Influence of Anticoagulation on Blood Loss Following Dental Extractions D. DECLERCK, F. VINCKIER, and J. VERMYLEN' Catholic University of Leuven, School for Dentistry, Oral Pathology and Maxillofacial Surgery, Capucijnenvoer 7, B-3000 Leuven, Belgium; and 'Centre for Thrombosis and Vascular Research, Herestraat 49, B-3000 Leuven, Belgium The coagulation activity level at which oral surgical procedures can be performed in anticoagulated patients without triggering bleeding complications and without enhancing the risk of developing thrombo-embolic events remains controversial. The objective of the present study was to evaluate blood loss following dental extractions at different levels ofanticoagulation and to determine its effect on wound closure rates. Blood loss was measured following the removal of four front teeth in warfarinized rabbits. Immediate blood loss was evaluated by determining the tooth socket bleeding times and by using a technique based on hemoglobin determinations. Long-term blood loss was assessed by comparison of labeled red-blood-cell disappearance curves. The results showed that blood loss following dental extractions was significantly greater in animals anticoagulated at a therapeutic level than in non-anticoagulated control animals. Determination of blood loss at different levels of anticoagulation clearly demonstrated that complete correction of the coagulation activity was unnecessary. Partial correction (INR values of 1.6-1.8) allowed extractions to be performed without extensive blood loss. With this technique of partial correction, the period of interruption of the anticoagulation could be kept very short, and the risk of postoperative bleeding complications was minimal. Wound closure rates were negatively influenced in anticoagulated animals. J Dent Res 71(2):387-390, February, 1992

Introduction. The management oforal surgery patients who are taking coumarin derivatives remains controversial. The clinician must balance the risk of reducing or stopping the anticoagulant therapy with that of triggering post-operative bleeding complications. Three different treatment philosophies can be distinguished. A first group of authors recommends leaving the anticoagulant dose unaltered. They state that dental extractions are possible at therapeutic anticoagulation levels (Behrman and Wright, 1961; Bailey and Fordyce, 1983). According to these authors, the small but potentially hazardous effect of discontinuing anticoagulant therapy is not justified in patients anticoagulated within the therapeutic range. Other studies have reported on tooth extractions in patients taking coumarin derivatives without the occurrence of severe bleeding episodes (Waldrep and McKelvey, 1968; Askey and Cherry, 1956). Leaving the anticoagulant dose unaltered is a simple technique, and development of thrombo-embolic events is prevented. However, the fear of possible hemorrhagic complications increases stress in the cardiac patient, and this induces the release ofsubstances that promote fibrinolytic activity (Bump and Kolodny, 1973). Local hemostasis-promoting materials and techniques are widely used in patients having dental extractions with unaltered anticoagulation. However, side-effects have been reported: delay in Received for publication March 5, 1991 Accepted for publication September 10, 1991 Based on a thesis submitted in fulfillment of the requirements for the Doctor in Dental Sciences degree, Catholic University of Leuven The investigation was supported by a research fellowship from the Belgian National Fund for Scientific Research.

wound closure and healing, increase of local inflammatory response, and induction of foreign-body reactions (Hunt and Benoit, 1976; Benoit and Hunt, 1976;Vinckier and Vermylen, 1984; Ibarrola et al., 1985). A second group of authors feels that complete correction of the coagulation activity is necessary (Ziffer et al., 1957; Dugdale, 1983) and is possible without too much risk of thrombo-embolic events developing. Complete correction of the coagulation activity has the advantage that hemostasis is perfectly normal at the moment ofthe intervention, and blood loss will be minimal. The main disadvantages are the possible confusion for patients when changes in their medication are necessary, the necessity of repetitive blood sampling to follow the coagulation activity level after withdrawal of the medication and after re-institution of therapy, the unavoidable postponement of the intervention until the coagulation activity is completely corrected, and the risk of a thrombo-embolic event developing in the period of interruption of the anticoagulation. A third philosophy is characterized by partial correction of the coagulation activity. The level should be sufficient for achieving immediate hemostasis following the intervention and, at the same time, limit the risk for thrombo-embolic events. According to Spouge (1964), an optimal balance is obtained when the prothrombin time is 50% above normal. Johnson and Leary (1988) adjusted the anticoagulant dose to allow for a prothrombin time 1.5 times that of the control. Mulligan and Weitzel (1988) considered the range ofsafety to lie between 1.5 and 2.0 times the control prothrombin time. Partial correction of the coagulation activity allows normal hemostasis to occur with minimal blood loss, and considerably reduces the period of interruption of the anticoagulating medication. Comparison ofthe different studies reporting on dental surgery in anticoagulated patients is extremely difficult. Most studies are clinical trials on rather small numbers of subjects. The basis for selection of cases is poorly defined, and, in most studies, little attention is paid to factors such as gingival health, blood pressure, multi- or single-rooted teeth, level ofdifficulty ofthe extraction, type of anesthesia (general or local, with or without addition of vasoconstrictor), material and technique used for suturing, local wound treatment (socket packing), antibiotic coverage, and intake of analgesics. In most studies, evaluation of the amount of blood lost is based on the patient's impression and is not objectively measured. Another important difficulty is the poor comparability of the levels of anticoagulation among the different studies. This is a consequence of differences in sensitivity of the test reagents used in the laboratories and different methods for reporting of the results (time, ratio, index, etc.). This problem is now solved by the introduction of the International Normalized Ratio (INR) (Van de Besselaer et al., 1984). Manufacturers determine, for each batch of test reagent, the ISI (International Sensitivity Index) by comparison with a reference thromboplastin. This enables laboratories all over the world to convert their prothrombin time results into INR values. An INR of 1.0 corresponds with a normal hemostatic capacity. Therapeutic levels of anticoagulation are reached with INR values between 2.0 and 3.5. When the INR is 5 or higher, there is a serious risk of spontaneous bleeding episodes. The aim of the present study was to develop a technique for evaluation of blood loss following dental extractions and to measure blood loss at several levels of anticoagulation. The level of coagulation activity at which blood loss was not greater than in non-

Downloaded from jdr.sagepub.com at UNIV OF PITTSBURGH on December 18, 2014 For personal use only. No other uses without permission.

387

388

DECLERCKet al.

J Dent Res

February 1992

Groups: 1

-_

0

2 3

on

_._ 4 5 _ 6

N

0 0

of

I

V]

0

0

*la 0

:so

0 4J

ot /

odo

0:

~

VA A

\\

*V\

"\__

-44 I-

0

-'\\

\1

X..~ \

10

20

30

40

.

50

60

Time after extraction (min)

Fig. 1-Tooth-socket bleeding times in control animals (group 1) and in anticoagulated animals with INR 2.0-3.0 (group 2), INR 1.8-2.0 (group 3), INR 1.6-1.8 (group 4), INR 1.4-1.6 (group 5), and INR 1.3-1.4 (group 6).

1

3

4

6

8

11

Time after extraction (days)

Fig. 2-Disappearance curves of labeled red blood cells in control animals (group 1) and in animals with coagulation activity levels INR 2.03.0 (group 2), INR 1.8-2.0 (group 3), INR 1.6-1.8 (group 4), INR 1.4-1.6 (group 5), and INR 1.3-1.4 (group 6).

anticoagulated animals was determined, and the effect on wound healing was evaluated.

afterward was collected on gauze tissues. The tissues were stored in a container filled with 50 mL of distilled water. This allowed the correct dilution ofthe collected blood to be known. At the end ofthe Materials and methods. collection period, the hemoglobin level of the solution in each Dutch rabbits ('Hollanders') between 16 and 20 weeks old at the container was determined, and the exact amount of blood contained start of the experiment were used, and they had body weights in the beakers could be calculated (Roller and King, 1983). A 10-mL blood sample was taken from each animal. Red blood between 2 and 2.5 kg. The rabbits were caged separately and received dry pellet food. Blood samples were obtained by puncture cells were radioactively labeled with Na251CrO4 by method A, deof the marginal ear vein. scribed by the ICSH Panel on Diagnostic Applications of RadioisoIn each rabbit, four front teeth were removed: the two upper topes in Hematology (1971). The labeled cells were re-injected into central incisors and the two lower incisors. In this way, four well- the respective animals, and 20 min later reference blood samples defined extraction wounds were -obtained. There was no interfer- were taken (2.5 mL in EDTA). Immediately afterward, the four ence with the normal food intake of the animals. The extractions teeth were removed. Blood sampling was repeated on days 1, 3, 4, were carried out as atraumatically as possible, with use of a Kocher 6, 8, and 11. At the end of the experiment, the radioactivity of each hemostatic forceps, after careful syndesmotomy. All extraction blood sample was determined with a gamma-counter. The value of wounds were sutured with nonresorbable silk sutures (3/0). The the sample obtained on day 0 (before the extractions) was regarded as 100% (reference sample). Radioactivity counts of the samples sutures were removed on day 5. The rabbits received neuroleptanalgesia during the interven- taken on the following days were expressed in percentages of the tion. This enabled the teeth to be removed atraumatically (no value ofthe reference sample. Radioactivity disappearance curves unexpected movements) and eliminated pain sensation for the were drawn for each experimental group of animals. Animals were anticoagulated by oral administration ofwarfarin rabbits. Neuroleptanalgesia was induced by subcutaneous administration of a solution offluanisone and phentanylcitrate (Hypnorm®, sodium (Marevang, Evans Medical Ltd., London, England), a couDuphar). The injection of 0.6 mLper kg body weight was scheduled marin derivative. Haugen (1967) used warfarin for anticoagulation 15 min before the extractions. of rabbits and showed that a well-controlled treatment may be Three different, complementary methods were used for assess- carried out in rabbits. Warfarin tablets were crushed, and the ment of blood loss following the extractions. Bleeding during the powder was administered to the rabbits on a teaspoon. On the first first hour following intervention was evaluated by determination of day, the animals received 1.0 mg/kg body weight, on the second day the tooth-socket bleeding time, and blood loss was measured by 0.5 mg/kg b.w., on the third and all following days a laboratory means of a direct method based on hemoglobin determinations. control of the coagulation activity was carried out (prothrombinLong-term blood loss (two weeks following intervention) was as- time determination), and the dose ofthe anticoagulant was adjusted sessed by comparison ofradioactivity disappearance curves (labeled according to the results of this test and the desired level of red-blood-cell technique). anticoagulation. Anticoagulant medication was started seven days The tooth-socket bleeding time was determined in the following before the extractions. way. All tooth sockets were inspected at five-minute intervals, up to The coagulation activity was measured with ThrombotestTm 60 min after the extractions. Oozing wounds were counted. A socket (Nyegaard) on venous blood. The test was performed within one was considered "oozing" when fresh blood appeared after gentle hour of blood collection. The Thrombotest percentages (TT) were pressure with a gauze tissue (analogy with Duke ear-bleeding time). derived from a calibration curve that was especially designed for A few minutes before the extractions, a blood sample was taken rabbit venous blood. from each animal, and the hemoglobin level was determined. All the Wound healing was evaluated by studying the rate of epithelial blood each animal lost during intervention and in the first hour overgrowth or wound closure. Epithelial cells grow only in the Downloaded from jdr.sagepub.com at UNIV OF PITTSBURGH on December 18, 2014 For personal use only. No other uses without permission.

A

~~.

a)

Groups:

Groups:

4r 04

40

2 3

__

30

0

4.)

.,,4 4) ol U to

4

5

A~~~~~~~~

20

0ud

10

_:/

* 2

4

6

8

:i _._4-i0.

0

_

4k)

'U

389

ANTICOAGULATION AND BLOOD LOSS

Vol. 71 No. 2

4'

4J

3

.....-3

14

M

ri

10 :3

P4J

3

Ua)

40L

0

a)

201

1

_2

80

60i

____ 5

.rs

0)

U.6

2

4

6

8

10

12

14

16

Time after extraction (days) 10

Time after extraction (days)

Fig. 3-Decrease of labeled red blood cells in control animals (group 1) and in animals with coagulation activity levels INR 2.0-3.0 (group 2), INR 1.8-2.0 (group 3), INR 1.6-1.8 (group 4), INR 1.4-1.6 (group 5), and INR 1.31.4 (group 6) (% decrease/24 h).

presence of vital supporting tissue. Therefore, wound closure rates will reflect the healing capacity of the underlying tissues. On the first day following the intervention, photographic slides were made of each extraction wound in all rabbits. Photographs were taken with a reference frame in place. This enabled correction to be made for deformities and differences in magnification between the different slides. Before the photographs were taken, the wounds were stained with a solution of lugol. This made the exact location of the wound margins more clear. The surface areas of the different wounds were calculated. The extraction wounds were inspected daily, and the time of complete epithelial closure was recorded for each wound. The experimental groups and a control group (group 1-no anticoagulation) were made up of eight animals each. The experimental animals were anticoagulated at the following levels: group 2 with INR between 2.0 and 3.0, group 3 with INR between 1.8 and 2.0, group 4 with INR between 1.6 and 1.8, group 5 with INR between 1.4 and 1.6, and group 6 with INR between 1.3 and 1.4. The following tests were used for statistical analysis: Results obtained by the labeled red-blood-cell technique and results ofblood loss in the first hour (hemoglobin method) were processed by oneway analysis of variance (ANOVA). When this test yielded a significant F value (5% level), the different pairs of treatments were compared (multiple comparisons) by means of the Tukey test. Results of tooth-socket bleeding times and wound-closure rates were tested by Chi-square statistics.

Results. Results of tooth-socket bleeding times are shown in Fig. 1. Differences were significant at the 1% level between non-anticoagulated animals (group 1) and anticoagulated animals of groups 2, 3, and 4. For groups 5 and 6 (INR value < 1.6), the results were not significantly different from those obtained in control animals. Results obtained with the hemoglobin determination method are shown in the Table. Differences in immediate blood loss measured by this technique were significant at the 5% level(ANOVA) between control animals and animals with INR values > 1.8 (groups 2 and 3). Figs. 2 and 3 show the results of long-term blood loss evaluation.

Fig. 4-Wound-closure rates of extraction wounds of upper incisors in control animals (group 1) and in animals with coagulation activity levels with INR 2.0-3.0 (group 2), INR 1.8-2.0 (group 3), INR 1.6-1.8 (group 4), INR 1.4-1.6 (group 5), and INR 1.3-1.4 (group 6).

When compared with results from control animals, the difference in bloodloss was significant at the 5% level on all days for anticoagulated animals with INR values > 1.8 (groups 2 and 3). The increase in blood loss occurred mainly on days 1 and 4. At coagulation activity levels with INR values between 1.4 and 1.8 (groups 4 and 5), blood loss was significantly greater (5% level) only on day 1. Results in animals with INR values lower than 1.4 (group 6) were not different from those in controls. Fig. 4 shows closure rates of the extraction wounds of upper incisors in the different experimental groups. Differences in epithelial closure rates for wounds of upper incisors were significant at the 1% level between control animals and all experimental groups.

Discussion. Radioactivity disappeared from the peripheral circulation of the experimental animals for three main reasons: first, as the result of breakdown ofthe radioactive element (The half-life of chromium is 27.8 days); second, as a consequence of the turnover of red blood cells [The survival time of rabbit red blood cells lies between 45 and 68 days, and Kozma et al. (1974) reported a mean erythrocyte life span TABLE BLOOD LOSS IN THE FIRST HOUR FOLLOWING TOOTH EXTRACTIONS (HEMOGLOBIN DETERMINATION METHOD)

Group

INR Value, Day 0

Blood Loss (mL)

(mean ± SD)

(mean ± SD)

1

1.04 ± 0.06

1.75 ± 0.38

2

2.50 ± 0.39

4.94 ± 1.43

3

1.97 ± 0.26

3.90 ± 1.17

4

1.71 ± 0.12

2.80 ± 0.41

5

1.52 ± 0.07

2.82 ± 0.79

6

1.37 ± 0.07

2.55 ± 0.69

INR = International Normalized Ratio; SD = standard deviation.

Downloaded from jdr.sagepub.com at UNIV OF PITTSBURGH on December 18, 2014 For personal use only. No other uses without permission.

390

DECLERCK et al.

J Dent Res

February 1992

of 50 days.]; and third, as a result of blood loss (The animals lost BEHRMAN, S.J. and WRIGHT, I.S. (1961): Dental Surgery during Continublood during repeated blood sampling and following tooth removal.). ous Anticoagulant Therapy, JAm Med Assoc 175:159-164. The first two factors (breakdown of the radioactive element and BENOIT, P.W. and HUNT, L.M. (1976): Comparison of a Microcrystalline turnover of red blood cells) can be regarded as identical in all Collagen Preparation and Gelatin Foam in Extraction Wounds, J Oral animals. This leaves post-extraction blood loss as the only variable Surg 34:1079-1083. that can be held responsible for differences in the disappearance BUMP, R.L. and KOLODNY, S.C. (1973): Fibrinolysis: A Possible Factor in rates of the labeled cells in the different experimental groups. the Control of Postoperative Hemorrhage in the Patient with HemoAt therapeutic anticoagulation activity levels, blood loss was philia, Oral Surg Oral Med Oral Pathol 36:195-200. significantly greater than in non-anticoagulated animals. During DUGDALE, H. (1983): The Patient with Bleeding Problems, Dent Clin the experimental period, 2 out of 8 animals in group 2 did not North Am 27:271-288. survive, and this was due to excessive blood loss. These results HAUGEN, J. (1967): Warfarin Dosage. In: Anticoagulant Treatment of suggest clearly that dental extractions at this coagulation activity Rabbits, Bergen: Norwegian University Press, pp. 2-11. level are clinically unacceptable. HUNT, L.M. and BENOIT, P.W. (1976): Evaluation of a Microcrystalline The results also demonstrate that complete correction of the Collagen Preparation in Extraction Wounds, J Oral Surg 34:407-414. coagulation activity is unnecessary. At INR values between 1.3 and ICSH (1971): Recommended Methods for Radioisotope Red Cell Survival 1.4, the amount of blood lost lay within the normal range. At INR Studies. ICSH Panel on Diagnostic Applications of Radioisotopes in Hematology, Br JHaematol 21:241-250. values between 1.6 and 1.8, blood loss was clinically acceptable. This will certainly be the case when local hemostasis-promoting materi- IBARROLA, J.; BJORENSEN, J.; AUSTIN, B.P.; and GERSTEIN, H. (1985): Osseous Reactions to Three Hemostatic Agents, JEndod 11:75als and techniques are used. However, it should be stressed that wound closure rates were negatively influenced in anticoagulated 83. animals (Fig. 4), possibly because of the poorer quality (lower fibrin JOHNSON, W. and LEARY, J. (1988): Management of the Dental Patient With Bleeding Disorders: Review and Update, J Oral Surg 66:297-303. content) of the blood clot formed. Since most socket-packing techniques are reported to retard wound healing, this should be kept in KOZMA, C.; MACKLIN, W.; CUMMINS, L.M.; and MAUER, R. (1974): mind when they are being used. Anatomy, Physiology and Biochemistry of the Rabbit. In: The Biology It is remarkable that blood loss in the different groups of of the Laboratory Rabbit, S.H. Weisbroth, R.E. Flatt, and A.L. Kraus, Eds., London: Academic Press, pp. 50-69. anticoagulated animals was greater than that in control animals, not only on day 1 but also on day 4. A possible explanation for this MULLIGAN, R. and WEITZEL, K.G. (1988): Pretreatment Management of increase in blood loss on day 4 could be the existence of a discrepancy the Patient Receiving Anticoagulant Drugs, JAm Dent Assoc 117:479483. between two steps in the wound-healing process. Fibrinolytic degradation of the blood clot progresses at a physiological rate, while ROLLER, N.W. and KING, O.H. (1983): A Simple Way to Determine the the proliferation of young connective tissue is slowed because of the Amount of Blood Lost During Oral Surgery, J Oral Maxillofac Surg lower fibrin content of the blood clot. This finding in our animals 41:618-620. corresponds well with the earlier reported finding that the fibrino- SPOUGE, J.D. (1964): Hemostasis in Dentistry, With Special Reference to lytic activity (plasminogen activator activity) reaches a maximum Hemocoagulation. III: Principles Underlying Clinical Hemostatic Pracbetween day 4 and day 6 (Wells, 1976). This finding also corretices in Patients with Faulty Coagulation, Oral Surg Oral Med Oral sponds very well with our clinical experience that bleeding episodes Pathol 18:701-712. occur mainly on the third and fourth post-operative days. The same VAN DE BESSELAER, A.M.P.H.; GRALNICK H.R.; and LEWIS, S.M. finding was observed in a clinical study in hemophiliacs (Vinckier (1984): Joint ICTHAICSH Proposed Policy Statement with Respect to and Vermylen, 1985). Reporting the Prothrombin Time in Oral Anticoagulant Control. The Hague: Martinus Nijhoff. VINCKIER, F. and VERMYLEN, J. (1984): Wound Healing Following Acknowledgments.

We thank Dr. W. Goossens and the laboratory personnel of the Department of Hematology for the laboratory analyses and advice. REFERENCES

ASKEY, J.M. and CHERRY, C.B. (1956): Dental Extractions during Dicumarol Therapy, CA Med 84:16-17. BAILEY, B.M.W. and FORDYCE, A.M. (1983): Complications of Dental Extractions in Patients Receiving Warfarin Anticoagulant Therapy, Br Dent J 155:308-310.

Dental Extractions in Rabbits. Effects of Tranexamic Acid, Warfarin Anticoagulation and Socket Packing, J Dent Res 63:646-649. VINCKIER, F. and VERMYLEN, J. (1985): Dental Extractions in Hemophilia: Reflections on 10 Years' Experience, Oral Surg Oral Med Oral Pathol 59:6-9. WALDREP, A.C. and McKELVEY, L.E. (1968): Oral Surgery for Patients on Anticoagulant Therapy, J Oral Surg 26:374-380. WELLS, T.J. (1976): A New Concept in the Control of Acute Gingival Hemorrhage, J Oral Surg 34:435-437. ZIFFER, A.M.; SCOPP, I.W.; BECK, J.; BAUM, J.; and BERGER, A.R. (1957): Profound Bleeding after Dental Extractions During Dicoumarol Therapy, N Engl J Med 256:351-353.

Downloaded from jdr.sagepub.com at UNIV OF PITTSBURGH on December 18, 2014 For personal use only. No other uses without permission.

Influence of anticoagulation on blood loss following dental extractions.

The coagulation activity level at which oral surgery procedures can be performed in anticoagulated patients without triggering bleeding complications ...
488KB Sizes 0 Downloads 0 Views