Influence of oral contraceptives on coagulation tests in native blood and plasma

Friedebert KuDZ, MD: Christoph Pechlaner, MD: Marco Tabarelli, MD,b Elisabeth Solder, MD,b and Wolf-Dieter Zwierzina, MD:

Innsbruck, Austria

Routine coagulation laboratory tests, clotting times in native (not anticoagulated) whole blood, platelet-rich and platelet-poor plasma, and recalcification times in citrated whole blood, platelet-rich and platelet-poor plasma were performed in 14 healthy premenopausal women. Blood was taken before and after one or two cycles of low-dose oral contraceptives. After oral contraceptives a reduction in clotting time in native platelet-rich plasma and activated partial thromboplastin time were observed. Recalcification times in whole blood and platelet-rich plasma were shorter than clotting times in their native counterparts. The observed changes are compatible with a procoagulant effect seen soon after the start of oral contraceptive use. The absence of these changes in the recalcification times in citrate systems suggests a masking effect of citrate. The reduction in clotting times in native platelet-rich but not in platelet-poor plasma indicates that the hypercoagulability in oral contraceptives users is mainly related to platelets. (AM J OBSTET GVNECOL

1990;163:417-20.)

Key words: Oral contraceptives, coagulation, native blood, activated partial thromboplastin time

There is a vast body of literature on changes in blood coagulation in users of oral contraceptives (OCs) and the induction of a hypercoagulable state, which is de­fined as procoagulatory changes in some coagulation tests (accelerated clotting or changes in clotting factors or inhibitors).l-s Much of the published data is conflict­ing, which partly reflects technical differences in the assay systems, as well as problems in an accurate di­agnosis of deep vein thrombosis.6

. 7 The insufficient re­liability of epidemiologic data reporting an association between OC use and deep vein thrombosis has been discussed extensively.s

There is also no consensus on the relevance of the reported hemostaseologic findings, for example, a de­crease in antithrombin III or a rise in fibrinogen con­centration.9 On the other hand, divergent effects may compensate each other, as may be the case with re­ported procoagulatory and pro fibrinolytic effects.1O

To date there is no coagulation laboratory test avail­able for assessment of individual thromboembolic risk.

In our laboratory we developed clotting tests in native (i.e., not anticoagulated) blood and plasma, which re­vealed hypercoagulability in coronary heart disease and diabetes mellitus (unpublished data) not disclosed by the conventional citrate test systems. This prompted us to examine, with these test systems, the effects of low-

From the Departments of Internal Medicine" and Gynecology/Ob­stetrics/ University of Innsbruck. Reprint requests: Friedebert Kunz, MD, Gerinnungslabor, Univ. Kli­nih fur Innere Medizin, Anichstr. 35, A-6020 Innsbruck, Austria. 6/0/19595

dose OC use on blood coagulation, with test persons serving as their own control.

Material and methods

Fourteen healthy women taking no other medication (particularly no medication known to influence coag­ulation or platelet function for at least 14 days) partic­ipated after they gave informed consent. All used low­dose OC formulations (containing ethinyl estradiol, 0.03 mg in 13 patients and 0.04 mg in 1 patient). They were examined in the second half of the menstrual cycle before and between the eighteenth and twenty-first day of the first (nine women) or second (five women) cycle ofOC use.

All clotting time measurements were performed in duplicate on a hook coagulometer (Amelung, Lemgo, West Germany) at 370 C. Blood was obtained by atrau­matic puncture of an antecubital vein through a 1.2-gauge needle. In citrated samples 9 volumes of blood were mixed with 1 volume of 0.11 moUL sodium citrate (Sarstedt tubes, Vienna, Austria).

Immediately after blood collection, 0.3 ml of native whole blood was transferred to the coagulometer, and clotting time was recorded.

The preparation of native platelet-rich and native platelet-poor plasma was carried out at 40 to 100 C by differential centrifugation (5 minutes at 400 g for native platelet-rich plasma; 10 minutes at 2000 g for native platelet-poor plasma). Cooling is essential to prevent coagulation. Contact with glass surfaces was avoided. Native platelet-rich plasma was adjusted to a platelet

417

418 Kunz et al.

Clolling lime of NPRP in DC users

1000

p ( 0.05 800

u CLJ

600 tn

400 t....J ::t:: -f- 200

before after

Fig. 1. Clotting times in native platelet-rich plasma (NPRP) before and after OC use. Bars denote the means of 13 values. (Data are missing in 1 of 14 cases.)

concentration of 250 GIL by addition of autologous native platelet-poor plasma. Clotting times were re­corded in 0.3 ml samples.

In the citrated systems, recalcification times in 0.2 ml samples were recorded after the addition of 0.2 ml of 0.025 mollL CaCI2 •

A platelet count was performed on a Coulter counter in whole blood, platelet-poor plasma, and before and after adjustment of the platelet concentration in platelet-rich plasma (both in native and citrated samples).

Routine coagulation tests were performed on cit­rated blood or plasma, according to the instructions of the manufacturers: activated partial thromboplastin time (cephaloplastin Dade, Merz + Dade, Diidingen, Switzerland), partial thromboplastin (Boehringer, Mannheim, West Germany), fibrinogen, II antithrombin III (Boehringer), and thrombelastography12 (Hellige thrombelastograph D, Freiburg, West Germany). A blood count was carried out on a Coulter counter.

The Wilcoxon matched-pairs signed-ranks test was used to test for statistically significant differences; cor­relations were tested according to standard nonpara­metric procedures. BMDP statistical package subpro­grams 3D and 8D, respectively, (BMDP statistical soft­ware, Inc., Los Angeles, Calif.) were applied.

Results

A significant reduction in clotting times was observed in native platelet-rich plasma (Fig. 1) and activated par­tial thromboplastin time (Fig. 2) after OC use but not in the other test systems (Table I). Neither the increase in fibrinogen concentration nor the decrease in anti­thrombin III was statistically significant (Table I).

u CLJ U)

aPTT in DC users

35

July 1990 Am J Obstet Gynecol

30 ~~ p ( 0 .01

25 ~

20

15

10

before after

Fig. 2. Activated partial thromboplastin time (aPTT) (in ci­trated plasma) before and after OC use. Bars denote the means of 14 values. Some points and lines are overlapping.

Clotting times in native whole blood and platelet-rich plasma were significantly longer than in their citrate counterparts before OCs (P < 0.01; Table I); after OCs only the whole blood methods differed significantly (P < 0.05; Table I).

Platelet counts in native and citrated platelet-rich plasma before adjustment to 250 GIL did not differ significantly (339 ± 112 vs 348 ± 82 GIL).

Coagulation test changes did not differ between first and second cycle. No correlation was observed between changes in clotting times of native platelet-rich plasma and activated partial thromboplastin time.

Comment

There is still ongoing debate on whether OC use induces the development of hypercoagulability or in­creases the risk of deep venous thrombosis, especially if the newer low-dose formulations are used. The es­tablishment of a correlation between OC use and clin­ically manifest thromboembolic disease is confounded by a rather high rate of false-positive and false-negative errors in clinical diagnosis. However, clinical experi­ence suggests an influence of OC use, because in young women deep vein thrombosis appears to occur almost exclusively in OC users or in very obese persons. In an ongoing study at our clinic, 24 of 37 premenopausal women with deep vein thrombosis received OCs (prev­alence of OC use in our area is 25%); of the remaining 13 women, 8 were obese.

The most commonly reported blood coagulation alterations in OC users are a decrease in antithrombin III and an increase in platelet aggregability and fi­brinogen, all compatible with a shift to hypercoagula-

Volume 163 Number I, Part 2

Native blood clotting tests and OCs 419

Table I. Mean ± SD of the tested variables before and after OC use

Before After Unit of

Clotting times Mean ± SD Mean ± SD measure p value

Native Whole blood 409.0 ± 155.4 364.9 ± 128.4 sec Platelet-rich plasma 521.9 ± 227.9 378.0 ± 145.7 sec <0.05 Platelet-poor plasma lO52.5 ± 734.3 974.2 ± 689.1 sec

Citrate Whole blood 264.4 ± 50.3 279.8 ± 114.3 sec Platelet-rich plasma 385.3 ± 152.9 346.6 ± 149.2 sec Platelet-poor plasma 715.4 ± 579.3 779.0 ± 639.6 sec

aPTT 30.2 ± 1.7 28.0 ± 1.7 sec <0.01 Other coagulation tests

Platelets 237.5 ± 34.6 223.5 ± 43.7 gm/L Prothrombin time 91.6 ± 21.5 92.9 ± 9.0 % Fibrinogen 249.3 ± 56.2 278.8 ± 63.1 mg/dl Antithrombin III lO6.6 ± 16.4 96.9 ± 11.4 % Thrombelastography r* 8.1 ± 2.9 7.9 ± 1.9 mill Thrombelastography k* 5.8 ± 2.1 6.0 ± 2.1 min Thrombelastography ma* 50.1 ± 7.0 51.2 ± 4.6 mm

*The thrombelastography parameters are1': r, reaction time; k, thrombus formation time; ma, maximal amplitude.

bility.'-5. 8 However, not all investigations confirmed these results13 ; in our study the lack of significant dif­ferences might be because of the small number of pa­tients. On the other hand, several papers9

. 14 report an increase in fibrinolytic activity, which may compensate for procoagulant changes.

The significant shortening of clotting time in native platelet-rich plasma but not in the other native test systems observed in our study suggests a role of plate­lets in mediating the effects of OCs. An increase in platelet activation has been described in other reports, however, not unanimously.'

The additional influence of erythrocytes 15 is probably the main reason why these changes between before and after OC use were not seen in whole blood. Another reason might be activation of platelets andlor coagu­lation by cooling; this, however, appears to be of minor relevance, because in uncooled citrate the clotting time of platelet-rich plasma was shorter than in cooled native platelet-rich plasma.

The lack of differences in recalcification times in the citrate systems before and after OC use points to a masking effect of citrate, probably because of a platelet­aggregating effect. This is in agreement with results of another study15 in native blood in which a sex difference in platelet aggregation with stronger platelet aggre­gation responses in women, not detectable in citrate, was observed. The differences between native and ci­trate systems may be caused by altered platelet function as a consequence of reduction as well as Ca'+ availability of increased pH in citrate. 16. 17 Since the platelet con­centration before adjustment to 250 GIL was the same in native and citrate platelet-rich plasma, a different platelet distribution can be ruled out as the reason for the differences between native and citrate systems.

As illustrated in Figs. 1 and 2, the decrease in clotting time was not uniformly observed in all subjects, which indicates differing individual reactions. Moreover, changes in the same persons were not unidirectional in native platelet-rich plasma and activated partial throm­boplastin time, which is not surprising, since partly dif­ferent phases of coagulation were tested.

The observed reduction of activated partial throm­boplastin time has also been reported by some other groups," 14, 18 which suggests changes in intrinsic plasma coagulation factors in OC users, for example, increased factor XII concentration.9

One of the several reviews8 on the subject came to the conclusion that "at this time, it is not known whether oral contraceptives do or do not trigger the hemostasis system to develop a thrombosis or by which mechanism they could trigger the event." Our data speak in favor of an effect of OC use (including the low-dose for­mulations) on blood coagulation, with a shift toward faster clotting demonstrable after one to two pill cycles. This seems to be mediated at least in part by increased reactivity of blood platelets, which is detectable in native but not in citrate blood.

The impact of the type of sex steroids used and other cofactors (e.g., smoking) needs further study. Whether the detected changes will still be observed after a longer period of OC use still needs to be determined.

We are indebted to Mrs. Sedighe Ahrabian for ex­cellent technical assistance and to Dr. Arnulf Lochs for expert help on statistical and computational problems.

REFERENCES 1. Shapiro S. Oral contraceptives-a time to take stock.

N Engl] Med 1986;315:450-1. 2. Dugdale M, Masi AT. Hormonal contraception and

Kunz et al.

thromboembolic disease: effects of the oral contraceptives on hemostatic mechanisms. J Chron Dis 1971;23:775-90.

3. Beller FK, Ebert C. Effects of oral contraceptives on blood coagulation. A review. Obstet Gynecol SUTY 1985;40:425-36.

4. Notelovitz M. Oral contraception and coagulation. Clin Obstet Gynaecol 1985;28:73-83.

5. Bonnar J. Coagulation effects of oral contraception. AM J OBSTET GYNECOL 1987;157:1042-8.

6. Goldzieher JW. Hormonal contraception: benefits versus risk. AMJ OBSTET GYNECOL 1987;157:1023-8.

7. Drill VA. Oral contraceptives and thromboembolic dis­ease. JAM A 1972;219:583-92.

8. Mammen EF. Oral contraceptives and blood coagulation: a critical review. AMJ OBSTET GYNECOL 1982;142:781-90.

9. Stadel BV. Oral contraceptives and cardiovascular disease. N EnglJ Med 1981;305:612-8.

10. Gevers Leuven JA, Kluft C, Bertina RM, Hessel LW. Ef­fects of two low-dose oral contraceptives on circulating components of the coagulation and fibrinolytic systems. J Lab Clin Med 1987;109:631-6.

11. Clauss A. Gerinnungsphysiologische Schnellmethode zur Bestimmung des Fibrinogens. Acta Haematol (Basel) 1957;17:237-46.

July 1990 Am J Obstet Gynecol

12. Hartert H. Blutgerinnungsstudien mit der Thrombelas­tographie, einem neuen Verfahren. Klin Wochenschr 1948;26:577-83.

13. Farag AM, Bottoms SF, Mammen EF, Hosni MA, Ali AA, Moghissi KS. Oral contraceptives and the hemostatic sys­tem. Obstet Gynecol 1988;71:584-8.

14. Ghezzo F, Mele A, Pegoraro L. Effects of oestro­progestogenic oral contraceptives on blood coagula­tion, fibrinolysis, and platelet aggregation. Haematologica 1980;65:1-9.

15. Zwierzina WD, Kunz F, Kogelnig R, Herold M. Sex­related differences in platelet aggregation in native whole blood. Thromb Res 1987;48:161-71.

16. Rogers AB. The effect of pH on human platelet aggre­gation induced by epinephrine and ADP. Proc Soc Exp Bioi Med 1972;139:1100-3.

17. Zwierzina WD, Kunz F. A method of testing platelet ag­gregation in native whole blood. Thromb Res 1985;38:91-100.

18. Notelovitz M, Kitchens CS, Coone L, McKenzie L, Carter R. Low-dose oral contraceptive usage and coagulation. AM J OBSTET GYNECOL 1981;141:71-5.

Hemostasis profile in women taking low-dose oral contraceptives

J. L. David, MD,. U. J. Gaspard, MD, PhD,b D. Gillain, MSC,b R. Raskinet,. and M. R. Lepot, MDb

Liege, Belgium

Thirty-six young, healthy, nonsmoking women have been selected to check the effect of low-dose oral

contraceptives on hemostasis. Two identical groups were treated by Marvelon (a monophasic oral contraceptive containing ethinyl estradiol and desogestrel) or Trigynon (a triphasic oral contraceptive containing ethilyl estradiol and levonorgestrel) for a 6-month period. In the absence, previously controlled,

of substantial differences between the effects of each treatment on hemostasis, all the results were pooled at the third and sixth month of the study. The effects of oral contraceptive treatment were as follows: (1) platelet number, platelet aggregating ratio, and plasma ~-thromboglobulin level were not significantly altered, and (2) antithrombin III activity was not reduced despite a slight decrease or antigen concentration. The von Willebrand factor parameters, factor VIII: C, factor VII: C, and clottable fibrinogen

were significantly increased. Plasminogen (activity and antigen concentrates) and (X2-antiplasmin levels

were also significantly increased. Activated partial thromboplastin time and euglobulin lysis time measured

after venous occlusion were significantly Shortened. Although statistical analysis did not show dramatic

changes in all these parameters, some individual extreme values were substantially altered. Therefore we

believe that these later values are worthy of cautious consideration for weighing the role that hemostasis

factors might play in individual thrombotic risk. (AM J OBSTET GVNECOL 1990;163:420-3.)

Key words: Oral contraception, platelets, von Willebrand factor, coagulation, fibrinolysis

Despite an overall reduction of thrombotic events in users of low-dose oral contraceptives, recent epidemi-

From the Department of Medicine" and Department of Obstetrics and Gynaecolog;y,h State University of Liege, Centre Hospitalier Uni­versitaire du Sart-Tilman (B 35). Reprint requests: J. L. David, MD, Thrombosis and Haemostasis Unit, Centre Hospitalier Universitaire Liege (4000, Sart-Tilman), Belgium. 6/0117881

420

ologic data are consistent with an increased risk of deep venous thrombosis and pulmonary embolism in treated women. 1 The thrombotic risk linked to oral contracep­tive use is probably multifactorial, involving diverse al­terations in hemostasis. 2 It is generally accepted that these later alterations are induced by the estrogen com­ponent of oral contraceptives and are dose related.3• 4

The estrogenic effect may be under a modulating in­fluence of progestogens.