SL Bridgers MD

Department of Cardiology, Department of Veterans Affairs Medical Center, Clinical Campus, Yale University School of Medicine

Neurology

Principal Investigator

MD Ezekowitz MD, PhD

Department of Cardiology, Department of Veterans Affairs Medical Center, Clinical Campus, Yale University School of Medicine

Cardiology

Principal Investigator

KE James PhD

Department of Health Services Research, VA Medical Center

Biostatistics

Statistician

KE James PhD

Department of Health Services Research
VA Medical Center
3710 SW US Veterans Hospital Rd.
Portland, OR
USA 97201

Department of Veterans Affairs

VA

n.r.

n.r.

Yes, all publications are reviewed by the Executive Committee

n.r.

n.r.

R Shabetai MD, FR Rickles MD, SM Nazarian MD, CL Colling PHARMD, MS, JF Kurtzke MD, CC Gornick MD, MD Ezekowitz MD, PhD, KE James PhD, SL Bridgers MD

Neurologist, cardiologists, statisticians, etc.

Guides the study and has access to the talents of appropriate consultants. This committee is responsible for recommending changes to the protocol and for monitoring the performance of each medical center. Finally, it controls all data releases; all publications arising from this study must be cleared through this committee, which will meet at twelve-month intervals.

Yes

n.r.

SM Nazarian MD

SM Nazarian MD, SR Gupta MD, SL Bridgers MD

n.r.

Copies of all cranial CT scans obtained in association with cerebral endpoints are classified by a neuroradiologist into 1) cerebral infarction or no infarction, 2) hemorrhagic infarction, or 3) cerebral hemorrhage. The neuroradiologist's decision will determine the ultimate classification of a specific endpoint as cerebral infarction (categories 1 and 2) or cerebral hemorrhage (category 3).

n.r.

Blinded to treatment received

n.r.

n.r.

n.r.

Boston VAMC

Boston, MA

E Ascher MD

n.r.

VA

Long Beach VAMC

Long Beach, CA

A Alzarka MD

n.r.

VA

Newington VAMC

Newington, CT

MJ Radford MD

n.r.

VA

San Diego VAMC

San Diego, CA

R Shabetai MD

n.r.

VA

Tampa VAMC

Tampa, FL

S Zachariah MD

n.r.

VA

West Haven VAMC

West Haven, CT

E Winter MD

n.r.

VA

Baltimore VAMC

Baltimore, MD

NH Carliner MD

n.r.

VA

Washington, DC VAMC

Washington, DC

P Carson MD

n.r.

VA

Seattle VAMC

Seattle, WA

NH Carliner MD

n.r.

VA

Baltimore VAMC

Baltimore, MD

J Stratton MD

n.r.

VA

Roseburg VAMC

Roseburg, OR

C Carter MD

n.r.

VA

Northport VAMC

Northport, NY

G Mallis MD

n.r.

VA

Minneapolis VAMC

Minneapolis, MN

CC Gornick MD

n.r.

VA

Little Rock VAMC

Little Rock, AK

SM Nazarian MD

n.r.

VA

Cincinnati VAMC

Cincinnati, OH

L Wexler MD

n.r.

VA

Hines VAMC

Hines, IL

SR Gupta MD

n.r.

VA

Palo Alto VAMC

Palo Alto, CA

E Atwood MD

n.r.

VA

Durham VAMC

Durham, NC

n.r.

n.r.

VA

Ann Arbor VAMC

Ann Arbor, MI

n.r.

n.r.

VA

Oklahoma City VAMC

Oklahoma City, OK

n.r.

n.r.

VA

Cerebral Infarction

Clinically evident focal neurologic deficit, not attributable to dysfunction of a single cranial nerve, the spinal cord or the peripheral nervous system, with persistence of the deficit, or some portion thereof, for longer than 24 hours, with absence of intracerebral hemorrhage, other than hemorrhagic infarction, on the first cranial CT scan obtained following onset of the deficit, and with absence of evidence of intracranial tumor or infection in association with the appearance of the neurologic deficit. "Stroke" is considered synonymous with clinically evident cerebral infarction.

5% per year

Based on Sherman, et al. Chest, 1986; 89(Suppl 2):82S-98S, and others.

2.5% per year

Powered to detect a 50% reduction in the incidence of stroke in the three-year follow-up of patients treated with oral anticoagulation as compared to placebo patients.

80%

n.r.

0.05

2-tailed

556

As described in the official trial protocol, and referenced in T. Colton, Statistics in Medicine, Little, Brown and Company, Boston, 1974, p. 169.

750

Assuming that 25% of randomized patients may be lost to follow-up

Ordered

Rank sum

n.r.

Differences in severity, of cerebral endpoints and of non-cerebral hemorrhage between the treatment and control groups are tested using the Wilcoxon rank sum test.

Continuous

Chi-square

n.r.

The chi-square test will be performed to test for the difference in the frequency of cerebral and other endpoints.

Time to event

Kaplan-Meier

n.r.

A survival function will be estimated for each treatment group by the Kaplan-Meier product limit method, and a log-rank test will be performed to see if there is evidence of treatment difference.

To compare the incidence of cerebral infarction over a three year follow-up period in warfarin- and placebo-treated patients.

Comparison of frequency of mortality, non-cerebral hemorrhage, vascular events other than cerebral infarction, and changes over time in hemoglobin and hematocrit levels, to better understand the extent of complications and efficacy of low intensity oral anticoagulation treatment.

Silent Cerebral Infarction

Patients will be classified according to the presence or absence of subclinical infarction based on CT scan, and the relaitve risk of cerebral infarction or cerebral hemorrhage occurrence will be computed for placebo and treated groups, and compared by chi square test.

Over 70

Those greater than 70 years of age.

M Dunn MD

M Dunn MD, WM O'Fallon, DA Triplett, LR Caplan

n.r.

This committee monitors the study as an outside group and periodically files a report giving its view of the progress of the study. The Committee also approves protocol modifications and continuation of the study. The Committee will meet at 12-month intervals.

See information on stopping rule

Not blinded to treatment received

n.r.

n.r.

Also called the Operations Committee

1

Jan 11, 1988

Changed exclusion criterion for atrial fibrillation secondary to hyperthyroidism:

1. If atrial fibrillation first was detected in the setting of hyperthyroidism but the patient is now euthyroid and is no longer a cardioversion candidate, then this will not be an exclusion.
2. Atrial fibrillation secondary to hyperthyroidism.

2

Apr 24, 1987

Changed exclusion criterion for hospitalization with acute or unstable cardiac disease to:

1. Current hospitalization for cardiac disease other than atrial fibrillation.
2. Acute or unstable cardiac disease.

3a

Apr 24, 1987

Changed criterion for patients on long term anticoagulation -- Arm I

1. Current long-term (greater than 1 month) oral anticoagulation therapy does not disqualify patients. Patients on oral anticoagulant therapy for 1 month or less must be off therapy for 1 month at time of randomization. Patients on oral anticoagulant therapy greater than 1 month must be off therapy for 6 months at time of randomization.

3b

May 12, 1987

Changed criterion for patients receiving anticoagulation -- Arm II

1. Patients may have been receiving continuous oral anticoagulant therapy at time of screening. Those receiving anticoagulation therapy for 6 months or less must be off therapy for one month prior to randomization. If therapy has lasted greater than six months, patients must be off therapy for at least 6 months at time of randomization.

3c

May 12, 1987

..., in order to avoid restricting patient intake, we have allowed for oral anticoagulation for an unspecified period prior to Arm I and II entry, subject to restrictions in 3a and b above.

4

Mar 6, 1989

Changed criterion for patients with prior TIA:

Patients with no remote event of a TIA within the past five years will not be excluded.

5a

May 12, 1987

Additions to entry evaluation:

*Cardiac doppler* and *carotid doppler ultrasound* and noncontrast cranial computed tomography.

5b

May 12, 1987

Functional assessments added to entry criteria:

Baseline functional status...

5c

May 12, 1987

Cardiac doppler added.

1+= detection of regurgitant stream less than 25% of the distance to the posterior atrial wall.

3+= detection of regurgitant stream more than 50% of the distance to the posterior atrial wall, but short of the wall itself.

CT scan required for Arm I and II patients.

1. Cranial computed tomography: An entry non-contrast CT scan will be required for all patients in Arm I and Arm II. In Arm I patients, the presence of clinically silent cerebral infarction evidence on CT scan will be reported. For Arm II patients, CT obtained in connection with the last clinically evident cerebral infarction may be used if this CT revealed an infarction appropriate to the clinical presentation. Copies of all CT scans will be forwarded to the Chairman's Office.
   4+= regurgitant stream extending to the posterior atrial wall.
   2+= detection of regurgitant stream 25-50% of the distance to the posterior atrial wall.
2. Cardiac doppler: Assessment based on the mitral regurgitant stream detected from the apical view using the following grading scale:

6

May 12, 1987

Change in treatment of patients who convert to sinus rhythm:

Patients who convert spontaneously to sinus rhythm or any heart rhythm will remain on blinded treatment with the study drug.

7

May 12, 1987

Change definition of major hemorrhage. Delete definitions of minor and occult hemorrhage.

8

May 12, 1987

Study drug may be reinstated after temporary treatment withdrawal.

9

Apr 27, 1987

Completion of FDA Form 1639:

The Adverse Drug Reaction Form will not be completed for hemorrhagic events. However, FDA Form 1639 must be completed for all major hemorrhages and all end point cerebral hemorrhages, as well as cutaneous necrosis and allergic reactions.

Reporting of adverse drug reactions.

10

May 12, 1987

Added procedure for defining cortical cerebral infarction:

11

May 12, 1987

Grading of cerebral endpoints deleted.

12

May 12, 1987

Ocular ischemic events added as intercurrent events.

13a

May 12, 1987

Clarify information given to patients at completion of study.

Patients will be informed of their treatment at the end of the 3-year period by the unblinded physician. Members of the Blinded Team will not be informed of treatment in specific patients until completion of the study.

13b

May 12, 1987

Patients completing the study who request unblinding at the time of their completion may be informed of the treatment assignment by the Unblinded Co-Investigator.

Normal prothrombin time

Prothrombin time within the normal range of the participating center

AND

Male veterans

Male veterans of any age, able to give informed consent

AND

Atrial fibrillation

Atrial fibrillation documented on two ECG's at least four weeks apart, with atrial fibrillation present at study entry. Atrial fibrillation is defined as the absence of p-waves or any other continuous rhythmic atrial activity on ECG, including saw-tooth waves if occurring at a regular rate.

Valvular cause of atrial fibrillation

Clinical evidence of mitral stenosis or mitral regurgitation with echocardiographic evidence of rheumatic valvular disease. Atrial fibrillation in association with mitral valve prolapse detected by echocardiography will not be considered valvular atrial fibrillation for this study.

OR

Male veterans

Male veterans of any age, able to give informed consent

OR

Atrial fibrillation due to hyperthyroidism

Atrial fibrillation secondary to hyperthyroidism

If atrial fibrillation first was detected in the setting of hyperthyroidism but the patient is now euthyroid and is no longer a cardioversion candidate, then this will not be an exclusion.

OR

Planned conversion to sinus rhythm

Pending institution of quinidine, procainamide or related anti-arrhythmic agent

OR

Pending electrical cardioversion attempt

OR

Recent sinus rhythm

Sinus rhythm on ECG in the month prior to entry

OR

Valve prosthesis

Mechanical or biological cardiac valve prosthesis

OR

Intracardiac thrombus

Evidence of intracardiac thrombus by echocardiography

OR

Atrial tumor

Evidence of atrial tumor by echocardiography

OR

History of bacterial endocarditis

History of bacterial endocarditis within one year of entry

OR

Acute or unstable cardiac disease

Myocardial infarction within one month of entry

OR

Current hospitalization for cardiac disease other than atrial fibrillation

OR

Pending procedure

Pending surgical or invasive radiographic procedure

OR

Low life expectancy

Any disorder making survival for one year unlikely

OR

Hemostasis disorder

Any existing disorder of hemostasis

OR

History of GI bleeding

History of gastrointestinal hemorrhage within two years of screening

OR

History of GI ulcer

History of documented gastric or duodenal ulcer within two years of screening

OR

Esophageal varices

Known esophageal varices

OR

Alcoholism

Known chronic alcoholism, or history of treatment in an inpatient alcohol rehabilitation program within two years of screening. Active participation in Alcoholics Anonymous or their support group does not exclude.

OR

Hypertension

Arterial hypertension at screening with systolic blood pressure greater than 180 mmHg, or diastolic blood pressure greater than 105 mmHg.

OR

Contraindication to anticoagulation

Dipyridamole and sulfinpyrazone lack clear-cut support as effective anti-thrombotic agents, and may be withdrawn for study entry. If this is unacceptable to a particular patient or his responsible physician, the patient will be excluded. Likewise, patients who require, or prefer to continue, ongoing chronic therapy with nonsteroidal anti-inflammatory agents other than aspirin would be excluded.

Any condition that, in the opinion of participating physicians, contraindicates oral anticoagulation.

OR

Very recent anticoagulation

Patients on oral anticoagulant therapy for 1 month or less must be off therapy for 1 month at time of randomization.

Current long-term (greater than 1 month) oral anticoagulant therapy does not disqualify patients.

AND

Patients on oral anticoagulant therapy greater than 1 month must be off therapy for 6 months at time of randomization

OR

Recurrent venous thromboembolism

Recurrent venous thromboembolism before age 50.

OR

Required anticoagulation

Active venous trhomboembolic disease

OR

Any other condition that, in the opinion of an attending physician, mandates oral anticoagulation

OR

Inherited or acquired hypercoagulable state

OR

Aspirin therapy

Current aspirin therapy for indications supported by randomized prospective studies (transient ischemic attack, unstable angina within tow years, coronary artery bypass graft within two years). Patients receiving aspirin for other reasons may have their aspirin withdrawn in order for them to participate in this study, provided both the patient and the attending physician responsible for initiation of maintenance of aspirin therapy agree.

OR

History of transient ischemic attack

Patients with no remote event of a TIA within the past five years will not be excluded.

Monocular visual loss, lasting less than 24 hours.

OR

One of more episodes of clearly defined focal neurologic deficit not attributable to the spinal cord, peripheral nervous system.

OR

History of prior cerebral infarction

Clearly documented focal neurologic deficit, not attributable to involvement of the spinal cord, peripheral nervous system or a single cranial nerve, and not attributable to intracranial pathology other than infarction, of greater than 24 hours duration, with persisting focal neurologic deficit evident on physical examination or CT scan evidence of cerebral infarction in a locus consistent with the reported deficit.

OR

History of systemic embolus

History of clinically evident systemic embolus

OR

History of intracranial bleeding

History of prior intracranial bleeding other than traumatic

OR

Unlikely compliance and follow-up

Any social or psychological condition or combination of conditions rendering an individual unlikely to adhere to the required regimen of medication and follow-up evaluation.

OR

Lab abnormalities

Hematocrit less than 32%

OR

Platelet count less than 100,000

OR

Partial thromboplastin time outside the normal range

OR

Total bilirubin outside the normal range

OR

SGOT, SGPT or alkaline phosphatase twice the upper limit of the normal range

OR

Serum T4 above the normal range

OR

More than 5 red blood cells per high powered field on microscopic examination of the urine sediment

OR

Positive stool test for blood

Note: "normal range" means the normal range for the clinical laboratory where the test is performed.

Randomized

Subject/Patient

Generated by Coordinating Center

The Palo Alto Cooperative Studies Program Coordinating Center provided each study site with a drug randomization lists for each study site before the commencement of patient entry

Blinded, or masked

Drug packages are labeled with the randomization code by a central pharmacy. Neither the blinded study team nor the patient will be informed of the treatment until the conclusion of the study. However, the unblinded team will be informed in order to make appropriate dose adjustments.

Randomization is blocked by study center.

No delays

N

260

265

525

n.r.

Age

67.33 +/- 7

67.35 +/- 7

n.r.

0.975

Sex

Males

260 (100.%)

265 (100.%)

525 (100.%)

n.r.

Females

0 (0.%)

0 (0.%)

0 (0.%)

n.r.

Active angina

57 (21.9%)

60 (22.6%)

n.r.

0.843

Prior myocardial infarction

45 (17.3%)

55 (20.8%)

n.r.

0.315

260

265

525

Patient

Didn't know

144 (55.4%)

207 (78.1%)

351 (66.9%)

Guessed correctly

74 (28.5%)

42 (15.8%)

116 (22.1%)

Guessed incorrectly

42 (16.2%)

16 (6.04%)

58 (11.0%)

Care Provider

Didn't know

203 (78.1%)

228 (86.0%)

431 (82.1%)

Guessed correctly

47 (18.1%)

26 (9.81%)

73 (13.9%)

Guessed incorrectly

10 (3.85%)

11 (4.15%)

21 (4.0%)

1

Induction

2 mg

Other

PO

2 mg tab

n.r.

n.r.

2

Maintenance

2 mg

Other

PO

2 mg tab

n.r.

n.r.

1

Induction

4 mg

QD

PO

2 mg tab

n.r.

Prothrombin time will be determined weekly, and dose adjustment performed as per protocol. It is presumed that patients who are compliant will have reached a stable dose by the end of the induction period.

2

Maintenance

customized n.r.

Other

PO

2 mg tab

n.r.

To achieve target prothrombin time of 1.2 to 1.5 times control. Patients will report monthly for prothrombin time determination. In the event of a prothrombin time result outside the therapeutic range, patients will re-enter the dose alteration schedule per protocol.

Clinically evident focal neurologic deficit, not attributable to dysfunction of a single cranial nerve, the spinal cord or the peripheral nervous system, with persistence of the deficit, or some portion thereof, for longer than 24 hours, with absence of intracerebral hemorrhage, other than hemorrhagic infarction, on the first cranial CT scan obtained following onset of the deficit, and with absence of evidence of intracranial tumor or infection in association with the appearance of the neurologic deficit. "Stroke" is considered synonymous with clinically evident cerebral infarction.

Statistical Analysis

Analysis Method

chi-square

Program Name

n.r.

Censoring

Click for details

Any death not subsequent to either of the other two endpoints (cerebral infarction or cerebral hemorrhage).

Statistical Analysis

Analysis Method

chi-square

Program Name

n.r.

Censoring

Click for details

Defined by any of the following: the patient was admitted to an intensive care unit; the patient received a transfusion of whole blood or packed cells; the patient underwent an emergency surgical procedure or intensive medical procedure to terminate bleeding or remove a hematoma.

Statistical Analysis

Analysis Method

chi-square

Program Name

n.r.

Censoring

Click for details

Episodes when a patient presents for acute evaluation of bleeding but is not hospitalized, or episodes which are clinically apparent and subsequently result in elective hospitalization.

Statistical Analysis

Analysis Method

chi-square

Program Name

n.r.

Censoring

Click for details

Assessment Details

Blinded to treatment?

Yes

Blinded to results?

n.r.

Method

Copies of all cranial CT scans obtained in association with cerebral endpoints are classified by a neuroradiologist into 1) cerebral infarction or no infarction, 2) hemorrhagic infarction, or 3) cerebral hemorrhage. The neuroradiologist's decision will determine the ultimate classification of a specific endpoint as cerebral infarction (categories 1 and 2) or cerebral hemorrhage (category 3).

Validity of Method

n.r.

Reproducibility of Method

n.r.

Assessed by

CT Scan Committee

When

End of study

In Subgroups

Placebo

Warfarin

Comment

Each study center's neurologist is blinded to the patient's assigned treatment.

Assessment Details

Blinded to treatment?

Yes

Blinded to results?

n.r.

Method

Autopsy evidence of cerebral parenchymal hemorrhage will be acceptable in lieu of CT scan evidence, in the event death occurs following a cerebral event before a CT scan is obtained.

Validity of Method

n.r.

Reproducibility of Method

n.r.

Assessed by

Determination of the occurrence and classification of cerebral endpoints will be the responsibility of the study neurologist at each study center

When

End of study

In Subgroups

Placebo

Warfarin

Comment

n.r.

Medical History

Yes

Only baseline medical history taken.

Neurological Exam

Yes

At baseline, and at 36 months.

Cranial CT scan

Yes

At baseline only.

Stool test

Yes

At baseline, and every three months until 36 months.

Tablet count

No

Every week from 1 to 12 weeks, then every month from month 3 to month 36.

Treatment Assignment

|

------------------------------------------------------------------------------

|

|

|

|

Timing of Outcomes Assessment

Primary Outcome
Cerebral Infarction
Mean: 1.7 years - Range: n.r.

Secondary Outcomes
Cerebral Hemorrhage
Mean: 1.7 years - Range: n.r.
Death
Mean: 1.7 years - Range: n.r.

Timing of Outcomes Assessment

Primary Outcome
Cerebral Infarction
Mean: 1.8 years - Range: n.r.

Secondary Outcomes
Cerebral Hemorrhage
Mean: 1.8 years - Range: n.r.
Death
Mean: 1.8 years - Range: n.r.

|

|

|

|

Treatment Assignment

|

------------------------------------------------------------------------------

|

|

|

|

82 (31.5%)

56 (21.1%)

N

260

265

525

n.r.

Age

67.33 +/- 7

67.35 +/- 7

n.r.

0.975

Sex

Males

260 (100.%)

265 (100.%)

525 (100.%)

n.r.

Females

0 (0.%)

0 (0.%)

0 (0.%)

n.r.

Active angina

57 (21.9%)

60 (22.6%)

n.r.

0.843

Prior myocardial infarction

45 (17.3%)

55 (20.8%)

n.r.

0.315

Warfarin

260

4 (1.54%)

Placebo

265

19 (7.17%)

Total

525

23 (4.38%)

Relative Risk Reduction

0.79 (0.52 to 0.9)

chi-square

p = 0.001

Warfarin

260

1 (.385%)

Placebo

265

0 ( 0.0%)

Total

525

1 (0.19%)

Relative Risk Reduction

n.r.

chi-square

n.r.

Warfarin

260

15 (5.77%)

Placebo

265

22 ( 8.3%)

Total

525

37 (7.05%)

Relative Risk Reduction

0.31 (-0.29 to 0.63)

chi-square

p = 0.19

6 Months

97.9

95.5 to 100

3

0

142

12 Months

97.2

94.5 to 100

4

0

141

18 Months

95.1

91.6 to 95.1

7

3

135

Prior CVA

1.89

0.23

5.55

0.004

8.72

5.66 to 13.2

Blood Pressure

1.46

0.62

5.55

0.02

4.31

1.28 to 14.52

Major hemorrhage

6 (2.31%)

4 (1.51%)

10 ( 1.9%)

p = 0.54

Minor hemorrhage

64 (24.6%)

46 (17.4%)

110 (21.0%)

p = 0.04

Easy bruising

84 (32.3%)

51 (19.2%)

135 (25.7%)

n.r.

Low-intensity anticoagulation with warfarin prevented cerebral infarction in patients with nonrheumatic atrial fibrillation without producing an excess risk of major hemorrhage. This benefit extended to patients over 70 years of age.

n.r.

On the basis of the results of all completed studies, patients with intermittent or chronic atrial fibrillation should be offered warfarin unless they have lone atrial fibrillation or contraindication to warfarin therapy, in which case aspirin therapy may be considered.

1996

Subgroup or Secondary Results

n.r.

Peer reviewed

Initiating and maintaining warfarin anticoagulation therapy in outpatients with atrial fibrillation: the importance of monitoring. (Circulation 1996; 94(8) Page(s) I-34-n.r.. *Abstract only*

MJ Radford MD, FR Rickles MD, KE James PhD, MD Ezekowitz MD, PhD

1995

Subgroup or Secondary Results

n.r.

Peer reviewed

Silent cerebral infarction in patients with nonrheumatic atrial fibrillation. Circulation 92(8):2178-2182.

SPINAF Investigators

1994

Subgroup or Secondary Results

n.r.

Peer reviewed

Asymptomatic stroke in patients with chronic nonrheumatic atrial fibrillation. JACC Page(s) 441A-n.r.. *Abstract only*

SPINAF Investigators

1992

Subgroup or Secondary Results

n.r.

Peer reviewed

Cerebral infarction rate and effect of warfarin in patients with nonrheumatic atrial fibrillation with or without a previous stroke. Circulation 1992; 86 (4):Suppl 1) Page(s) I-147-n.r.. *Abstract only*

SPINAF Investigators

1992

Main Results

n.r.

Peer reviewed

Warfarin in the prevention of stroke associated with nonrheumatic atrial fibrillation. NEJM 327(20):1406-1412.

SPINAF Investigators

1992

Editorial

Randomized trials of warfarin for atrial fibrillation. NEJM 327(20):1451-1453.

DE Singer MD