THROMBOSIS RESEARCH 68; 33-43,1992 0049-3848/92 $5.00 + .OOPrinted in the USA. Copyright (c) 1992 Pergamon Press Ltd. All rights reserved.

Heparin Study in Internal Medicine (HESIM): design and preliminary

results

J. Harenbergl, P. Roebruckz, G. Stehlel, W. Habscheids, M. Biegholdt?, D.L. Heenel the HESIM-Group*

and

1st Department of Internal Medicine, Faculty of Clinical Medicine Mannheim, University of Heidelberg, 6800 Mannheiml, Institute for Medical Biometry and Informatic, University of Heidelberg, 6900 Heidelbergz, Department of Medicine, University of Wthzburg, 8700 Wtirzburg3, and Sanofi Winthrop GmbH, 8000 Milnchen4, Germany (Received 19.3.1992; accepted in revised form 2.7.1992 by Editor K. Lechner) (Received by Executive Editorial Office 24.8.1992)

ABSTRACT The Heparin Study in Internal Medicine (HESIM) compares the efficacy and safety of an unfractionated (UF) heparin with a low molecular weight (LMW) heparin (CY 216 D) for prevention of proximal deep vein thrombosis @VT) and pulmonary embolism (PE) in medical inpatients with a high risk for development of thromboembolism. Patients are randomized and receive three times daily 5000 IU UF heparin or once daily 3100 IU LMW heparin and two placebo injections subcutaneously for 10 days. All patients are screened for the presence of proximal DVT at day 1 and 10 by real-time B-mode compression sonography and for PE by repeated clinical examinations. Perfusion scintigraphy is used for confirmation of the clinical diagnosis of PE. The study protocol includes a stratified randomization of patients on admission to the hospital according to one of the following main diagnoses: malignant disease, cardiovascular disease, bronchopulmonary disease, neurologic disease, and other diseases. The present study may serve as a model for further clinical trials in medical inpatients using the biometric approach of statistical analysis for proving equivalence of drug efficacy, and to adopt less sensitive but noninvasive methods for the detection of primary endpoints. * HESIM - group: SteerinP Committee: D.L. Heene, Mannheim; J. Harenberg, Mannheim; U. Schmitz-Huebner, Herford; F. Heinrich, Bruchsal; P. Roebruck, Heidelberg; P. dAzemar, Paris; M. Biegholdt, Mtinchen. E-e: K. Lechner, Wien, M. Georgi, Mannheim, W. Habscheid, Wurzburg, F.A. Spengel, Mznchen. Critical Event Committee: G. Mtiller-Berghaus, Bad Nauheim; H. Rasche, Bremen; Th. Gasser, Zurich. Particinants: W. Becker, Bochum; D.L. Heene, J. Harenberg, Mannheim; D. Heinrich, Wetzlar; F. Heinrich, Bruchsal; R. Lorenz, Mtinchen; P.C. Ostendorf, Hamburg; U. Schmitz-Huebner, Herford; C. Stautner-Brtickmann, Mtinchen; H. Stiegler, Mtinchen; W. Theiss, Mtinchen. Correspondence to: J. Harenberg, MD, I. Medizinische Klinik, Klinikum Mannheim, Fakultlt fur Klinische Medizin Mannheim, Universitat Heidelberg, Theodor-Kutzer-Ufer, D-6800 Mannheim 1 Key words: Prophylaxis of thromboembolism, clinical trial, B-mode compression sonography, heparin, low molecular weight heparin 33

34

HEPARIN STUDY IN CLINIC

Vol. 68, No. 1

INTRODUCTION Clinical studies have demonstrated an equal or superior effect of low molecular weight (LMW) heparins in reducing the incidence of postoperative thromboembolism compared to unfractionated (UF) hepain (1.2). Advantages of low molecular weight heparin were the improved pharmacodynamics (3,4) and lower incidence of bleeding complications (5). In internal medicine UF heparin has been demonstrated to effectively reduce the incidence of venous thrombosis in several diseases. In myocardial infarction the incidence of DVT ranges from 17 to 34% without prophylaxis and is reduced to 3-14% with low dose heparin (6,7,8). In acute stroke the occurrence of deep vein thrombosis (DVT) is decreased from about 70% without, to 12-25% with use of low dose heparin (9, 10). Today the use of radiolabelled fibrinogen for detection of thrombosis is ethically not acceptable for clinical studies in medical patients. Therefore, other screening methods for detection of the development of DVT are needed. B-mode compression sonography (11, 12, 13) as well as duplex sonography (14,15) have been demonstrated to be sensitive and accurate for diagnosis of proximal DVT, sensitivity and specificity being above 92 %, as compared to phlebography. The sensitivity for diagnosis of distal thrombosis varies from 30 to 85% (11, 12, 16, 17, 18). Therefore, the use of B-mode compression or duplex sonography in clinical studies should be restricted to screening for proximal DVT. Clinical symptoms of pulmonary embolism (PE) occur in about 0.5-1.0 % of patients in internal medicine (19). Clinical parameters such as age, history of DVT or PE, carcinoma, infection, ischemic heart disease or chronic obstructive lung disease are connected with an increased risk of PE (20). Clinical variables predict the presence of PE with a reasonable sensitivity of 85 % but a low specificity of 35 % (19). Thus, objective testing by perfusion lung scanning is clearly indicated in patients with suspected PE (21). However, general screening for PE by repeated perfusion scintiscans is not feasible because of ethical reasons and because specificity is only 60% as compared with pulmonary angiography (21). Three small studies have demonstrated the safety of low molecular heparin for prophylaxis of deep vein thrombosis in medical patients (22,23,24). We now report the study protocol of the first randomized, controlled double blind multicentre trial for comparison of the efficacy of a low molecular heparin with unfractionated heparin for the prevention of proximal deep vein thrombosis and pulmonary embolism in medical patients. Patients are expected to have a high risk to develop thromboembolism due to the severity of the underlying disease and the concomitant risk factors. The efficacy of low dose heparin in medical inpatients for prophylaxis of thromboembolism is generally accepted as outlined above. Therefore a placebo controlled study was regarded as not acceptable ethically. Primary endpoints are the occurrence of proximal deep vein thrombosis and/or pulmonary embolism. The ethical committee of the Faculty for Clinical Medicine Mannheim of the University of Heidelberg has accepted the study protocol. An independent monitoring is performed according to the “Good Clinical Practice” (GCP) guidelines (25). STUDY

DESIGN

Patients and Endpoints Inclusion Criteria: The inclusion criteria were chosen to define a patient population at high risk for development of thromboembolic complications. Thus the following patients are included into the study: age between 50 and 80 years, expected duration of bed rest and hospital stay >lO days, indication for prophylaxis of thromboembolism, and informed consent. One or more risk factors for development of thrombosis should be present in patients and are recorded: obesity, varicosis, chronic venous insufficiency, postthrombotic syndrome, intake of oral contraceptives or estrogen, thrombocytosis > 450,000/~1 or hyperviscosity syndrome

Vol. 68, No. 1

HEPARIN STUDY IN CLINIC

35

(fibrinogen > 500 mg/dl), previous myocardial infarction, thrombotic cerebral infarction or peripheral arterial ischemia. Exclusion Criteria: The following patients are excluded from the study: known intolerance towards heparin, thrombocytopenia (c 80,000/~1), hereditary or acquired coagulation disorder, acute DVT, pretreatment with heparin other than study medication, regular intake of medication influencing blood coagulation, indication for treatment with medication with effect on blood coagulation, fibrinolysis or platelets, indication for high dose treatment with heparin, diseases with unfavorable short term prognosis, septicemia with gramnegative bacteria, disseminated intravascular coagulation, known vascular aneurysm, fixed hypertension, severe diabetic microangiopathy, history of cerebral bleeding, recent hemorrhage of the gastrointestinal tract, the urogenital tract and other organs, severe renal impairment (creatinine > 3 mg/dl), acute glomerulonephritis, necrotic pancreatitis, severe liver dysfunction (prothrombin time/Quick c 60 %) and/or decompensated hepatic cirrhosis, and endocarditis lenta. The postoperative phase is no exclusion criterion. Randomization: The study protocol includes a stratified randomization of patients on admission to the hospital according to one of the following subgroups: malignant disease, cardiovascular disease, bronchopulmonary disease, neurologic disease, or other diseases. Endpoints: Primary endpoints are proximal deep vein thrombosis @VT) of the lower limbs and/or pulmonary embolism (PE). The presence of a primary endpoint is screened in all patients using repeated clinical examinations and compression sonography (see below). Only one sndpoint per patient is counted for statistical evaluation. Secondary endpoints are venous thromboses of other locations, arterial embolism, myocardial infarction, and death within the study pernod. The comparison of the treatment efficacy relates to patients with an observation period of 8 to 10 days. Treatment and Methods Medication: One group of patients is treated with LMW heparin (CY 216 D), 3100 IU (anti Xa) / 0.3 ml and two placebo injections per day. The second group receives 5000 IU of calcium heparin three times daily. LMW heparin CY 216 D is obtained from unfractionated calcium heparin by chemical depolymerization followed by ethanol extraction. The mean molecular weight is 4500 daltons. The compounds are dissolved in distilled water and supplied in syrimges containing a volume of 0,3 ml. Placebo is isotonic sodium chloride solution in identical syringes containing 0,3 ml. Treatment protocol: Patients meeting the inclusion criteria are entered into the study within 12 hours after admission to the hospital. For each stratification group of every center subgroups of n=20 are randomized to ensure a balanced distribution of the two treatments within the five strata and the centers. The injection are given three times daily at 8 hours intervals. In the LMW heparin group the first injection contains the active compound and the other two injections the placebo. The first injection (verum) is given within 12 hours after admission of the patients to the hospital, the other injections in the LMW heparin group at day one containing placebo. Subsequently the first injection is given between 6 and 7 a.m., the second injection between 2 and 3 p.m. and the third injection between 10 and 11 p.m. Examination protocol: Patients are to be examined at days 1, 4, 6, 8 and 10 for clinical symptoms of DVT. The examination protocol includes the signs of Payr and Homan, spontaneous tenderness or pain on compression of the thigh, popliteal space, adductory channel, and groin (Fig. 1).

36

Vol. 68, No. 1

HEPARIN STUDY IN CLINIC

admission to hospital start of prophylaxis 1 1

day

2

clinical examination

x

B-mode compression sonography

__ x __

laboratory test

4(‘1)

6(?1)

X

X

8- 11 X

X

X*

X

Figure 1: Diagram of diagnostic procedures included in the study protocol in the HESIM trial _ heparin study in internal medicine (x = examination, * = before start of the prophylaxis).

Deep-vein thrombosis: The diagnosis of proximal DVT is ma& by comparison of the results of the compressibility of the veins using B-mode or duplex sonography at the end of the study (day 8 - 11) with the initial (day 1) findings. Proximal veins are the popliteal, femoral and iliac veins. B-mode compression sonography is performed as described by Habscheid et al.( 17). The compressibility of the veins using duplex-sonography is carried out according to Millewich et al (15). All positive or undecisive sonographic results are documented photographically or by video. Incomplete or complete incompressibility of a vein is considered to prove the presence of venous thrombosis. Phlebography is performed in patients with a negative or inconclusive sonographic result but clinical symptoms suggestive of thromboembolism (26).

I

B-mode compression sonography

negative: no DVT t medication continued

I

positive: DVT t medication stopped

+ validation committee +

El

event

Figure 2: Flow chart of the screening for DVT in the HESIM trial heparin study in internal medicine

Vol. 68, No. 1

HEPARIN STUDY IN CLINIC

37

In case the patient refuses phlebography, an inconclusive sonographic result will be recorded as negative. All positive or unclear findings will be evaluated by the validation committee. Patients with no clinical signs of DVT but positive sonographic examination at day 1 are to be excluded from the study. Positive sonographic results at the initial examination will also be evaluated by the validation committee (Fig. 2). Quality control of sonographic examinations will be performed by a member of the validation committee by mailing of 50 examination records to the participating centers in portions of 10 to 15 records during the study. This quality control procedure during the study period should ensure that all investigators identify positive sonographic results as positive and negative sonographic results as negative which is critical for the outcome of the study. Pulmonary embolism: Patients will be examined for clinical symptoms of PE on day 1, 4, 6, 8 and 10 (Fig. 1). The examination protocol includes the following: tachycardia, dyspnea, tachypnea of recent onset , cough with or without hemoptysis, signs of arterial hypoxemia in blood gas analysis, chest pain when breathing, rise of body temperature, pleural friction sound or signs of infarction pleuropneumonia. If pulmonary embolism is clinically suspected, chest radiography, ECG, laboratory tests and perfusion scintiscan are performed. If possible, a ventilation scintiscan is made additiqnally. Otherwise it is attempted to perform a control perfusion scintiscan. If necessary for diagnosis of pulmonary embolism, pulmonary angiography will be performed in addition to the scintigraphic procedures (27) (Fig. 3). Safety and adverse reactions: The size of hematomas at injection sites is measured, every other day (i.e. day 4, 6, 8, 10) and the number of hematomas with a diameter above 2.5 cm is recorded. Patients are also examined for hematuria and hematomas at others than the injection sites, and side effects such as alopecia, pruritus, or allergic reactions. Clinical chemistry analyses are performed on days 1 and 10 of the study and include the following parameters: asparaginaminotransferase, alanin-aminotransferase, gamma-glutamyltranspeptidase, cholesterol, triglycerides, lactat dehydrogenase, alkaline phosphatase, urea, and serum creatinine. Hematological evaluation includes hematocrit, erythrocyte, leucocyte and thrombocyte count, prothrombih time,

I

clinical signs dayltoll

I

medication stopped

validation committee

c3 event

Figure 3: Flow chart of screening for PE in the HESIM trial - heparin study in internal medicine

38

HEPARIN STUDY IN CLINIC

Vol. 68, No. 1

antithrombin III, and fibrinogen. All clinical chemistry parameters are measured using commercially available test systems. All adverse reactions are classified according to their severity as slight, moderate or severe. The relationship between the adverse reaction and the study medication will be classified by the investigators as: not related, uncertain, possibly related, probably related or definitely related Treatment of patients with study heparin is to be terminated in the following cases: DVT before or at the end of the treatment period of 10 days, PE before or at the end of the treatment period, hemorrhage which requires therapy and necessitates discontinuation of heparin treatment, thrombocytopenia < 8O,OOO/fiand other severe adverse reactions which occur during the treatment period. Sample size and biometric

considerations

Data for calculation of the sample size: In the I. and II. Department of Internal Medicine of the Mannheim University Hospital, 1858 out of 6253 patients received low dose heparin for more than 7 days during their hospital stay in 1986. Eleven of these heparin-treated patients reportedly developed DVT, additional 15 patients developed PE. This amounts to 26 out of 1858 patients or an incidence of 1.4 %. Only 2 out of the 26 patients were outside a range between 5080 years of age. No screening for the development of DVT or PE was carried out in these patients in 1986. The frequency of DVT in medical patients with prophylactic anticoagulant treatment ranges from 3 - 25 % using the radiofibrinogen uptake test (6-10). About 20 % of these thromboses occur in the proximal veins (28). As mentioned above, the sonographic procedures employed only detect proximal DVT with adequate sensitivity and specificity. In consequence, a rate of about 3 % of occurrence of DVT in severely ill medical inpatients with a high risk to develop thromboembolism can be expected if B-mode compression or duplex sonography will be utilized as screening methods. The incidence of clinical signs of PE in severely ill medical inpatients ranges from 0.5-1.0 % (18). Since general screening of all patients by perfusion lung scanning is not feasible, the estimated combined rate of occurrence of DVT and PE has to be set at 4 % for a clinical trial in medical inpatients on prophylaxis of thromboembolism with low dose heparin. In postoperative medicine the superiority of the LMW heparin used in this trial has been demonstrated compared with UF heparin (2,5). Therefore the aim of the study was to demonstrate the superiority of one daily S.C.injection of the LMW heparin CY 216 D compared with three times daily S.C.calcium heparin. Initial hypothesis and sample sizes: Pn and Ps are probabilities of thromboembolic endpoints with the new and the standard treatment, respectively. Based on the expected incidences mentioned above it was planned to test with Fisher’s exact test the two-sided hypothesis H’: Pn = Ps with a type one error probability a = 0.05 and with a power 1 - /3= 0.80 for hypothetical values of Ps = 0.04 and Pn = 0.02 or 0.07. With two interim analyses a maximal number of about 1300 patients per group results from Bonferroni’s inequality and the approximation according to Casagrande et al. (29). Adaptation of the hypothesis: After about 800 patients had been included in the trial, a global incidence for thromboembolic events of approximately 1% was observed. The low incidence of primary endpoints was mainly due to the fact, that an unexpectedly high proportion of about 1.5% of the patients exhibited clinically inapparent DVT upon the initial sonographic examination and had to be excluded from the study. This substantial deviation from the expected average of 3% endpoints necessitated reconsideration of the original hypothesis of superiority of the new treatment . In order to prove superiority of one treatment with overall incidences as low as those observed, sample sizes would be necessary which am two to three times as large as those calculated initially. Therefore the Steering Committee decided to change the aim of the study from showing superiority of one treatment to proving equivalence of both treatments in order to make a meaningful conclusion of the study possible. Thus the sample size can be reduced by switching from a two-sided to an one-sided test.

Vol. 68, No. 1

HEPARIN STUDY IN CLINIC

39

Revised planning of the study: To prove the equivalence of the new and the standard treatment means to test the null-hypothesis H: Pn 2 Ps + A that the new treatment is inferior against the alternative hypothesis K: Pn c Ps + A which is interpreted as equal efficacy. For further considerations it was decided to assume Ps = 0.01 and A = 0.015. The sample size calculation is based on the difference of the incidences rather then on the oddsratio due to the uncertainty regarding the assumption on Ps. With the figures above the odds-ratio would be 2.6. The same odds-ratio with e.g. Ps = 0.02 however would result in a difference of A = 0.03, which seems to be unacceptable as equivalence of such low incidences. At present no exact test is available for equivalence problems based on differences. Thus an approximation will be used which relates to the asymptotic normality of observed rates of thromboembolic events. Since the sample sizes are quite large this approximation should be acceptable. Sample size calculations are performed as cited above (29). The sample sizes according to others (30) generally are judged to be too small. Therefore in our calculation of the sample size a modification of the formula was used which gives slightly greater values and takes into account a continuity correction in analogy to the derivation (31). With type one error probability a = 0.05 and power 1 - p = 0.80 if actually Pn = Ps, about 1100 patients are necessary for each treatment group. In this calculation one interim analysis (after 700 patients per group) is taken into account which could permit an early termination of the study with the acceptance of K (equivalence) or with stating that the new treatment is inferior to the standard treatment. Within the context of sample size determination “patients” always means evaluable patients according to the study protocol. The overall error probabilities are controlled by means of Bonferroni’s inequality. The nominal type one error probabilities for H will be 0.01 in the interim analysis and 0.04 for the final test. The resulting power is at least 0.,84 if Pn = Ps. Thus an interim test with significance level 0.04 can be used to detect major inferiority of the new treatment i.e. to test the null hypothesis Pn I Ps against Pn > Ps preserving an overall power of 0.80 for the equivalence problem. The necessity of a separate analysis against inferiority results from the formulation of the one-sided central test. PRESENT

STATUS

OF THE TRIAL

The study started in May 1990 with 10 trial centers. So far 1672 patients have been included in the tial. 1115 study protocols have been validated. The inclusion rate is about 90 patients per month and has been constant during the trial (Fig. 4). The drop-out rate is about 15 %. In approximately 1.5 % of the patients recent proximal DVT without clincial signs is detected at the first examination by B-mode compression sonography. The incidence of primary endpoints amounts to about 1 to 1,5 %. Based on the validated case reports 13 % of the patients have been stratified to group 1 (malignant disease), 34 % to group 2 (cardiovascular disease), 11 % to group 3 (bronchopulmonary disease), 19 % to group 4 (neurologic disease), and 23 % to group 5 (other diseases). Fifty six percent of the patients were female and 44 % male. The mean age of the patients was 70 years. Seventeen percent were aged from 50 to 60 years, 28 % from 60 to 70 years, and 55 % from 70 to 80 years. DISCUSSION Several small scale clinical studies have demonstrated the benefit of low dose heparin for the prophylaxis of DVT in patients with myocardial infarction and recent stroke. However, large scale studies are not available so far. Nevertheless, low dose heparin is widely accepted in severely ill patients in internal medicine (32). Since a placebo-controlled study does not seem ethically acceptable, we have decided to perform a controlled clinical trial comparing subcuta

HEPARIN STUDY IN CLINIC

40

Vol. 68, No. 1

(4 2000 1800 1600 1400 1200 1000 800 600 400 200 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 month Figure 4: Cumulative inclusion rate of patients (n) in the HESIM trial - heparin study in internal medicine. neous unfractionated heparin three times daily with low molecular heparin once daily plus two placebo injections. Screening for the development of DVT in operative medicine is generally performed using the radiofibrinogen uptake test. Taking into consideration the low incidence of thromboembolic events and consequently the large numbers of patients required for such a randomized study in internal medicine, most patients would be subjected to radioactivity without a clear-cut clinical necessity. In addition, commercially available radiolabelled fibrinogen preparations may not be absolutely free of viral antigen (33). Radiolabelling of autologous fibrinogen of each patient does not seem feasible in large clinical studies. For these reasons radiolabelled fibrinogen was not used as screening method for detection of DVT in the present protocol. In the last few years new non-invasive techniques, such as impedance plethysmography and different sonographic methods for detection of deep vein thrombosis have been introduced (11-17). Real time B-mode compression sonography and duplex-sonography have been shown to be nearly as sensitive and specific as phlebography in detection of proximal DVT of the lower extremity. In contrast, diagnosis of distal lower limb thrombosis appears to be rather difficult and conflicting, using the new non-invasive techniques (11, 12, 16, 17). In patients with symptomatic DVT the sensitivity and specificity of B-mode compression and duplex sonography ranges from 87 to 100 % and from 91 to 100 %, respectively (11-17). The sensitivity of the sonographic techniques of the popliteal and femoral veins in asymptomatic patients varies from 80 to 90 % and the specificity from 94 to 100 % (14, 34, 35). The sensitivity of the sonography of asymptomatic calf vein thrombosis is only 39 to 43% (18). Being aware of these limitations, we decided to adopt the sonographic techniques for screening of DVT in the large scale clinical trial and to restrict statistical evaluations to results obtained upon ultrasound examination of the popliteal and femoral area of the lower limbs. Incidence of symptoms of PE in internal medicine is approximately 0.5 - 1.0 % (17). Asymptomatic pulmonary thrombi may occur more frequently. The high fibrinolytic potential of the pulmonary circulation resolves small emboli within a short time. In order to prevent irrelevant statistical figures, we decided not perform pulmonary scintigraphy in all patients routinely. Clini-

Vol. 68, No. 1

HEPARIN STUDY IN CLINIC

41

cal variables have been shown to predict the presence of pulmonary embolism with a reasonable sensitivity of 85 % but a low specificity of 35 % (20). Therefore, we decided to screen patients for the presence of PE by repeated clinical examination and subject patients with clinical signs of PE to lung scanning (21). In the present study approximately 1.5 % of the patients undergoing the initial ultrasound examination exhibited acute venous thrombosis without obvious clinical symptoms. Thus, the total incidence of DVT in the study population is lowered from 3% of the initial planning to about 1.0 to 1.5 %. Accordingly, the interim analysis gave evidence that the global incidence of primary endpoints will not exceed 1.5 % in the study population. Other reasons for this unexpectedly low incidence could be the lack of an acute event as in operative medicine, and the rather short observation period as compared to the long duration of the illness. Due to these facts the study committees decided to change the hypothesis of the trial from superiority of the new drug to equivalence of the two treatment groups. The monthly inclusion rate of the patients by the participating study centers indicate that patients are not selected and that they are included continuously upon admission to hospital. Thus the study will be completed within the foreseen period. REFERENCES

1.

KAKKAR, V.V., DJAZAERI, B., FOK, J., FLETSCHER, M., SCULLY, M. and WESTWICK, J. Low molecular weight heparin and prevention of post-operative deep vein thrombosis. Br. Med. J. 284.75-379, 1982.

2.

THE EUROPEAN FRAXIPARIN STUDY (EFS) GROUP. Comparison of a low molecular weight heparin and unfractionated heparin for the prevention of deep vein thrombosis in patients undergoing abdominal surgery. Br. J. Sure 15, 1058-1063, 1988.

3.

FAREED, J., KUMAR, A., WALENGA, J., EMANUELE, R., WILLIAMSON, K. and HOPPENSTAEDT, D. Antithrombotic actions and pharmacokinetics of heparin fractions and fragments. Nouv. Rev. Fr. Hematol, 26,267-275, 1984.

4.

HARENBERG, J., WURZNER, B., ZIMMERMANN, R. and SCHEI’TLER, G. Biovailability and antagonization of the low molecular weight heparin CY 216 in man. Thromb. & 44, 549-554, 1986.

5.

KAKKAR, V.V. and MURRAY, G. Efficacy and safety of low molecular weight heparin (CY 216) in preventing postoperative venous thromboembolism: a cooperative study. &.. J. Surea, 786-91, 1985.

6.

WARLOW, C.H., BEA‘ITIE, A.G., TERRY, G., OGSTON, D., KENMURE, C.F. and DOUGLAS, A.S. A double-blind trial of low doses of subcutaneous heparin in the prevention of deep-vein thrombosis after myocardial infarction. Lancet 2.934937, 1973.

7.

PITT, A., ANDERSON, S.T., HABERSBERGER, P.G. and ROSENGARTEN, D.S. Low dose heparin in the prevention of deep vein thromboses in patients with acute myocardial infarction. Am. Heart J,B, 574-578,198O

8.

EMERSON, P.A. and MARK, S.P. Preventing thromboembolism after myocardial infarction - effect of low dose heparin or smoking. Br. Med. L 2&l,, 18-20, 1977

9.

MC CARTHY, S.T., TURNER, J.J., ROBERTSON, D. and HAWKEY, C.J. Low-dose heparin as a prophylaxis against deep-vein thrombosis after acute stroke. Lancet 1, 800801, 1977.

42

HEPARIN STUDY IN CLINIC

10. HEINRICH, F. and CZECHANOWSKI, frischem ischHmischen zerebrovaskultiem 1981.

Vol. 68, No. 1

B. Prophylaxe venbser Thrombosen bei Insult. Q&I&. Med. Ws&lQ!j, 1254-1260,

11. CRONAN, J.J., DORFMAN G.S. and GRUSMARK, J. Lower extremity deep venous thrombosis: further experience and refinement of US-assessment. Radiologv 168, lOl107, 1988. 12. LENSING, A.W.A., PRANDONI, P., BRANDJES, D., HUISMAN, P.M., VIGO, M., TOMASELLA, G., KREKT, J., TEN CATE, J.W., HIUSMAN, M.V. and BULLER, H.R. Detection of deep-vein thrombosis by real-time B-mode ultrasonography. N. Frtgl& Els;d m, 342-345, 1989. 13. HABSCHEID, W. and WILHELM, T. Diagnostik der tiefen Beinvenenthrombose Real-time-Sonographie. Dtsch. Med. Wschr,m, 586-591, 1988.

durch

14. BARNES, R.W., LAVENDER, R.C. and ERNEST, J.F. Perioperative asymptomatic veft;;;thrombosis. Role of duplex scanning versus venography. J. Vast. Surg, 2,25 l-260, 15. MILLEWICH, LA., BEDFORD, G.R., BEACH K.W. and STRANDNESS, D.E. Diagnosis of deep venous thrombosis, a prospective study comparing duplex scanning to contrast venography. Circulatiou29,810-814, 1989. 16. ELIAS, A., LECORFF, G., BOUVIER, J.L., BENICHOU, M. and SERRADIMIGNI, A. Diagnosis of deep venous thrombosis: value of real time B mode ultrasound imaging and continuous wave doppler (echography-doppler) compared to venography. h Anpio.6, 175-182, 1987. 17. HABSCHEID, W., HGHMANN, M., WILHELM, T. and EPPING, J. Real-time ultrasound in the diagnosis of acute deep venous thrombosis of the lower extremity. Angiology 41, 599-608, 1990. 18

JONGBLOETS, L.M.M., LENSING, A.W.A., BULLER, H.R. and ten CATE, J.W. The value of real-time compression ultrasonography in detecting deep-vein thrombosis in asymptomatic high-risk patients. Thromb. Res,& Suppl. 1, C 88, 1992

19. HIRSH, J., HULL, R.D. and RASCOB, G.E. Epidemiology and pathogenesis of venous thrombosis. J. Am. Coll. Cardiol, S, 104B-113, 1986. 20. HULL, R.D., RASKOB, G.E., CARTER, C.J., COATES, G., GILL, G.J., SACKETT. D.L., HIRSH, J. and THOMPSON, M. Pulmonary embolism in outpatients with pleuritic chest pain. Arch. Intern. Med .I& 838-844, 1988. 21. HULL, R.D., HIRSH, J., CARTER, C.J., RASKOB, G.E., GILL, G.J., JAY, R.M., LECLERC, J.R., DAVID, M. and COATES, G. Diagnostic value of ventilation-perfusion lung scanning in patients with suspected pulmonary embolism. m & 819-828, 1985. 22. DAHAN, R., HOULBERT, D., CAULIN, C., VILTART, C., WOLER, M. and SEGRESTAA , J. M. Prevention of deep vein thrombosis in elderly medical inpatients by a low molecular weight heparin: a randomized double-blind trial. Haemosta& 16, 159-164, 1986. 23. PONIEWIERSKI, M., BARTHELS, M., KUHN, M. and POLIWODA, H. Uber die Wirksamkeit niedermolekularen Heparins (Fragmin) in der Thromboembolieprophylaxe bei intemistischen Patienten. Med. Klin, s;Z, 241-245, 1988.

Vol. 68, No. 1

HEPARIN STUDY IN CLINIC

43

24. HARENBERG,

J., KALLENBACH, B., MARTIN, U., DEMPFLE, C.E., ZIMMERMANN, R., KUBLER, W. and HEENE, D.L. Randomized controlled study of heparin and low molecular weight heparin for prevention of deep-vein thrombosis in medical patients. Thromb. Res, 5p, 639-650, 1990.

25.

EEC note for guidance: good clinical practice for trials on medicinal products in the European Community. Phan-nacol Toxic01 a,361-372,199O

26.

RABINOV, K. and PAULIN, S. Roentgen diagnosis of venous thrombosis in the leg. Arch. Surg,m, 134-144, 1972.

27.

BIELLO, D.R., MATTAR, A.G., MC NIGHT, R.C. and SIEGAL, B.A. Interpretation of ventilation-perfusion studies in patients with suspected pulmonary embolism. Am. J. Radial. lu, 1033-1037, 1979.

28.

MUCK, A.O., MUCK, S. and HEINRICH, D. Veniise thromboembolische Komplikationen bei frischem zerebrovaskularem Insult. II. Prophylaxe mit niedrigdosiertem Heparin. Med. Welt a, 843-848, 1989.

29. CASAGRANDE,

J.T., PIKE, M.C., and SMITH, P.G. Improved approximate formula for calculating sample sizes for comparing two binomial distributions. Biometrics 2,483486, 1978

30. BLACKWELDER,

Trials 2,345353,

W.C. “Proving the null hypothesis” in clinical trials. Corm. Clin, 1982

31. RODARY, C., COMNOUGUE,

C., and TOURNADE, M.F. How to establish equivalence between treatments: A one-sided clinical trial in paediatric oncology. &rt. in Me& 8, 593-598, 1989

32. HEENE, D.L. Antikoagulantientherapie

und Myokardinfarkt. Hlmostaseolo&

1, 25-29,

1981. 33. BUNDESGESUNDHEITSAMT.

Aufbereitungsmonographie

der Kommission 11. B Anz,

15. Februar, 790, 1991. 34. BORRIS,

L.C., CHRISTIANSEN, H.M., LASSEN, M.R., OLSON, A.D. and SCHOTI’, P. Real-time B-mode ultrasonography in the diagnosis of postoperative deep vein thrombosis in non-symptomatic high-risk patients. Eur. J. Vast. Sure. 4, 473-475, 1990.

35. WOOLSON, S. T., MC CRORY, D.W., WALTER, J.F., MALNEY, W.J., WAIT, J.M.

and CAHILL, P.D. B-mode ultrasound scanning in the detection of proximal venous thrombosis after total hip replacement. J. Bone Joint Su rg, m, 983-987,199O.