FINAL DIAGNOSIS Prominent platelet satellitism causing pseudothrombocytopenia.
The phenomenon of platelet satellitism was initially described by Field and MacLeod in 1963 in blood taken from a 14 year old boy being worked up for a neurologic disorder 1. Since it was initially described there have been multiple case reports of both healthy patients and patients presenting with various medical conditions (e.g. coronary bypass graft surgery, chronic liver disease, neurologic disease, mantle cell lymphoma, chronic lymphocytic leukemia etc.) and, to date, no particular disease association has been made. The only commonality that has been drawn is that platelet satellitism is only observed in blood which is collected with the anticoagulant ethylenediaminetetraacetic acid (EDTA), a calcium ion chealator, and is therefore thought to be an in vitro phenomenon.1,2,3,4,5,6,7,8,9
In a 1974 study, this phenomenon was observed in two patients during the same week: one admitted for a leg fracture and one for treatment of alcohol withdrawal. To better understand this occurrence, the blood from these two patients was studied and it was determined that platelet satellitism was independent of the age of the blood (fresh vs. "several-hours-old" blood), temperature (4-37oC), and staining (unstained vs. stained with Wright-Giemsa). Platelet satellitism was observed in all areas of the slide and not limited to the feathered edge; however, the phenomenon ceased to be observed when blood was collected in double oxalate, sodium citrate or heparin. Interestingly, platelet satellitism was also not observed when the patients' plasma, platelets and white blood cells were mixed with platelets and white blood cells from normal patients.6
This particular study further investigated the association by performing electron microscopy. The images taken demonstrated that the interaction was predominately driven by either a pseudopod extending from the neutrophil to the platelet or a short process extending from the platelet to the neutrophil. In such associations, an electron-lucent area was observed between the two apposing membranes. Neutrophil phagocytosis of the platelets was observed as well. In a 3 week post-discharge follow-up of both patients in that study platelet satellitism was no longer observed.6
This electron-lucent area between platelets and neutrophils was not widely agreed upon. Subsequent, similar studies utilizing electron microscopy instead described this space as containing a low electron density material that appears proteinaceous and, in some instances, appeared to strand. This observation led to the hypothesis of a "bridging substance". Also observed in a subsequent electron microscopy study was the finding that participating platelets tended to have higher glycogen content and were called "stress platelets." In this study, stress platelets were also seen in a control group of patients treated with steroids. This led to the hypothesis that stress platelets have a higher propensity to rosette around neutrophils.1
A study of fourteen patients with platelet satellitism in 1995 demonstrated that the anti-neutrophil activity of the plasma was completely abolished when the plasma was adsorbed with normal platelets before mixing with leukocytes and platelets, suggesting a single antibody mediated the process. Inhibition studies demonstrated that this antibody was an IgG autoantibody directed against the platelet glycoprotein IIb/IIIa complex and Fc gamma receptor III (Fc gamma RIII) on the neutrophils. This was confirmed by mixing the sera from the patients with platelet satellitism with the platelets of patients with Glanzmann's Thrombocythemia (congenital lack of platelet glycoprotein IIb/IIIa) and congenital Fc gamma RIII absence on neutrophils (NAnull Phenotype) which abolished the phenomenon. This antibody creates the bridge between the platelets and the neutrophils and fit with the "bridging substance" hypothesis previously described.2 Why neutrophils appear to be solely involved has been linked to the differences in expression of Fc gamma RIII. Neutrophils expression appears to be limited to the phosphatidyl inositol-linked form while natural-killer cells and macrophages express a transmembrane form.9
Another cause of pseudothrombocytopenia is platelet clumping. Similar to platelet satellitism, platelet clumping is considered an in vitro phenomenon also observed only in the presence of EDTA. The mechanism is thought to be related to the EDTA calcium chelation causing a conformation change in the platelet membrane glycoproteins and subsequent binding of IgG, IgM, or IgA antibodies. The conformational change in the membrane is thought to expose cryptantigens which may not be normally exposed on platelets and some antibodies with GP IIb specificity have been described.3,8 It is as yet unclear why some patients develop antibodies only to platelets, causing platelet clumping, and some patients develop antibodies with specificity for platelets and neutrophils, causing platelet satellitism. The incidence of pseudothrombocytopenia has been reported as 1.9% for inpatients and 0.15% for outpatients 10. Platelet clumping is a more common cause of pseudothrombocytopenia, with platelet satellitism having an incidence reported at approximately 1:30,000. Reportedly, these two phenomena have never been observed together.9
In recent years, another cause of pseudothrombocytopenia has been identified and linked to the use of the monoclonal antibody abciximab. This drug, which is used to prevent thromboembolic events during percutaneous coronary intervention, has specificity for the B3 subunit of GP IIb/IIIa. The pseudothrombocytopenia incidence related to abciximab has been reported between 2-27% (depending on the criteria) and is, again, only observed in blood collected with the anticoagulant EDTA. The mechanism of action is thought to be the same, with the calcium ion change causing a conformational change in the membrane and either act as an agglutinating antibody itself or aid in exposing antigens which promote agglutination.9
Platelet satellitism, while rare, represents an important cause of pseudothrombocytopenia. Without recognition of this entity patients may receive unnecessary platelet transfusions.1 In the case of abciximab related pseudothrombocytopenia, the artificially decreased platelet count could trigger the drug to be stopped or a necessary procedure to be cancelled.9 However, the effect is in vitro and only observed in the presence of EDTA. Patients do not have an increased propensity to bleed and although one study has suggested these patients have a higher propensity for thrombosis and development of auto-immune conditions; this link has not been clearly established and platelet function is maintained.1, 9 As such, in order to get an accurate platelet count, it is recommended that blood instead by drawn in sodium citrate when platelet satellitism is observed.
Contributed by Brian K. Theisen, MD and Raymond E. Felgar MD, PhD