Pseudothrombocytopenia Secondary to the Effects of EDTA in a Dog

Introduction

Pseudothrombocytopenia (PTCP) is an in vitro occurrence resulting in platelet clumping and a low, invalid platelet concentration reported from automated hematology analyzers and also manual platelet counts. It can occur with traumatic venipuncture and platelet activation, leading to in vitro platelet clumping. Additionally, feline platelet concentrations are often falsely decreased because of platelet clumping when automated hematology analyzers are used. Pseudothrombocytopenia can also be induced by some anticoagulants. In most cases, anticoagulant-induced PTCP is caused by ethylenediaminetetraacetic acid (EDTA), often used for complete blood counts (CBCs).^2–4 Pseudothrombocytopenia is less commonly seen in other anticoagulants, such as citrate or oxalate.^1,2 In human medicine, PTCP occurs in approximately one out of 1000 people, irrespective of the presence or absence of disease.^2,5 Pseudothrombocytopenia secondary to the effects of EDTAhas also been described in horses and in one case of a miniature pig.^6,7 This case demonstrates that EDTA-dependent PTCP can also occur, even if very rarely, in dogs.

Case Report

A 6-year-old, intact male Labrador retriever was examined at Washington State University’s teaching hospital with a history of hypothyroidism and chronic dermatitis. Physical examination revealed a bright and alert dog with a dull hair coat and mild ventral neck alopecia. No other abnormalities were observed on physical examination.

A CBC revealed a leukocytosis (18,500/μL; reference interval 5800 to 11,700/μL) with a neutrophilia (12,950/μL; reference interval 3000 to 7100/μL), monocytosis (2035/μL; reference interval <800/μL), and eosinophilia (925/μL; reference interval <100/μL). The packed cell volume (PCV) was 43% and within the reference interval (38% to 59%). A thrombocytopenia (20,000/μL; reference interval 157,000 to 394,000/μL) was also noted, with platelet aggregates observed on a Giemsa-Wright-stained blood smear. The venipuncture was nontraumatic, and large platelets were not present to suggest a regenerative response in the platelet line. The dog also did not exhibit any bleeding tendencies, and a repeat platelet concentration was measured using EDTA blood 1 week later. At this time, the platelet concentration remained decreased (24,000/μL), with marked platelet clumping observed on the blood smear. The presence of marked platelet aggregation on the blood smear, a clean venipuncture to help minimize platelet activation, and lack of clinical signs compatible with a true thrombocytopenia suggested the reported thrombocytopenia in this dog was due to EDTA-dependent PTCP or platelet hyperreactivity causing in vitro platelet aggregation.

Diagnosis of EDTA-dependent PTCP requires documentation of normal platelet concentrations in a different anticoagulant or rapid evaluation of blood without an anti-coagulant using blood from the same blood draw, as well as repeatability of the PTCP in EDTA.¹ To determine if the thrombocytopenia in this dog was secondary to EDTA, platelet concentrations were compared in whole blood anticoagulated with EDTA and in whole blood anticoagulated with 3.8% sodium citrate.

Venipuncture was performed in the clinical pathology laboratory to minimize time between the blood draw and the first platelet concentration measurement. One single, non-traumatic venipuncture was performed with a 19-gauge butterfly needle using a jugular vein. Blood (5 mL) was drawn and discarded to eliminate platelets that may have been activated during venipuncture. Next, 5 mL of blood was drawn into two separate syringes—one containing 15% K₃-EDTA (0.01 mL EDTA per 1 mL of blood) and the other containing 3.8% sodium citrate (0.1 mL sodium citrate per 1 mL of blood). The platelet concentrations were measured within 1 minute of blood draw (time 0) and then at 30 minutes, 60 minutes, and 120 minutes after the blood draw. Separate samples of EDTA anticoagulated blood were incubated at 37°C starting at 20 minutes after the blood draw, and platelet numbers were evaluated at 30 minutes (blood warmed for 10 minutes) and at 60 minutes (blood warmed for 40 minutes) to help determine if temperature contributed to the low platelet concentrations. Mean platelet volume (MPV) was also documented.

Platelet concentrations were evaluated on all samples using an automated hematology analyzer.^a Blood smears were also evaluated at time 0 to evaluate for platelet aggregates. Platelet numbers in EDTA blood were low and variable [see Figure] from time 0 to 120 minutes, ranging from 12,000 to 81,000/μL (mean 43,750/μL). The MPV was also elevated in the EDTAblood (mean 11/μL; reference interval 6.5 to 9.9 μL).³ The EDTA platelet concentrations 30 minutes after blood collection at room temperature and warmed to 37°C were comparable with counts of 12,000/μL and 13,000/μL, respectively. After 60 minutes, the room-temperature EDTA sample contained 41,000/μL and the warmed sample contained 10,000/μL. The platelet concentrations from the sodium citrate anticoagulated blood were within the reference interval and ranged from 322,000 to 394,000/μL (mean 341,500/μL; reference interval adjusted 10% for dilution of platelets in sodium citrate was 172,700 to 433,400/μL). Smears of EDTA blood at time 0 contained marked platelet clumping, while blood smears from sodium citrate whole blood contained only rare, small platelet aggregates.

Discussion

The automated platelet concentrations in this dog were clearly decreased in the EDTA blood compared to the sodium citrate samples. A clean venipuncture was performed, and 5 mL of whole blood was initially discarded to eliminate the possibility of platelet activation secondary to venipuncture. The PTCP in this dog was compatible with a diagnosis of EDTA-dependent PTCP. The MPV was increased in the samples with the lowest platelet counts at time 0, 30 minutes, and 60 minutes; theMPV decreased into the reference interval at 120 minutes and 4 hours. While the MPV can increase in EDTA after storage,⁸ the MPV decreased in this dog with time, indicating the increased MPV was not due to storage artifact. The platelet count increased as the MPV decreased, suggesting that small platelet clumps may have been detected by the analyzer at time 0, 30 minutes, and 60 minutes, which may have contributed to the mildly increased MPV. The platelet concentrations were still decreased when the anticoagulated whole blood was warmed to 37°C, suggesting the EDTA-dependent PTCP was not temperature dependent in this dog. However, the platelets were not warmed immediately after the blood draw, and it is possible that platelet clumping at cooler temperatures would not be completely reversed with warming to 37°C.

In humans, EDTA-induced thrombocytopenia is attributed to immunologically mediated platelet clumping; however, the specific mechanism has not been determined.^1,5,9 One human study demonstrated EDTA-dependent antibody reactivity with all platelets except those in patients having Glanzmann’s disease; such individuals lack the platelet GPIIb/IIIa receptor. This finding suggested the GPIIb/IIIa receptor played an important role in this phenomenon.⁹ An additional study found that platelet aggregation in human patients with EDTA-dependent PTCP was prevented by blocking the von Willebrand factor and fibrinogen-binding sites on GPIIb/IIIa; these findings further supported this receptor’s involvement.¹

Alterations in calcium have also been implicated in EDTA-induced PTCP in humans. Calcium is chelated by EDTA, which can result in conformational changes in the membrane, including the GPIIb/IIIa receptor, and the exposure of surface antigens that are usually hidden.^10,11 These exposed antigens, neoantigens, can be targets for circulating antibodies.^5,10 Antiplatelet antibodies—including immunoglobulin G, immunoglobulin M, and occasionally immunoglobulin A— have been documented in human patients with EDTA-dependent PTCP.^5,12,13 These antibodies are hypothesized to represent autoantibodies that could recognize neoantigens, although their exact origin has not been determined.⁵ Citrate also has a chelating effect on calcium, but to a lower degree.⁵ Other anticoagulants, such as heparin, are reported to cause PTCP.¹³ Heparin does not chelate calcium, suggesting that calcium chelation is not a sole contributor to anticoagulant-induced platelet aggregation.¹³ Platelets from human patients with EDTA-dependent PTCP also have upregulation of activation markers including P-selectin (CD62P), suggesting that the binding of antibody to the GPIIb/IIIa receptor causes platelet activation in these patients. The cause of EDTA-dependent PTCP has not been elucidated in veterinary medicine, and further investigation is necessary to determine an immunological role.

In humans, EDTA-dependent PTCP can occur suddenly. While an underlying cause may be found in some patients, this phenomenon has not been correlated to gender, age, a particular disease, or drug administration.⁵ Once it is documented in humans, EDTA-dependent PTCP also appears to be a persistent finding.⁵ This dog was lost to follow-up, and subsequent platelet concentrations could not be evaluated to determine the persistence of the EDTA-dependent PTCP.

Conclusion

Pseudothrombocytopenia can be considered clinically significant if it is not differentiated from a diagnosis of true thrombocytopenia, because it may result in unnecessary diagnostics and medical therapy; however, PTCP is not a pathological process by itself. This phenomenon can often be prevented by decreasing platelet activation during sample acquisition through nontraumatic venipuncture and transferring the sample quickly into EDTA anticoagulant. Properly drawn samples in EDTA rarely have clumped platelets. Traumatic venipuncture should be avoided, as it is a common cause of platelet activation, platelet clumping, and PTCP; also, slow collection into syringes should be avoided.

In cases of thrombocytopenia, scanning a prepared blood smear for platelet clumps is always recommended. Some analyzers also report the number of platelet clumps detected. A repeat, nontraumatic venipuncture may be enough to obtain an accurate platelet count for the dog. However, rare cases in the dog may require blood drawn into a different anticoagulant, such as sodium citrate, to help discriminate a true thrombocytopenia from an EDTA-dependent PTCP. Platelet clumping in sodium citrate can still occur, but it is generally reduced. In conclusion, EDTA-dependent PTCP is recognized in horses and should also be considered in other veterinary species, including dogs, that have a significant thrombocytopenia and no clinical bleeding tendency.