Platelet Parameters and Aggregation in Essential ...

Platelet Parameters and Aggregation in Essential and Reactive Thrombocytosis EPHRAIM SEHAYEK, M.D., NURIT BEN-YOSEF, M.Sc, MICHAELA MODAN, M.Sc, ANGELA CHETRIT, B.A., AND DINA MEYTES, M.D.

ESSENTIAL THROMBOCYTOSIS (ET) is a well-defined entity.16 However, in some patients extreme reactive thrombocytosis (RT) in which the cause is not readily apparent can initially mimic ET. Additional structural and functional platelet characteristics may be of help in differentiating between the two entities. The platelets in myeloproliferative disorders were found to differ from normal in morphologic characteristics,17'20 biochemistry,8 and physical characteristics.4 Defective hemostatic function of platelets in these disorders has been reported by many authors. Functional disorders, including reduced adhesiveness,6 prolonged bleeding time,1 and defective aggregation in response to adenosine diphosphate (ADP), epinephrine (EPN), and collagen (COL), were reported.51013 Received November 30, 1987; received revised manuscript and accepted for publication April 11, 1988. Address reprint requests to Dr. Meytes: Head, Hematology Department, The Edith Wolfson Hospital, Tel-Giborim, Holon 58100, Israel.

Departments of Hematology, The Edith Wolfson Hospital, and Clinical Epidemiology, Chaim Sheba Medical Center, Tel Aviv University Sackler Faculty of Medicine, Israel

Automated blood counting devices added novel information concerning size distribution of blood particles, including platelets. In thrombocytopenia the mean platelet volume (MPV) has been found to be of clinical importance.7 Whether platelet size indices are of clinical relevance in thrombocytosis has not yet been established. Three recent publications deal with the characteristics of platelet populations in states with thrombocytosis. Giles9 reported that platelet count and MPV were inversely related; Holme and associates12 found smaller MPV values in reactive thrombocytosis and similar ones in thrombocytosis related to myeloproliferative disorders as compared with controls. Small and colleagues19 suggested that the megathrombocyte index in association with MPV and the dispersion of platelet volume distribution are of clinical use in distinguishing between ET and RT. In the present study, platelet size indices were compared in patients with the two forms of thrombocytosis as well as normal controls, and their relationship to in vitro platelet aggregation was established. Materials and Methods Patients All routine automated blood counts of adult inpatients that were performed at the Edith Wolfson Hospital during a period of five months were reviewed. One hundred three consecutive patients with platelet counts over 500 X 109/L (500 X 103/ML) were identified. These patients were interviewed and examined so that the cause of thrombocytosis could be assessed. In addition to pertinent clinical and x-ray investigation, the following relevant blood chemistry was also tested: blood fibrinogen, prothrombin, and partial thromboplastin time; fibrinogen splitting products (FSP); serum iron; iron-binding capacity; and uric acid. Of the 103 patients, 2 were diagnosed as having ET and 89 as having RT; in 12 patients a definite diagnosis of ET or RT could not be

431

Downloaded from http://ajcp.oxfordjournals.org/ by guest on April 10, 2016

Platelet characteristics were assessed in 15 patients with essential thrombocytosis (ET), 89 patients with reactive thrombocytosis (RT), and 23 normal controls. A platelet volume distribution width (PDW) greater than or equal to 10.5 was found in 50%, 21%, and 14% of the three groups, respectively (P = 0.01 between patients with ET and patients with RT; P = 0.02 between patients with RT and controls), reflecting an excess of extreme values at both ends of the distribution. Compared with controls, the increase in platelet number in patients with RT was about twofold throughout the platelet volume range, whereas ET was characterized by afivefoldincrease in small platelets less than 7.5 fL and threefold increase in larger size platelets. Mean platelet volume (MPV) was significantly lower in patients with ET versus patients with RT and in patients with RT versus controls (mean ± SD 7.5 ±1.2 vs. 8.8 ± 0.1 and 10.2 ± 1.8 fL, respectively, P < 0.01). Rate of in vitro platelet aggregation greater than or equal to 50% was significantly lower in patients with ET versus patients with RT and in patients with RT versus controls (0%, 23%, and 45%, respectively, P < 0.01). Aggregation rate was positively correlated with MPV (r = 0.54; P < 0.0001). Aggregation rate in patients with ET was significantly lower (P = 0.01) than expected from their reduced MPV alone. Despite these group differences, the overlap of individual platelet characteristics between the three groups precludes their usefulness for diagnostic purposes. (Key words: Essential thrombocytosis; Reactive thrombocytosis; Platelet volume; Platelet aggregation) Am J Clin Pathol 1988;90:431-436

SEHAYEK ET AL.

432

A.J.C.P. • October 1988

FIG. 1. Platelet volume (fL) distribution curve in a normal control.

7.5-10,4 10,5- 13,4 > U.5 4.5-7,4 PLATELET VOLUME GROUPS fL

made. These latter patients did not have any obvious known cause for thrombocytosis, but because the platelet count was not sufficiently elevated to permit the diagnosis of ET according to the criteria of the Polycythemia Vera Study Group (PVSG),16 they were excluded from analysis. An additional 13 patients with ET under follow-up by the hematology clinic were also analyzed. The study group thus included three diagnostic categories: (1) fifteen patients with ET, who met the criteria of PVSG and who, at the time of the inclusion in the study, were not receiving or recovering from a recent course of chemotherapy (in patients with known, previously treated ET, the study was done at a time when an increasing platelet count was observed before readministration of the required therapy); (2) eighty-nine patients with RT, the causes being recent major surgery (47 patients), malignancy (29 patients), and inflammation (13 patients); (3) a control group of 23 healthy persons with a normal platelet count from the hospital staff. Laboratory Studies For each participant, the platelet count and all platelet characteristics were determined twice on two consecutive days. All analysis was based on the means of these two determinations. Blood was drawn into ethyl diamine tetra-acetic acid (EDTA)-anticoagulated plastic tubes. Specimens were routinely counted within two hours of drawing by the Coulter Counter Model S-Plus® (Coulter Electronics Inc., Hialeah, FL). A complete blood count, including total platelet count, MPV, and platelet distribution width (PDW), was obtained. The latter value, which is automatically displayed by the Coulter Counter, is an

expression of the spread of the platelet size distribution, based on the estimated 20th and 80th size percentiles, computed as follows: PDW =

80th percentile - 20th percentile XC 80th percentile + 20th percentile

C = calibration factor In addition, a detailed individual platelet volume relative distribution histogram and minimal and maximal values of platelet volume were obtained from the counter (Fig. 1). In this histogram the relative proportion of counts perceived by each of 64 channels representing volume ranges from 2 to 20 fL is plotted. To estimate the actual number of platelets of different sizes, each histogram was divided by hand into five volume categories (Fig. 1). As the histogram is received from the counter plotted on a grid, the number of area units within each volume category was counted. The platelet number in each platelet volume category was determined by multiplying the percentage of the histogram area within that category by the total platelet count. Although the counter omits platelets less than or equal to 2 fL and greater than or equal to 20 fL from the total count as well as from the platelet size distribution curve, the effect of this omission is negligible, because their number in all likelihood is relatively small. This assumption is based on the fact that all curves were monomodal, with the mode invariably between 3-8 fL, the relative frequency approaching zero rate toward the extreme platelet volume ranges. Platelet aggregation was assessed after it was ascertained that no medication known to affect it had been taken by the participant within the preceding two weeks. Tests were performed by use of the light transmission


1-HA

Vol. 90 • No. 4

PLATELET PARAMETERS IN THROMBOCYTOSIS

Statistical Analysis Differences in MPV, PDW, and platelet volume distribution between patients with ET, patients with RT, and controls were analyzed by one-way analysis of variance with internal comparisons. Differences in the distribution of aggregation rates in the three aggregation categories were analyzed by Fisher's exact test for 2 X 3 tables. Association of MPV with total platelet count, as well as of summed aggregation with MPV and number of platelets less than 7.5 fL, was analyzed by testing the significance of the difference of Pearson's correlation coefficient (r) from zero. Results The MPV by total platelet count in the three groups combined is presented in Figure 2, demonstrating a highly significant inverse correlation between the two parameters (r = -0.504; P < 0.0001). MPV values were lowest in patients with ET (mean ± SD 7.5 ± 1.2 fL), intermediate in patients with RT (8.8 ± 1.1), and highest in controls (10.2 ± 1.8) (patients with ET vs. patients with RT and controls, P < 0.001; patients with RT vs. controls, P < 0.01). The prevalence of MPV values less than 7.5 fL in patients with ET, patients with RT, and

•X.«.. ..*...*-*....*.

:ooo

•

.*....«....»....«....«....«....«....«....*.

[

(

I t

t too.

• I

I

I I I

too.

I

i

l

I

« l I I n l I It I II It u t l I MM M i l IM ItMt Ml I I • t t t i m t i

•

I

• I I M •

• C

45. to. B • 127 r. '-.sojf mi 91-10

79. '0.

c c C CC C C c

19. 10.

ct t C C It c c

tJ. 00.

1

C

109 110

1

119 110

\a 120

mpi

FIG. 2. Relationship between MPV (fL) and platelet count (in 109/L [103/jiL]) in controls (C), patients with RT (R), and patients with ET (E).

controls was 66.7%, 10.1%, and 4.3%, respectively. The low MPV in patients with ET as compared with those with RT could not be explained entirely by the higher counts in the former group. When MPV values in patients of both groups were analyzed within corresponding ranges of total platelet counts (Table 1), it was found that the mean MPV of patients with ET in each count range was significantly lower than in those with RT. PDW values were highest in patients with ET (mean ± SD 10.5 ± 0.6), intermediate in patients with RT (10.3 ± 0.4), and lowest in controls (10.1 ± 0.3) (patients with ET vs. patients with RT and controls, P < 0.01; patients with RT vs. controls, P = 0.02). PDW greater than or equal to 10.5 was found in 50%, 21%, and 14% of ET, RT, and control groups, respectively. This greater width resulted from excess at both ends of the distribution, reflected in the platelet volume distribution curves of the three groups presented in Figure 3. All curves were monomodal. The curve for patients with RT was of a similar shape to that of controls: there was a proportional addition of platelets (2.2- to 2.8-fold compared with controls) in each platelet size category. In contrast, with patients with ET the curve was shifted to the left because of a much more prominent addition of small


aggregometer (Chrono-Log Corporation, Havertown, PA) according to Miale.18 Platelet-rich plasma (PRP) was prepared from citrated venous blood in a 1:10 ratio by centrifugation (Beckman, Model T-JG®, Palo Alto, CA) at 120 X g at room temperature for 5 minutes. Platelet-poor plasma (PPP) was prepared from the PRP by spinning at 2,750 X g for 10 minutes. PPP was used to adjust the final platelet count to 250 X 109/L (250 X 103/ML). All tests were performed within one hour of blood drawing with the use of 0.4-mL aliquots of the adjusted PRP. After preliminary stirring for 1 minute, the aggregating reagents were added in the desired concentration in a volume of 0.1 mL. The following reagents were used: ADP, 4.6 and 1.15 /*mol/L final concentration; COL, 0.4 g/L final concentration; and EPN, 2.7 and 0.67 jtmol/L (TEVA Pharmaceutics, Jerusalem, Israel). The change in light transmission was recorded at a chart speed of 50 mm/minute by use of the B-500 chart recorder (Houston Instruments, Houston, TX). Aggregation rate was classified into three categories according to the percentage change in light transmission: no response (no detectable change in light transmission); response less than 50% (increase in light transmission by less than 50%); and response greater than or equal to 50% (increase of 50% or more). Summed aggregation rate was the sum of percentage changes in light transmission incurred by the maximal concentrations of ADP and EPN and by COL. Aggregation time was defined as the time needed to reach maximal aggregation.

433

Table 1. Mean MPV Values (fL) in Patients with ET and Patients with RT within Corresponding Ranges of Platelet Counts in 109/L (10 3 /ML) (patient numbers in parentheses)

Table 2. Distribution (%) of Response to Platelet Aggregation Reagents in Patients with ET, Patients with RT, and Controls*

Platelet Count Ranges Group

600-699

700-799

>900

Patients with ET (X ± SD) Patients with RT (X ± SD) P

7.4 ± 1.5 (2) 8.5 ± 1.0 (26)