Circadian variation in platelet function in healthy volunteers

MISCELLANEOUS

Circadian Variation in Platelet Function in Healthy Volunteers Syed M. Jafri, MD, Mike VanRollins, PhD, Tsunenori Ozawa, MD, Eberhard F. Mammen, MD, A. David Goldberg, MD, and Sidney Goldstein, MD

Circadian variation in hemostatic factors may contribute to a higher frequency of cardiac events observed in the morning and with activity. Diurnal changes in these factors were investigated by measuring in vitro platelet aggregability in response to epinephrine and adenosfne diphosphate together with @-thromboglobulin and platelet factor 4 as indexes of in vivo platelet activation. Activation of coagulation was measured by thrombinantithrombin Ill complexes and D-Dimers. Tests were performed in 9 normal healthy subjects. Circadian changes occurred in /3-thromboglobulin (p
ircadian variation of factors that increase the risk of thrombus initiation or decreasethe rate of thrombus removal can be an important trigger mechanism for coronary thrombosis. Platelet activation and antifibrinolytic tendency have been shown to follow a diurnal rhythm and possibly relate to the peak time of onset of myocardial infarction and sudden death.lv2 Platelet responsesto activation include adhesion,aggregation and secretion. Platelet aggregability is determined by evaluating in vitro responsesto platelet aggregating agents.3,4Increased plasma levels of platelet release proteins &thromboglobulin and platelet factor 4 reflect platelet release reaction and are considered an index of in vivo platelet activation.5.6This study was performed to investigate the relation of diurnal changes in in vitro platelet aggregation’**to the simultaneous in vivo releaseof platelet releaseproteins. Furthermore, we investigated the diurnal changes in markers of activation of coagulation, thrombin-antithrombin III complexes and D-Dimers.7,s

C

METHODS Study group: The study group comprised 9 healthy subjects(4 men and 5 women, mean age 51 f 10 years) who had no known diseaseand had not received any medication within the preceding 2 weeks before the study. The protocol was approved by the human rights committee, and written informed consent was obtained from all patients. Study pro&d: Patients were admitted to a private room, and venipunctures were performed with the patient in a seatedposition at 3, 6 and 9 P.M. on day 1 of the study. Patients were ambulatory until 10 P.M., and slept at 11 P.M. On day 2, blood sampleswere obtained with the patient resting and in a supine position at 7 and 8 A.M. Between9 and 10 A.M. they walked down the hallways, climbed stairs and useda stationary exercisebicycle, and remained active for the remainder of the day. Blood sampleswere obtained in a seatedposition at 10 A.M., 12 noon and 3 P.M. Blood sample collection: Precautions were taken to minimize the potential for activation of platelets and clotting during blood sampling. A separatevenipuncture was performed at each time point. An initial 3 ml blood sample was discarded. The next 27 ml of blood were collected in a plastic syringe containing 3 ml of 1.24 M From the Heart and Vascular Institute, Henry Ford Hospital, and sodium citrate and 1.36 M glucose (pH 6.5) for platelet Wayne State University Schoolof Medicine, Detroit, Michigan. Manu- aggregation studies. The next 9 ml of blood were colscript received September 16, 1991; revised manuscript received De- lected in a plastic tube containing sodium citrate, cember 6,199 1, and acceptedDecember 7. Address for reprints: Syed M. Jafri, MD, Heart and Vascular theophylline, adenosine and dipyridamole (Diatube-H, Institute K-14, Henry Ford Hospital, 2799 W. Grand Blvd., Detroit, American Bioproducts, Parsippany, New Jersey), and placed on crushed ice. The next 9 ml of blood were colMichigan 48202. CIRCADIAN CHANGES IN PLATELET FUNCTION

951

lected in a plastic tube containing 1 mg of 2.8% sodium citrate solution, and also placed in ice. These blood specimenswere centrifuged at 4’C at 2,000 g for 30 minutes, and plasma sampleswere frozen at -70°C for batch analysis. The upper third of the plasma was separated from the first tube for analysis of platelet factor 4 and @-thromboglobulin. D-Dimer and thrombin-antithrombin III complexeswere assayedusing the titrated plasma. Platelet aggregation sWii: Platelet-rich plasma was prepared by centrifuging the anticoagulated blood at 150 g for 15 minutes. Blood samples were centrifuged at 3,330 g for 10 minutes to obtain plateletpoor plasma. After counting the platelets with the Coulter Counter (Model 2B1, Coulter Electronics, HighCal, Florida), the platelet-rich plasma was diluted with platelet-poor plasma to 200,000 platelets/mm3. Platelet aggregation was monitored photometrically (Monitor IV Aggregometer, Helena Laboratories, Beaumont, Texas)ly2;platelet-rich and platelet-poor plasma absorptions were used to calibrate the 0 and 100%aggregation responses,respectively, of the instrument. To test platelet thresholds, 450 aliquots of platelet-rich plasma were heated at 37’C, and mixed at 1,000 rpm for 2 minutes with adenosine diphosphate (ADP) (Chronolog Corp., Havertown, Pennsylvania) or epinephrine (Parke-Davis, Morris Plains, New Jersey). The minimum concentration neededto produce an irreversible, biphasic increase in light transmission was consideredthe threshold value. To find the threshold, 6 concentrations of epinephrine (0.2 to 50 FM) and ADP (0.5 to 15 PM) were used in order of increasing concentration. The tracings obtained were then analyzed, and the lowest concentration of the aggregating agent to produce biphasic aggregation was evaluated as the index of platelet aggregability. Platelet

factor 4, &thromboglobulin

and D-Dimer:

For assaysof platelet factor 4, fl-thromboglobulin and D-Dimer, enzyme-linked immunosorbent assays were used (Asserachrom, American Bioproducts).4,9 The plasma samples were placed on microwell titer plates that were precoated with specific antibodies of the 3

proteins. An antiplatelet factor, anti+thromboglobulin or anti-D-Dimer peroxidase conjugate was added to form a sandwich. The color change was read on a microplate reader (Bio-Tek Instruments, Model El-305, Vincoski, Vermont) at 492 nm. Calibration curves were then constructed by standard solutions provided with the kit. lhrombin-antithrombin III complexes: Thrombin-antithrombin III complexesin plasmaswere affixed to antibodies to thrombin, which were precoatedon the inner wall of a plastic tube of the enzymgnostkit (Behringwerke Marburg, Germany).7 Peroxidaseconjugated antibodies to antithrombin III were then added, which bind to the available free antithrombin III determinants. After removal of the free conjugate, the bound enzyme activity was measuredby adding hydrogen peroxide and orthophenylenediamine. The color intensity was again read at 492 nm on the microplate reader. A referencecurve was also obtained using a thrombin-antithrombin III complex standard plasma provided with the kit. Statistical analysis: Data in the figures are presented as mean f standard error. To assesswhether overall differences were present, analysis of variance with repeated measureswas performed. Log transformations were performed for some of the variables to adjust for unequal variancesamong the time points. Testing for all variables was performed at the 0.05 level. Pairwise comparisons were performed to compare differences for individual time points when overall significant variation or trends were detectedduring the 24-hour observationperiod. Pairwise testing was performed at the 0.01 level (instead of 0.05) to adjust for multiple comparisons. RESULTS Platelet fun&on studies: Circadian changesin plasma levels of platelet releaseproteins, fl-thromboglobulin and platelet factor 4 are summarized in Figure 1. Betathromboglobulin levels were highest at 3 P.M. on both days 1 (134 f 32 IU/ml) and 2 (147 f 29 IU/ml) of the study. Beta-thromboglobulin levels were lowest at 7

zoo-

150 -

e

E 3 s g m

.

FIGURE 1. Cii tinges in plasma le!vokof~roleaso~,platolof factor 4 (W 4) md /3-tflrombogkbulln (BTG). circa&a dmgos of in vitro platolotaggregcltionweroassodatodwithin vlvo roleaso of plate&f factor 4 ad & thrombogbbu&.

loo-

-

952

THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 69

BTG PI=4

APRIL 1, 1992

A.M. (43 f 10 IU/ml) and 8 P.M. (31 f 3.6 IU/ml) when patients were resting and in a supine position. Platelet factor 4 levels also increased with activity and peaked at 3 P.M. on both days 1 (34 f 10 IU/ml) and 2 (28 f 7.3 IU/ml). Platelet factor 4 levels were again lowest at 7 A.M. (5.0 f 2.0 IU/ml) and 8 A.M. (3.0 f 1.0 IU/ml). Circadian variation was detected for P-thromboglobulin (p <0.05) and reached borderline significance for platelet factor 4 (p cO.06). Furthermore, there were significant differences in @thromboglobulin and platelet factor 4 levels obtained at 7 and 8 A.M. when compared with those at 3 P.M. (p
trends of increasein both in vitro platelet aggregability in response to epinephrine and in vivo release of pthromboglobulin and platelet factor 4 occurring at the same time of the day. Markers of coagulation: An absence of circadian variation of thrombin-antithrombin III complexes (p = 0.36) and D-Dimer (p = 0.44) is displayed in Figure 3. An apparent increase in thrombin-antithrombin III complexesat 10 A.M. was due to abnormally elevated levels in 1 patient. Thus, in contrast to platelet function, circadian changesin activation of the coagulation system were not present in our study group. DISCUSSION The results indicate that platelet activation was lowest when patients were resting in a supine position and that it increasedwith activity. Changesin in vivo release of ,&thromboglobulin and platelet factor 4, were accompanied by similar trends in in vitro responseto platelet aggregating agent epinephrine. Circadian changes in

r”

0

--2 -5

FIGURE 2. Circedim dutnges in tlmdold

-4 3 '5

s J

i: -10

8

- -6

vitropieteM~waaklwesl withpauentslvMingandrrpineat7and 8 A.M., haeased with adivity and peaked et 3 P.M.

-15

- -6 -

EP~ ADP ! -10

-20 15:

16:

21:

07:

10:

06:

12:

15:

Tima of day (hour)

250-

-10

200-

-6

c

E a 150Q

FIGURE 3. Absence ef circedien chenges ecthman,tlBraminmerkersefthmbin binIII complexes (TAT) end ffMdyh,D-Dimers.

2 1

-6

9

z is * loo-

-4

50 -

-2

z

--P-

DDimer TAT 0

0 15:

16:

21:

07:

06:

10:

12:

15:

lima of day (hour)

CIRCADIAN CHANGES IN PLATELET FUNCTION

-I 953

activation of coagulation (as detected by plasma levels plaque. This study confirms circadian variation of plateof thrombin-antithrombin III complexesor D-Dimers) let activation. were not demonstrated. A circadian variation of cardiac ischemic eventshas beendocumentedfor silent myocardial ischemia,10angina pectoris,’ ’ myocardial infarction12 and suddencardi- REFERENCES 1. Tofler GE, BrezenskiDA, Schafer AI, Cziesler CA, Rutherford JD, Willich ac death.13A primary morning peak has been reported SN, Gleason RE, Williams GH, Muller JE. Concurrent morning increase in and a lesserprominent evening peak. Several physiolog- platelet aggregability and the risk of myocardial infarction and suddencardiac N Engl J &fed 1987;316:1514-1518. ic changes occur with activity, which can trigger rup- death. 2. Brezinski DA, Tofler GH, Muller JE, Pahjola-Sintonen S, Wellich SN, ture of an atherosclerotic plaque, and production of cor- Schafer AI, Czeisler CA, Williams GH. Morning increasein platelet aggregabilonary thrombosis. These changes include increases in ity association with assumption of the upright posture. Circulation 1988; serum catecholamines,14heart rate and blood pres- 78:35-40. 3. Born GVR. Aggregation of blood platelets by adenosinediphosphateand its sure.15Clinical evidencethat platelet activation may be reversal. Nature 1962;194:927-929. O’Brien JR. Variability in the human platelets by adrenaline. Nature an important trigger mechanism for coronary throm- 4.1964;202:1188-1190. bosis comes from analysis of the Physician’s Health 5. Kaplan KL. Owen J. Plasmalevelsof beta-thromboglohulinand platelet factor Study.16Patients receiving aspirin had a 59.3% reduc- 4 as indices of platelet activation in viva. B/o& 1981;57:199-202. Takahashi H, Yoshino N, Shibata A. Measurementof platelet factor 4 and tion in the incidence of myocardial infarction during 6. beta-thromboglobulin by an enzyme linked immunosorbentassay. C/in Chim morning hours compared with 34.1% during the re- Acta 1988;175:113-114. 7. Hook JA, Stark A, Cate JWJ, Lamping RJ, BerendsF, BoomJJJ. Laboratory maining hours of the day. clinical evaluation of any assayof thrombin/antithrombin III complexesin Circadian variation of activation of platelet and co- and plasma. C/in Chem 1988;34:3058-3062. agulation has been investigated as a contributing trigger 6. Marder VJ, Budzynski AZ. Degradation products of fibrinogen and cross linked fibrin-projected clinical applications. Thromb Diathi 1974;32:49-56. mechanism. In an earlier study, Tofler et all showed a 9. Whitaker AN, Elvis MJ, Masci PP, BundensenPG, Rylatt DB, Wilber FJ, morning increase in in vitro platelet aggregability. In Burice IH. Measurementof crosslinked fibrin derivativesin plasma:an immunotheir subsequentstudy, platelet aggregability was shown assay using monoclonal antibodies.J C/in Pathol 1984;37:882-887. 10. Rocco MB, Barry J, CampbellS. Circadian variation of transient myocardial to relate not particularly to the time of the day, but ischemia in patientswith coronary artery disease.Circulation 1987;75:3955400. to assumption of an upright posture and to activity.2 11. Muller JE, StonePH. Turi ZG, Rutherford JD, CzeislerCA, Parker C, Poole Results from our study confirm these latter observa- WK, PossamaniE, Roberts R, RobertsonT, Sobel BE, Willerson JT, Braunwald MILIS Study Group. Circadian variation in the frequencyof acutemyocardial tions, as platelet activation was related to activity with E, infarction. N Engl J Med 1985;313:1315-1322. changes noted to increase in the morning hours after 12. NademaneeK, Intrachot V, JosephsonMA, Singh BN. Circadian variation occurrenceof transient overt and silent myocardial ischemiain chronic stable standing, and peaking at 3 P.M. However, other studies inangina and comparisonwith prinzmetal angina in men. Am J Cardiol 1987;60: described in vitro and in vivo platelet aggregability to 494-498. peak in the morning hours.17,r8A consistent finding in 13. Muller JE, Lindmer PL, Wellich SN. Tofler GH, Ayllmar G, Klangos 1, PH. Circadian variation in the frequencyof suddencardiac death. Circulaall studies was that platelet activation is lowest when Stone tion 1987;75:131-138. patients were resting. Differences occurred in the detec- 14. Turton MG, DeeganT. Circadian variations of plasmacatecholamine,cortition of peak platelet activation. In our study, a delayed sol and immunoreactiveinsulin concentrationsin supinesubjects.C/in Chim Acta peak may representthe gradual increaseof platelet acti- 1974;55:389-397, 15. Millar-Craig MW, Bishop CN, Raftery EB. Circadian variation of blood vation over the course of the day with continued activi- pressure.Lmcet1978;1:795-797. 16. Ridker PM, Manson JE, Buring JE, Muller JE, HennekensCH. Circadian ty. Alternatively, a bimodal peak in platelet activation variation of acute myocardial infarction and the effect of low-dose aspirin in a could also exist as reported for cardiac events with randomized trial of physicians.Circulation 1990;82:897-902. 17. Jovicic A, Mandic S. Circadian variation of platelet aggregability and lihrinomorning and afternoon peaks in events.19 lytic activity in healthy subjects. Thromb Res 1991;62:65-74. The absence of fibrinolytic tendency in the early 16. Musumeci V, Rosa S, Caruso A, Zuppi C, Zappacosta B, Tutinelli F. morning hours is consideredto be another potential con- Abnormal diurnal changesin in-viva platelet activation in patients with atherotributing factor for the reported higher morning inci- sclerotic diseases.Atherosclerosis 1986;60:231-236. 19. Hjalmarson A, Gilbin EA, Nicod P, Dittoich H, Henning H, Engler R, denceof thrombotic cardiovascular events,as measured Blacky AR, Smith SC Jr, Ricou F, Ross J Jr. Differing circadian patterns of by tissue plasminogen activator and its inhibitor after a symptomsonsetin subgroupsof patientswith acute myocardial infarction. Circulation 1989;80:267-275. circadian rhythm. 20-22We did not detect thrombin acti- 20. Andreatti F, Davies GH, Hackett DR, Khan MI, DeBart ACW, Aber VR, vation as fibrinolysis after a diurnal rhythm. It is possi- Maseri A, Kluft C. Major circadian fluctuations in librinolytic factors and possible that the ability to lyse a thrombus is reduced in the ble relevanceto time of onsetof myocardial infarction, suddencardiac death and stroke. Am J Cardiol 1988;62:635-637. morning hours, but thrombin activation or librinolysis 21. Fearnley GR, Balmfooth G, Fearnley E. Evidenceof a diurnal librinolytic does not have diurnal variability in healthy subjects. rhythm: with a simple method of measuring natural librinolysis. C/in Sci In conclusion, investigations to elucidate triggers 1957;16:645-650. 22. Angleton P, Chandler WL, Gottfreid S. Diurnal variation of tissue type for coronary thrombosis are important to understand plasminogen activator and its rapid inhibitor (PAI-I). Circulation 198979: mechanismsprecipitating rupture of an atherosclerotic lOlllO6.

954

THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 69

APRIL 1, 1992