, but is well suited for platelets. When not centrifuged, it can serve as a substitute in the absence of EDTA blood. However, the centrifugation of that tube is often hard to prevent because of its frequent use for the production of citrated plasma for plasmatic coagulation diagnostics. This presents an automatism in the larger laboratory where the color code directs the work flow.
Platelet function tests and measurement of antibody loadings are best performed from platelet-rich plasma, which is obtained by simple centrifugation from citrated blood at only 100 g.
ACD (acid citrate dextrose) is an additive in stored blood cell concentrates and is suitable for counting in a similar manner to citrate. However, for other platelet tests, there are no reliable validation data.
Collection from infusion systems
The method of choice is the direct, slow drawing of venous blood using steel needles with as wide a lumen as possible, with the patient seated or, sometimes, lying down (effect on the hematocrit). Blood is drawn frequently from each patient in ICU every day. However, when blood is drawn from an infusion tube, this properly requires a preliminary drawing of at least 10 mL of blood, given the complex cabling and the thinning of the blood in the arm vein as a result of the infusion. After a week in intensive care, a quantity of blood is generally collected that corresponds to the amount of one unit of stored blood (concentrated red cells); this is unethical, as are any additional blood samples taken for research purposes without the patient’s consent. That preliminary drawing of blood is often omitted or forgotten, resulting in incorrect readings and misinterpretation like the beginnng of a disseminated intravasal coagulation. It also makes comparisons with previous levels more difficult and distorts study findings. The resulting question of whether thrombocytopenia is derived from the dilution in infusion systems or from infusion therapy can be clarified by another comparison of previous readings, which is usually easily possible with these patients since previous data exist, for example, from their time in the emergency room. In the case of artificial dilution from the infusion system, all cell concentrations are reduced by approximately the same factor, while with hemodilution by infusion therapy, the platelets are very quickly replenished from the bone marrow, in contrast to erythropoiesis, which reacts only slowly.
Mixing of the blood sample and control material
The sedimentation of the blood cells (10–15 mm/h and/or 2 mm in 10 min.) sets in as soon as the collection tube is placed in a sample rack. The mixing can be done manually or by machine. The best method is the overhead mixing of the sample (>10-fold). The control material requires special treatment. It should alternately be rolled manually between one’s palms and mixed overhead, because it needs to be mixed more thoroughly. If an analysis is done from an unmixed sample, the error cannot be reversed: the separation and the erroneous measurement will affect any future use as well. Since a person not properly trained will usually make the same mistake, it will be compounded each time. Another problem is dried blood, especially for systems with screw threads. The dried and disintegrating erythrocytes are counted as platelets. Dried blood and/or residual blood on the thread must therefore be removed with a paper towel. The control material should ideally be placed on the unit like a sample, but the tubes of the control material are often different from those of patient samples, and may not be compatible with the gripper arm or mixing head of the hematology unit. In practice, tumbling mixers are often used. They rotate the tube and simultaneously shake it (partial overhead mixing). There is no literature on the maximum duration of the mechanical mixing as described before cell damage sets in. The consensus is that the sample should be on the mixer for less than 30 min and that the control material should be mixed manually (eminence-based recommendation).
The samples should be analyzed in the laboratory as soon as possible. When shipping samples via a courier service over more than 12 h, precautions will have to be taken. A shipping temperature below +4°C must be avoided at all cost. An unprotected addition of frozen cooling elements in a styrofoam box containing the samples will lead to cell damage. Recorded curves of temperature loggers in the sample chamber have repeatedly shown this phenomenon. A temperature of 10–14°C is recommended. This recommendation stems from an unpublished study by a company on behalf of the British army. Even when laboratory courier services use thick-walled styrofoam containers, samples shipped in a truck overnight in winter will freeze with outside temperatures of below −15°C. Conversely, in summer, passenger cars (such as taxis) can quickly heat up to 50–70°C behind the windshield. Temperature loggers have also shown that airplanes used for shipping samples over greater distances (including within Germany) achieve a temperature in the cargo hold of around 0°C. All this makes it necessary to use temperature-stabilized packaging, which must be checked beforehand. Temperature loggers should also be used, at least on a spot-check basis.
Both in the specification of control materials as well as the regulatory standards of the German Medical Association, low platelet concentrations are a sensitive issue, which has not been solved so far; i.e., lack of controls and regulations in the range below 50,000 platelets per microliter. Internal quality control based on Levey-Jennings charts shows as well that the platelets are a sensitive issue: in the course of a batch (control period), they show a steadily increasing concentration due to the disintegration of erythrocytes and leukocytes, which stands out the most in the low control. We have tested all well-known manufacturers of stabilized control blood, and it appears that the problem of low platelet concentrations in the relevant ranges below 30,000/μL may be addressed only by way of separate control material with artificial microparticles. The situation could be improved only if interlaboratory testing institutions and associations were to draw attention to this issue and exert appropriate pressure on the manufacturers, but they have been reluctant to do so. Accreditation and fear of failing an interlaboratory test and certificate, or switching to a different interlaboratory test provider are counterproductive. In fact, the devices are capable of achieving good results through laser scattered light analyses, nucleic acid staining or immunological methods, while other devices and methods can only yield results like “Concentration <20,000/μL”.
Reference methods for platelet counts
The measurement of low particle and thus also platelet concentrations is by no means trivial: plastic particles, healthy platelets and platelets in ITP or MDS behave differently, like buffer solutions, centrifuged plasma and patient plasma as a matrix. All solutions of the device system, reagents, dilution and washing solutions must be free of particles, and the measuring distances between the suction needle and measuring cuvette must be short and, on the inside, made of finely polished metal. Hose systems give off impurities, while the cleaned hoses lead to the adhesion of platelets to the foreign surface, which is their task in vivo. For example, with ten sequential blank measurements to flush a device prior to the measurement of a sample having low concentrations, there may be at first a drop of counts, which is then followed by an increase of the blank, because the rinse solution removes impurities slowly from the inside of tubes and valves.
In the presence of high concentration of unwanted cells (erythrocytes), coincidences and dead times play a significant role in signal processing. The DIN Committee headed by the Physikalisch-Technische Bundesanstalt (PTB) in Berlin and the International Committee for Standardization in Hematology (ICSH) have published appropriate reference methods based on the immunophenotypic count of platelets [46, 47].
The counting of platelets in a counting chamber under a microscope (named hemocytometer) after erythrocyte lysis and contrast enhancement is suitable to check the plausibility of a measurement result, but is not suitable as a reference method or for follow-ups due to statistical errors esp. in the low count area.
In summary, in the diagnosis of thrombocytopenia, the aggregation tendency of platelets is an interference factor that must always be kept in mind, and the counting of low platelet counts is still technically challenging. The differential diagnoses for thrombocytopenia are manifold, and, as it may be a symptom of a serious disease, it always requires a clarification. In the case of thrombocytosis, there is a broad overlap between reactive and malignant, myeloproliferative causes. The PCR diagnostics of JAK2, calreticulin, MPL and BCR-ABL mutations that occur with myeloproliferative diseases associated with thrombocytosis, depending on the type in varying degrees, has brought some relief here. The combination of laser scattered light and immunocytometry, intelligent signal processing and/or imaging analysis will bring about a better quality, if the compensation systems are adapted accordingly. The cascaded use of different measurement systems in stepwise diagnostics, including manual inspection of the blood smear, has meanwhile become a viable option to intercept erroneous results due to interference.
This article is dedicated to Dr. Silke Heller on the occasion of her retirement from her active offices. Throughout many years of chairing the Laboratory Working Group of DGHO, she managed to bring together hematologists, laboratory physicians, transfusion specialists, pathologists and cytogeneticists in a joint effort to promote and improve the quality of hematological laboratories through professional development, interlaboratory testing and accreditation. On the various national and international boards and committees, she has always provided a competent representation of hematological laboratory diagnostics. For many years, she also acted as the editor in chief of the hematology section of this journal.
Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.
Research funding: None declared.
Employment or leadership: None declared.
Honorarium: None declared.
Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.
Nebe T, Bentzien F, Bruegel M, Fiedler GM, Gutensohn K, Heimpel H, et al. Multicentric determination of reference ranges for automated blood counts. Laboratoriumsmedizin-J Lab Med 2011;35:3–28.Google Scholar
Ault KA, Rinder HM, Mitchell J, Carmody MB, Vary CP, Hillman RS. The significance of platelets with increased RNA content (reticulated platelets). A measure of the rate of thrombopoiesis. Am J Clin Pathol 1992;98:637–46.Google Scholar
Ault KA. Flow cytometric measurement of platelet function and reticulated platelets. Ann NY Acad Sci 1993;677:293–308.Google Scholar
Kienast J, Schmitz G. Flow cytometric analysis of thiazole orange uptake by platelets: a diagnostic aid in the evaluation of thrombocytopenic disorders. Blood 1990;75:116–21.Google Scholar
Diquattro M, Gagliano F, Calabro GM, Tommasi M, Scott CS, Mancuso G, et al. Relationships between platelet counts, platelet volumes and reticulated platelets in patients with ITP: evidence for significant platelet count inaccuracies with conventional instrument methods. Int J Lab Hematol 2009;31:199–206.Google Scholar
Rinder HM, Munz UJ, Ault KA, Bonan JL, Smith BR. Reticulated platelets in the evaluation of thrombopoietic disorders. Arch Pathol Lab Med 1993;117:606–10.Google Scholar
Richards EM, Baglin TP. Quantitation of reticulated platelets: methodology and clinical application. Br J Haematol 1995;91:445–51.Google Scholar
Ogata H. Measurement of reticulated platelets in thrombocytopenia – methodology and clinical utility. Kurume Med J 1998;45:165–70.Google Scholar
Briggs C, Hart D, Kunka S, Oguni S, Machin SJ. Immature platelet fraction measurement: a future guide to platelet transfusion requirement after haematopoietic stem cell transplantation. Transfus Med 2006;16:101–9.Google Scholar
Zucker ML, Murphy CA, Rachel JM, Martinez GA, Abhyankar S, McGuirk JP, et al. Immature platelet fraction as a predictor of platelet recovery following hematopoietic progenitor cell transplantation. Lab Hematol 2006;12:125–30.Google Scholar
Takami A, Shibayama M, Orito M, Omote M, Okumura H, Yamashita T, et al. Immature platelet fraction for prediction of platelet engraftment after allogeneic stem cell transplantation. Bone Marrow Transplant 2007;39:501–7.Google Scholar
Thomas-Kaskel AK, Mattern D, Kohler G, Finke J, Behringer D. Reticulated platelet counts correlate with treatment response in patients with idiopathic thrombocytopenic purpura and help identify the complex causes of thrombocytopenia in patients after allogeneic hematopoietic stem cell transplantation. Cytometry B Clin Cytom 2007;72:241–8.Google Scholar
Briggs C, Kunka S, Hart D, Oguni S, Machin SJ. Assessment of an immature platelet fraction (IPF) in peripheral thrombocytopenia. Br J Haematol 2004;126:93–9.Google Scholar
Abe Y, Wada H, Tomatsu H, Sakaguchi A, Nishioka J, Yabu Y, et al. A simple technique to determine thrombopoiesis level using immature platelet fraction (IPF). Thromb Res 2006;118:463–9.Google Scholar
Hayashi S, Nishiyama M, Suehisa E, Kashiwagi H, Kurata Y, Tomiyama Y. [Comparison between two methods for the measurement reticulated platelet and their clinical significance – flow cytometry (FCM) method and IPF method using automated hematology analyzer (XE-2000)]. Rinsho Byori 2009;57:1039–44.Google Scholar
Koike Y, Miyazaki K, Higashihara M, Kimura E, Jona M, Uchihashi K, et al. Clinical significance of detection of immature platelets: comparison between percentage of reticulated platelets as detected by flow cytometry and immature platelet fraction as detected by automated measurement. Eur J Haematol 2010;84:183–4.Google Scholar
Meintker L, Haimerl M, Ringwald J, Krause SW. Measurement of immature platelets with Abbott CD-Sapphire and Sysmex XE-5000 in haematology and oncology patients. Clin Chem Lab Med 2013;51:2125–31.Google Scholar
Pons I, Monteagudo M, Lucchetti G, Munoz L, Perea G, Colomina I, et al. Correlation between immature platelet fraction and reticulated platelets. Usefulness in the etiology diagnosis of thrombocytopenia. Eur J Haematol 2010;85:158–63.Google Scholar
Provan D, Stasi R, Newland AC, Blanchette VS, Bolton-Maggs P, Bussel JB, et al. International consensus report on the investigation and management of primary immune thrombocytopenia. Blood 2010;115:168–86.Google Scholar
Kiefel V, Santoso S, Weisheit M, Mueller-Eckhardt C. Monoclonal antibody-specific immobilization of platelet antigens (MAIPA): a new tool for the identification of platelet-reactive antibodies. Blood 1987;70:1722–6.Google Scholar
Nguyen XD, Dugrillon A, Beck C, Kerowgan M, Kluter H. A novel method for simultaneous analysis of specific platelet antibodies: SASPA. Br J Haematol 2004;127:552–60.Google Scholar
Nguyen XD, Goebel M, Schober M, Kluter H, Panzer S. The detection of platelet antibodies by simultaneous analysis of specific platelet antibodies and the monoclonal antibody-specific immobilization of platelet antigens: an interlaboratory comparison. Transfusion 2010;50:1429–34.Google Scholar
Sachs UJ. [Diagnosis of idiopathic thrombocytopenic purpura]. Hamostaseologie 2008;28:72–6.Google Scholar
Zhao CH, Hou M. [Advances of research on abnormality of cell immunity in idiopathic thrombocytopenic purpura: review]. Zhongguo Shi Yan Xue Ye Xue Za Zhi 2006;14:1045–8.Google Scholar
Semple JW, Provan D, Garvey MB, Freedman J. Recent progress in understanding the pathogenesis of immune thrombocytopenia. Curr Opin Hematol 2010;17:590–5.Google Scholar
Toltl LJ, Nazi I, Jafari R, Arnold DM. Piecing together the humoral and cellular mechanisms of immune thrombocytopenia. Semin Thromb Hemost 2011;37:631–9.Google Scholar
Johnsen J. Pathogenesis in immune thrombocytopenia: new insights. Hematology Am Soc Hematol Educ Program 2012;2012:306–12.Google Scholar
Semple JW, Provan D. The immunopathogenesis of immune thrombocytopenia: T cells still take center-stage. Curr Opin Hematol 2012;19:357–62.Google Scholar
Zhong H, Bao W, Li X, Miller A, Seery C, Haq N, et al. CD16+ monocytes control T-cell subset development in immune thrombocytopenia. Blood 2012;120:3326–35.Google Scholar
Mustjoki S, Auvinen K, Kreutzman A, Rousselot P, Hernesniemi S, Melo T, et al. Rapid mobilization of cytotoxic lymphocytes induced by dasatinib therapy. Leukemia 2013;27:914–24.Google Scholar
Shih A, Nazi I, Kelton JG, Arnold DM. Novel treatments for immune thrombocytopenia. Presse Med 2014;43:e87–95.Google Scholar
Rumi E, Pietra D, Ferretti V, Klampfl T, Harutyunyan AS, Milosevic JD, et al. JAK2 or CALR mutation status defines subtypes of essential thrombocythemia with substantially different clinical course and outcomes. Blood 2014;123:1544–51.Google Scholar
Thiele J, Kvasnicka HM. [Thrombocytosis versus thrombocythemia – differential diagnosis of elevated platelet count]. Pathologe 2000;21:31–8.Google Scholar
Mohamed M. Refractory anemia with ring sideroblasts associated with thrombocytosis (RARS-T). Blood 2014;123:314.Google Scholar
Norman JE, Westbury SK, Jones ML, Mumford AD. How should we test for nonsevere heritable platelet function disorders? Int J Lab Hematol 2014;36:326–33.Google Scholar
Shah U, Ma AD. Tests of platelet function. Curr Opin Hematol 2007;14:432–7.Google Scholar
Pawelski H, Lang D, Reuter S. Interactions of monocytes and platelets: implication for life. Front Biosci (Schol Ed) 2014;6:75–91.Google Scholar
Fahim A, Crooks MG, Morice AH, Hart SP. Increased platelet binding to circulating monocytes in idiopathic pulmonary fibrosis. Lung 2014;192:277–84.Google Scholar
Rinder HM, Bonan JL, Rinder CS, Ault KA, Smith BR. Dynamics of leukocyte-platelet adhesion in whole blood. Blood 1991;78:1730–7.Google Scholar
Ritchie JL, Alexander HD, Rea IM. Flow cytometry analysis of platelet P-selectin expression in whole blood – methodological considerations. Clin Lab Haematol 2000;22:359–63.Google Scholar
Hagberg IA, Lyberg T. Blood platelet activation evaluated by flow cytometry: optimised methods for clinical studies. Platelets 2000;11:137–50.Google Scholar
Goossens W, Van Duppen V, Verwilghen RL. K2- or K3-EDTA: the anticoagulant of choice in routine haematology? Clin Lab Haematol 1991;13:291–5.Google Scholar
Phillis J, Coiner E, Smith D, Becker J, Leong G. Performance of K2EDTA- vs. K3EDTA-collected blood specimens on various hematology analyzers. Lab Hematol 1998;4:17–20.Google Scholar
Macey M, McCarthy D, Azam U, Milne T, Golledge P, Newland A. Ethylenediaminetetraacetic acid plus citrate-theophylline-adenosine-dipyridamole (EDTA-CTAD): a novel anticoagulant for the flow cytometric assessment of platelet and neutrophil activation ex vivo in whole blood. Cytometry B Clin Cytom 2003;51:30–40.Google Scholar
Menssen HD, Melber K, Brandt N, Thiel E. The use of hirudin as universal anticoagulant in haematology, clinical chemistry and blood grouping. Clin Chem Lab Med 2001;39:1267–77.Google Scholar
DIN 58932-5:2007-10. Hämatologie – Bestimmung der Blutkörperchenkonzentration im Blut – Teil 5: Referenzverfahren zur Bestimmung der Konzentration der Thrombozyten. In. Berlin: Beuth-Verlag.Google Scholar
Harrison P, Ault KA, Chapman S, Charie L, Davis B, Fujimoto K, et al. An interlaboratory study of a candidate reference method for platelet counting. Am J Clin Pathol 2001;115:448–59.Google Scholar
Original German online version at: http://www.degruyter.com/view/j/labm.2014.38.issue-5/labmed-2014-0034/labmed-2014-0034.xml?format=INT. The German article was translated by Compuscript Ltd. and authorized by the authors.