Therapy of venous thromboembolism in patients with brain metastases

493

Therapy of Venous Thromboembolism in Patients with Brain Metastases David Schiff, M.D., and Lisa M . DeAngelis, M.D.

Background. Deep venous thrombosis (DVT) and pulmonary embolism (PE) are common in patients with brain metastases. Few data exist to help guide the clini- cian's choice between the two therapeutic options of anticoagulation and inferior vena cava filter placement.

Methods. The authors reviewed their institutions' experience with the treatment of venous thromboembolism in 51 adult patients with known brain metastases since 1980.

Results. Ten patients were initially treated with Greenfield filters; four (40%) had recurrent nonfatal thromboembolic events (two PE and two DVT), and three required anticoagulation. Thirty-nine patients were treated initially with anticoagulation; none of these patients later received filters. Two patients with DVT were untreated and both died of PE. Among 42 patients who received anticoagulation, the duration of anticoagulation ranged from 5 to 563 days (mean, 100 days). Two patients who received anticoagulation experienced devastating central nervous system hemorrhage in the setting of supratherapeutic anticoagulation by conventional laboratory criteria. A third patient experienced a minor deterioration, possibly attributable to hemorrhage, for a 7% (3 of 42) incidence of serious central nervous system complications. Three asymptomatic patients developed hyperdensity within metastases on routine follow-up noncon- trast computed tomography scan, suggesting possible intratumoral hemorrhage. Three patients taking warfarin had recurrent DVT with prothrombin time between 15.1 and 17.7. Systemic bleeding complications were generally minor and occurred in only eight patients (19%).

Conclusion. Anticoagulation is more effective than Greenfield filters and acceptably safe when maintained in the therapeutic range in most patients with brain me-

From the Department of Neurology, Memorial Sloan-Kettering Cancer Center, and the Department of Neurology and Neuroscience, Cornell University Medical College, New York, New York.

The authors thank Jerome B. Posner, M.D., for his comments and Peter A. Forsyth, M.D., for his help with early drafts of the man- uscript.

Address for reprints: Lisa M. DeAngelis, M.D., Department of Neurology, Memorial Hospital, 1275 York Avenue, New York, NY 10021.

Accepted for publication August 13, 1993.

tastases and venous thromboembolism. Cancer 1994; 73:493-a.

Key words: brain metastasis, brain tumor, deep venous thrombosis, pulmonary embolism, venous thromboembolism, anticoagulation, heparin, warfarin, vena cava filter, intracerebral hemorrhage.

Venous thromboembolic disease is a common complication of both occult and established cases of cancer; 15% of patients with cancer experience clinical thromboembolic disease during their illness.' Patients with both primary and metastatic brain tumors appear to be even more prone to this complication, perhaps because many are hemiparetic or The optimal therapy for patients with brain tumors and deep venous thrombosis (DVT) or pulmonary embolism (PE) is unknown. The potentially catastrophic risks of systemic and intracranial hemorrhage lead many clinicians to es- chew anticoagulation and advocate placement of inferior vena cava filters to protect the patient from large PE.6-'o However, the efficacy of filters in patients with brain tumors is unproven, and recent data suggest an unexpectedly high rate of recurrent PE as well as recurrent DVT and filter thrombosis in such patients." Sev- eral retrospective series indicate that patients with primary brain tumors and venous thromboembolic disease can be safely anti~oagulated.'~''-'~ Although some case reports document the occurrence of intratumoral hemorrhage in anticoagulated patients with brain metas- t a s e ~ , ' ~ , ' ~ the actual magnitude of this risk is unknown. We retrospectively reviewed the experience at Memo- rial Sloan-Kettering Cancer Center of patients with brain metastases and concomitant venous thromboembolism to evaluate the efficacy and complications of filters and anticoagulation.

Methods

We used computerized data bases to identify 51 adult patients with brain metastases treated at Memorial

494 CANCER January 15, 1994, Volume 73, No. 2

Table 1. Clinical Features of All Patients With Brain Metastases and Venous Thromboembolism

All Anticoagulated Filter patients patients* patients*

No. of patients Age (Y r)

Median Range

Sex (MJF) Primary tumor

Non-small cell lung Breast Melanoma Melanoma and CML Small cell lung Renal Systemic lymphoma

Patients with DVT alone Patients with PE alone Patients with DVT and PE

51t

58

30/21

25 9 7 1 5 3 1

27 10 14

25-81

42

59 25-67 25/17

22 6 4 1 5 3 1

22 9

11

10

55 48-81

6/4

3 3 3 0 0 1 0 6 2 2

CML chronic myelogenous leukemia; DVT deep venous thrombosis; PE: pulmonary embolus. * Includes three patients who received a filter followed by anticoagulation. t Includes two oatients who received neither a filter nor anticoaeulation.

Sloan-Kettering Cancer Center between January 1, 1980, and July 1, 1992, who experienced DVT or PE. Patients with primary brain tumors or leptomeningeal metastases without parenchymal metastases or whose venous thromboembolism antedated diagnosis of brain metastasis were excluded. Information recorded in- cluded the patient’s age and sex, the primary tumor, evidence of hemorrhage in the brain metastasis before institution of therapy, history of radiation therapy or surgery for the metastasis, history of seizures, chemotherapy, the nature of the initial venous thromboembolic event, the initial therapy for venous thromboembolism, dosage and duration of anticoagulation, coagulation parameters, clinical or laboratory evidence of systemic hemorrhage, neurologic deterioration, results of subsequent brain imaging studies, changes in or discontinuation of therapy for DVT and PE, and date of last follow-up or death. All patients had computed tomography (CT) or magnetic resonance scans document- ing brain metastasis. The diagnosis of venous thromboembolism was based on clinical findings and was confirmed in 46 patients with venogram, pulmonary arteriogram, venous Doppler, or ventilation-perfusion scan and with autopsy in 1 patient; in four cases, confir- matory tests could not be performed. Therapy was at the discretion of the attending physician. When the ex- act date of discontinuation of anticoagulation was not specified, the last date of known anticoagulation was used in analyses. Patients receiving anticoagulation for

a nominal period around the time of Greenfield filter placement were not considered to have received anticoagulation; no such patient experienced hemorrhagic complications. For the purpose of analysis, we considered patients to be anticoagulated as long as they were receiving anticoagulation intended as treatment for established DVT or PE, regardless of the dose and coagulation parameters.

Results

Fifty-one patients were identified with a venous thromboembolic event after diagnosis of brain metastasis (Ta- ble 1). Two received no therapy for DVT, and both died of autopsy-proven PE. One was not anticoagulated because of a large hemorrhagic melanoma brain metastasis; the second patient had a prolonged prothrombin time (PT) from liver disease and was believed to be autoanticoagulated. Ten patients initially received Greenfield filters; Table 2 summarizes the reasons why they were not anticoagulated. Four of these patients were diagnosed with recurrent venous thromboembolism within 1 month of filter placement (Table 3), confirmed in two cases by high-probability ventilation- perfusion scans and in one case with a positive venogram contralateral to the initial DVT. Three of 10 (30%) subsequently received anticoagulation for control of recurrent venous thromboembolism.

The 3 patients who received anticoagulation after filter placement were grouped with the 39 patients initially treated with anticoagulation, to yield 42 patients with evaluable courses of anticoagulation. Although nu- merous anticoagulation regimens were used (Table 4) because of physician preference and intercurrentinedi- cal needs, most patients received traditional courses of heparin followed by warfarin. Target partial thromboplastin times (PTT) were generally 2-2.5 times control, and target PT were 1.5-2.0 times control. International normalized ratios were not reported over this period.

Table 2. Reasons Patients Received Filters Rather Than Anticoagulation

Reason No. of uatients

Hemorrhagic brain metastases on computed tomography scan or magnetic resonance imaging before anticoagulation

4 (3 melanoma, 1 non-small cell lung)

Pericraniotomy 3 Thrombocytopenia 1

for tissue diagnosis 1

Falling hematocrit level and need for bronchoscopy

Age (81 yr) and generally Door condition 1

Venous Thromboembolism/Schiff and DeAngelis 495

Table 3 . Patients With Complications of Greenfield Filters

Age (vr) Sex Tumor Reason for filter

52 F NSCLC DVT, hemorrhagic brain metastasis

5 3 M Renal DVT 2 days after craniotomy for resection of metastasis

6 3 M NSCLC PE, required bronchoscopy for

50 M Melanoma PE, metastasis hemorrhagic by MRI tissue diagnosis

thouah not bv CT

~~ ~

Time after filter Subsequent Complication placement (d) anticoagulation

Bilateral leg swelling, probable 18 No contralateral DVT (clinical diagnosis)

phlegmasia cerulea dolens (positive venogram)

ventilation-perfusion scan)

ventilation-perfusion scan)

Contralateral DVT with 21 Yes

Recurrent PE (high-probability 5 Yes

Recurrent PE (high-probability 31 Yes

DVT: deep venous thrombosis; NSCLC: non-small cell lung cancer; PE: pulmonary embolus; CT: computed tomography scan; MRI: magnetic resonance imaging.

Subcutaneous heparin regimens were particularly vari- able; some patients received as little as 10,000-15,000 units of heparin daily, whereas others received up to 60,000 units each day. The 42 patients were anticoagulated for a total of 421 1 days (mean, 100 days; median, 88 days; range, 5-563 days) or a total of 138.4 patient- months of anticoagulation. Two patients received two courses of anticoagulation. Twenty-one patients (50%) completed their planned course or courses of anticoagulation without interruption. Nine patients (21%) died of causes apparently unrelated to thromboembolic disease or anticoagulation, and two patients (5%) were lost to follow-up while receiving anticoagulation. Ten patients (24%) had courses of anticoagulation stopped: 9 (21%) because of hemorrhage and 1 (2%) for a surgical procedure.

Five patients (12%) were diagnosed as having recurrent nonfatal episodes of venous thromboembolism while receiving anticoagulation. One patient receiving intravenous heparin for a DVT had PTT consistently less than or equal to 1.5 times control and experienced a PE 4 days into therapy. Three patients receiving warfarin had recurrent DVT and one patient had both DVT and PE; one had a normal PT and three had low therapeutic PT of 15.1-17.7 seconds (1.2-1.4 times control)

Table 4. Anticoagulation Regimens in 42 Patients

Regimen No. of uatients

Intravenous heparin followed by warfarin Intravenous heparin Warfarin alone Intravenous heparin and subcutaneous heparin Intravenous heparin, subcutaneous heparin,

Duration of anticoagulation (d) and warfarin

Median Range

29 2 2 7

2

88 5-563

at the time of recurrence. The two patients with PE had high-probability ventilation-perfusion scans; the diagnosis of recurrent DVT was based on recurrence of leg pain and swelling after these symptoms had initially resolved after receiving anticoagulants. Four of these five patients had no further thromboembolic events after their anticoagulation regimen was altered; anticoagulation was discontinued in the fifth patient because of general medical deterioration. Two other patients had recurrent thromboembolic disease after anticoagulation had been discontinued, and each received a second course of anticoagulation.

Three of the patients receiving anticoagulation (7%) experienced serious or devastating cerebral hemor- rhages (Table 5). A DVT developed in one woman with a nonhemorrhagic frontal metastasis from breast cancer treated with radiation therapy. She was begun on intravenous heparin, but her dosage was excessive, resulting in a PTT of 93 seconds after 5 days (her PTT on admission was 23 seconds, normal < 37 seconds). The next day she experienced a massive intratumoral hemorrhage and died despite aggressive medical management. A second patient who had received radiation therapy for two left hemisphere metastases from non-small cell lung cancer received warfarin for a DVT. Anticoagu- lation was continued for more than 5 months despite gradual general deterioration and problems with compliance. He had progressive confusion for 2 weeks, but PT was not checked. A global aphasia and right prona- tor drift then developed. PT on admission was more than 60 seconds (more than 4.8 times control) and PTT more than 100 seconds (more than 2.7 times control), and a CT scan documented hemorrhage in the smaller of his two metastases. The patient improved slightly with vitamin K, fresh-frozen plasma, and dexametha- sone but remained confused and lethargic until his death 3 weeks later. The third patient had multiple brain metastases from non-small cell lung cancer treated with radiation therapy, and a DVT developed.

496 CANCER Juanuary 25, 2994, Volume 73, No. 2

Table 5. Complications of Anticoagulation in 42 Patients

Therapeutic Supratherapeutic coagulation coagulation parameter* parameterst Total

Complication (%) W O ) (%)

Systemic bleeding Major Minor

Guaiac-positive stool with stable hematocrit level

Gastrointestinal bleeding (present before anticoagulation)

Thigh hematoma Hernoptysis

Brain hemorrhage

0 0 0

Symptomatic 1 ( 2 ) 2 (4) 3 (7) Asymptomatic 3 (7) 0 3 (7)

PT: prothrombin time; PTT: partial thromboplastin time. * Defined as PT 5 2 times control and PTT c 2.5 times control. t Defined as PT > 2 times control or PTT > 2.5 times control. $ Required blood transfusion.

She received 9 days of intravenous heparin (28,800 units per day) and was then switched to subcutaneous heparin (20,000 units every 8 hours). Eight days later, increased dysarthria and gait ataxia developed. PT was 14.6 seconds (1.2 times control) and PTT was 65.1 seconds (1.7 times control), and a CT scan showed a small hemorrhage in the third ventricle. Anticoagulation was stopped, and her neurologic status returned to baseline. Thus, in one patient, anticoagulation was directly re- sponsible for death, in another it was a likely contribut- ing factor, and in a third case it led to potentially serious but reversible neurologic dysfunction. Seizures and thrombocytopenia did not contribute to any of these episodes, and no patient suffered a subdural hematoma.

Three additional patients receiving anticoagulants (7%), one each with brain metastases from melanoma, breast carcinoma, and non-small cell lung carcinoma, were asymptomatic but had routine follow-up CT scans revealing new hyperdensity of metastases precontrast, suggesting intratumoral blood. In each case, anticoagulation was discontinued, although warfarin was later reinstituted uneventfully in the patient with melanoma after a recurrent PE.

Systemic bleeding potentially related to or exacer- bated by anticoagulation occurred in 8 of 42 patients (190/,) but was generally minor (Table 5). Four patients had guaiac-positive stools not requiring transfusion, and two patients had mild hemoptysis (both had pulmonary lesions and one had a PT of 50 seconds). One patient had gastrointestinal bleeding before anticoagulation and continued to bleed with anticoagulation, thus requiring transfusion. The final patient, in the set-

ting of a PT of more than 40 seconds, hemorrhaged in his thigh and required transfusion.

Discussion

The optimal treatment of venous thromboembolism in patients with brain tumors is unknown. Primary and secondary brain tumors should be considered sepa- rately, because they differ considerably in the underly- ing propensity of the tumor to bleed”,” and in the ap- proach to diagnosis and treatment. For example, patients with primary brain tumors are more likely to undergo a neurosurgical procedure that may prohibit use of anticoagulants in the immediate postoperative period, whereas patients with metastatic brain tumors usually have sites of systemic disease that are vulnera- ble to hemorrhagic complications of anticoagulation. Al- though case reports document the occurrence of hemorrhage into brain metastases in patients receiving anticoagula t i~n , ’~ . ’~ no study has addressed the frequency with which this occurs. Even without anticoagulation, spontaneous hemorrhage into a brain metastasis is not rare; macroscopic hemorrhage is found in 2.9-9.4% of all brain metastases on pathologic specimen^.'^-'^ The risk varies with the histology of the primary tumor: 40% of melanomas and 70% of renal carcinomas metastatic to brain show evidence of intratumoral hemorrhage on CT scan,” and virtually 100°/~ of brain metastases from choriocarcinoma have frank hem~rrhage.’~ Cancer patients without central nervous system disease receiving anticoagulation have an increased risk of hemorrhagic complications compared with the general population, although most complications occur in patients

Venous Thromboembolism/Schifi and DeAngelis 497

who receive excessive anticoagulation by conventional laboratory standard^.'^

The potential risks of anticoagulation and the ready availability of inferior vena cava (IVC) filters have led some authorities to recommend filters in all patients with brain tumors and venous t h r o m b ~ e m b o l i s m . ~ ~ ~ ~ However, as a recent editorial26 indicates, there are no data comparing the safety and efficacy of filters versus anticoagulation in any setting. Infrequently reported complications include filter tilting or malpositioning, hematoma and cellulitis at the insertion site, air embolism, pneumothorax, filter migration, and erosion of filter struts through the wall of the IVC.27-30 None of our patients experienced these complications, but only 10 received IVC filters. A recent prospective study found that in 35% of patients receiving IVC filters through the femoral vein puncture, ultrasound-proven thrombosis developed at the puncture site.31 Further- more, a retrospective study of 42 patients with a mix- ture of primary and metastatic brain tumors receiving filters found that 12% experienced recurrent PE, whereas 57% experienced either recurrent DVT, IVC/ filter thrombosis, or postphlebitic syndrome." Thus, patients with brain tumors treated with IVC filters, whether because of hemiparesis, immobilization, craniotomy, or other biologic factors, appear to be particularly prone to progressive venous thromboembolism despite the presence of a filter.

In our study, anticoagulation effectively prevented recurrent thromboembolism. Only five (1 2%) patients who received anticoagulation experienced nonfatal recurrent DVT or PE when coagulation values were in the low therapeutic or normal range. In contrast, 4 of 10 patients (40%) receiving filters experienced recurrent thromboembolic disease, and 3 of them ultimately received anticoagulation. Serious central nervous system complications occurred in 7% of anticoagulated patients with brain metastases, but two of the three patients with significant complications had supratherapeutic PT or PTT, and the third was on a strikingly high dose of subcutaneous heparin despite a therapeutic PTT at the time of complication. Although evidence of systemic bleeding in anticoagulated patients occurred in 19% of these patients, it was generally minor and infrequently required transfusion or intervention. Be- cause a quarter of our population died outside the hospital or were lost to follow-up while receiving anticoagulants, we cannot be certain that the actual incidence of central nervous system or systemic hemorrhagic complications was not higher.

Our findings suggest that anticoagulation is more effective than IVC filters in patients with brain metastases and venous thromboembolism. Furthermore, anticoagulation is safer; IVC filters have a higher incidence

of complications. Although there are some circum- stances that may necessitate placement of an IVC filter (e.g., before craniotomy, poor compliance for any reason, inability to receive appropriate monitoring, fre- quent falls, and possibly anticipated prolonged thrombocytopenia from chemotherapy), anticoagulation is in general the preferred therapy for thromboembolic disease even in patients with brain metastases. The safety of anticoagulation can be enhanced with close monitoring of PT and PTT, particularly with the use of the international normalized ratios.32 The international normalized ratios not only control for variation in throm- boplastins between laboratories, but also have led to the awareness that with reagents commonly used in the United States, a PT of 1.3-1.5 is adequate to control most cases of venous thromb~embol i sm.~~ This lower degree of anticoagulation is likely to reduce the risk of clinically significant hemorrhage further.

Other important issues remain unresolved. The tendency of certain metastatic tumors such as melanoma, choriocarcinoma, renal cell carcinoma, and thy- roid carcinoma, to hemorrhage spontaneously may place such patients at increased risk from anticoagulation, although our data do not suggest this. Whether asymptomatic hemorrhage into a metastasis presages symptomatic hemorrhage and therefore mandates discontinuation of anticoagulation is unknown, although the need for continued therapy should be reassessed. How soon patients recovering from craniotomy can receive anticoagulation safely is similarly unknown. The role of low-molecular-weight heparins, which may carry a decreased risk of hemorrhagic complications for a given degree of antic~agulation,~~ has not been stud- ied in patients with brain tumors but may offer a safer therapeutic alternative to the currently available options. Certainly, both anticoagulation and IVC filters are imperfect remedies to the problem of venous thromboembolism in the patient with brain metastases. Until a randomized trial addresses the risks and benefits of filters compared with anticoagulation in patients with nonhemorrhagic brain metastases, we recommend that such patients receive carefully monitored anticoagulation as their initial therapy.

References

Luzzatto G, Schafer AI. The prethrombotic state in cancer. Se- min Oncol 1990; 17:147-59. Ruff RL, Posner JB. Incidence and treatment of peripheral venous thrombosis in patients with glioma. Anti Neuvol 1983; 13:334-6. Cheruku R, Tapazoglou E, Ensley 1, Kish JA, Cummings GD, Sarraf M. The incidence and significance of thromboembolic complications in patients with high-grade gliomas. Caricev 1991; 68:2621-4.

498 CANCER Iunuury 25, 2994, Volume 73 , No. 2

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

Constantini S, Kornowski R, Pomeranz S, Rappaport ZH. Thromboembolic phenomena in neurosurgical patients oper- ated upon for primary and metastatic brain tumors. Actn Neuro- rh i r (Wieti) 1991; 109:93-7. Sawaya R, Zuccarello M, Elkalliny M, Nishiyama H. Postopera- tive venous thromboembolism and brain tumors: part I. Clinical profile. ] Neurooricol 1992; 14:119-25. Moore FD, Osteen RT, Karp DD, Steele G, Wilson RE. Anticoagu- lants, venous thromboembolism, and the cancer patient. Arch Surg 1981; 116:405-7. Swann KW, Black PM, Baker MF. Management of symptomatic deep venous thrombosis and pulmonary embolism on a neurosurgical service. I Neurosurg 1986; 64:563-7. Levine M, Hirsh J , The diagnosis and treatment of thrombosis in the cancer patient. Semin Oncol 1990; 17:160-71. Cohen JR, Tenenbaum N, Citron M. Greenfield filter as primary therapy for deep venous thrombosis and/or pulmonary embolism in patients with cancer. Surgery 1991; 109:12-5. Cohen J R , Grella L, Citron M. Greenfield filter instead of heparin as primary therapy for deep venous thrombosis or pulmonary embolism in patients with cancer. Cancer 1992; 70:1993-6. Levin JM, Schiff D, Loeffler JS, Fine HA, Black PM, Wen PY. Complications of therapy for venous thromboembolic disease in patients with brain tumors. Neurology 1993; 43:1111-4. Choucair AK, Silver P, Levin VA. Risk of intracranial hemorrhage in glioma patients receiving anticoagulant therapy for venous thromboembolism. ] Neurosurg 1987; 66:357-8. Olin JW, Young JR, Graor RA, Ruschhaupt WF, Beven EG, Bay JW. Treatment of deep venous thrombosis and pulmonary em- boli in patients with primary and metastatic brain tumors. Arch lrit Med 1987; 147:2177-9. Altschuler E, Moosa H, Selker RG, Vertosick FT. The risk and efficacy of anticoagulant therapy in the treatment of thromboembolic complications in patients with primary malignant brain tumors. Neurosurg 1990; 27~74-7. So W, Hugenholtz H, Richard MT. Complications of anticoagulant therapy in patients with known central nervous system lesions. Can I S u r g 1983; 26:181-3. Wolfe MW. Intracranial hemorrhage during systemic anticoagulation in patients with cerebral metastases [letter]. Arch I i i l Med 1988; 48: 1878-9. Mandybur TI. Intracranial hemorrhage caused by metastatic tumors. Neurology 1977; 27:650-5.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

Scott M. Spontaneous intracerebral hematoma caused by cerebral neoplasms. Neurosurg 1975; 42:338-41. Globus JH, Sapirstein M. Massive hemorrhage into brain tumor: its significance and probable relationship to rapidly fatal termi- nation and antecedent trauma. ]AMA 1942; 120:348-52. Kondziolka D, Bernstein M, Resch L, Tator CH, Fleming JFR, Vanderlinden RG, et al. Significance of hemorrhage into brain tumors: clinicopathologic study. Wakai S, Yamakawa K, Manaka S , Takakura K. Spontaneous intracranial hemorrhage caused by brain tumor: its incidence and clinical significance. Neurosurg 1982; 10:437-44. Weisberg LA. Hemorrhagic metastatic intracranial neoplasms: clinical-computed tomographic correlations. Conzpuf Radiol 1985; 9:105-14. Vaughan HG, Howard RG. lntracranial hemorrhage due to metastatic chorioepithelioma. Neurology 1962; 12:771-7. Krauth D, Holden A, Knapic N, Liepman M, Ansell J. Safety and efficacy of long-term oral anticoagulation in cancer patients. Caiicer 1987; 59:983-5. Muchmore JH, Dunlap JN, Culicchia F, Kerstein MD. Deep vein thrombophlebitis and pulmonary embolism in patients with malignant glioma. South M e d ] 1989; 82:1352-5. Dorfman GS. Evaluating the roles and functions of vena caval filters: will data be available before or after these devices are removed from the market? Radiology 1992; 185:15-7. Carabasi RA, Moritz MJ, Jarrell BE. Complications encountered with the use of the Greenfield filter. A m JSurg 1987; 154:163-8. Kanter B, Moser KM. The Greenfield vena cava filter. Chest

Becker DM, Philbrick JT, Selby JB. Inferior vena cava filters: indications, safety, effectiveness. Arch Znt Med 1992; 152:1985- 93. Grassi CJ, Goldhaber SZ. Interruption of the inferior vena cava for prevention of pulmonary embolism: transvenous filter devices. Herz 1989; 14:182-91. Molgaard CP, Yucel EK, Geller SC, Knox TA, Waltman AC. Access-site thrombosis after placement of inferior vena cava filters with 12-14 F delivery sheaths. Radiology 1992; 185:257- 61. Hirsh J , Oral anticoagulant drugs. N EtzglIMed 1991; 324:1865- 75. Hirsh J, Levine MN. Low molecular weight heparin. Blood 1992; 7911-17.

Neurosurg 1987; 67:852-7.

1988; 93:170-5.

Documents

Therapy of venous thromboembolism in patients with brain metastases