Pre-admission warfarin use in patients with acute ischemic stroke and atrial fibrillation: The appropriate use and barriers to oral anticoagulant therapy

Sara L. Partington^a, Simona Abid^a, Koon Teo^a, Wesley Oczkowski^a and Martin J. O'Donnell ^{, a,
a}Divisions of Cardiololgy and Neurology, McMaster University, Hamilton Ontario, Canada
Received 24 August 2006;  revised 18 December 2006;  accepted 22 December 2006.  Available online 16 April 2007.

Warfarin reduces the risk of stroke in patients with atrial fibrillation. Despite strong guideline recommendations, studies continue to demonstrate the under-use of warfarin in clinical practice.

To determine the prevalence and predictors of warfarin use in patients presenting with atrial fibrillation and acute ischemic stroke who do not have a documented contraindication to anticoagulants.

We conducted a retrospective chart review of all patients admitted to the Hamilton General Hospital with a primary diagnosis of ischemic stroke and a coded diagnosis of atrial fibrillation between 1999 and 2004. Using a standardized data abstraction form, the following variables were recorded: baseline demographics, past medical history including risk factors for stroke and major bleeding and known predictors of warfarin under-use. In cases where warfarin was not prescribed, charts were also reviewed for documented contraindications to warfarin use. The following were considered valid contraindications to warfarin: patient refusal, non-compliance with INR monitoring, bleeding diathesis, history of major bleeding or significant alcohol consumption.

In total, 196 patients with ischemic stroke and atrial fibrillation were identified. Of these patients, 106 were considered to be appropriate candidates for anticoagulation after excluding patients with no known diagnosis of atrial fibrillation prior to admission (N = 59), a valid contraindication to warfarin use (N = 18), a CHADS2 score < 1 (N = 6) or a competing diagnosis for warfarin use (N = 7). Of the patients deemed to be suitable candidates for warfarin, 57 (54%) were receiving warfarin therapy on admission. On multivariable analyses, increasing age (OR 0.7; 95% CI 0.5–0.9) was associated with a reduced odds of warfarin use while a history of stroke or TIA (OR 2.6; 95% CI 1.1–6.5) and a history of congestive heart failure (OR 3.2; 95% CI 1.1–9.0) were associated with an increased odds of warfarin use in patients without a contraindication to warfarin. While 75% of patients < 75 years old were anticoagulated, only 33% of those > 85 years were prescribed warfarin on admission to hospital.

early half of all patients presenting with atrial fibrillation and acute ischemic stroke who were suitable candidates for anticoagulation were not prescribed warfarin. In patients not prescribed warfarin, very few had a documented contraindication. Advanced age appears to be the strongest predictor of warfarin non-use.

Figure 1. Proportion of patients receiving warfarin by age-group.

Table 1.
Baseline variables in entire study population

Variable	Participants (n = 196)
Age (mean ± SD) (years)	77.2 ± 10.3
Female	82/196 (42%)
Diagnosis of Afib prior to admission	137/196 (70%)
Warfarin prior to admission	69/196 (35%)
Previous stroke or TIA	73/196 (37%)
Mechanical valve	7/196 (4%)
Hypertension	138/195 (70%)
Diabetes Mellitus	30/196 (15%)
Congestive cardiac failure	41/186 (21%)
Previous gastrointestinal bleed	16/196 (8%)
Previous intracranial bleed	3/196 (2%)
Peptic ulcer disease	10/196 (5%)
Alcohol consumption	13/194 (7%)
Dementia	28/195 (14%)
Falls	16/194 (8%)
Living in Supportive care	19/196 (10%)
CHADS2 (mean ± SD)	2.5 ± 1.4
Bleeding score (mean ± SD)	1.81 ± 0.9

Table 2.
Univariate analysis comparing demographic and medical history variables between patients prescribed warfarin and not prescribed warfarin on admission in patients who are appropriate candidates for anticoagulant therapy

Variable	Warfarin (n = 57)	No warfarin (n = 49)	P-value
Age (mean ± SD)(years)	77.7 ± 8.6	82.0 ± 9.2	0.02
Female	27/57 (47%)	32/49 (65%)	0.05
Previous stroke or TIA	29/57 (51%)	16/49 (33%)	0.05
Hypertension	43/57 (75%)	36/49 (74%)	0.5
Diabetes Mellitus	10/57 (18%)	6/49 (12%)	0.3
Congestive cardiac failure	21/57 (37%)	8/46 (17%)	0.02
History of PUD or remote GI bleed	5/57 (8.8%)	6/49 (12.2%)	0.4
Alcohol consumption	2/57 (4%)	4/49 (8.2%)	0.3
Dementia	12/57 (21%)	10/48 (21%)	0.6
Falls	4/56 (7%)	7/48 (15%)	0.2
Living in Supportive care	9/57 (16%)	6/49 (12%)	0.4
CHADS2 (mean ± SD)	2.9 ± 1.3	2.5 ± 1.3	0.14
Bleeding score (mean ± SD)	2.1 ± 0.9	1.6 ± 0.8	0.003

Table 3.
Multivariable analysis assessing predictors of warfarin therapy among patients deemed to be appropriate candidates for anticoagulation

Variable	Odds ratio	95%CI	P-value
Age	0.7	(0.5–0.9)	0.01
Male	2.0	(0.7–5.5)	0.2
Previous stroke or TIA	2.6	(1.1–6.5)	0.04
Congestive cardiac failure	3.2	(1.1–9.0)	0.03
History of PUD or remote GI bleed	0.6	(0.1–3.1)	0.5

Age was coded in 10-year increments above 75 years (less than 75 years was the dummy variable).

Thrombosis Research
Volume 120, Issue 5, 2007, Pages 663-669

---

Prevalence and biologic profile of aspirin resistance in patients with angiographically proven coronary artery disease

Idoia Narvaez^{a, ,} , Jose Domingo Sagastagoitia^b, Marta Vacas^a, Yolanda Saez^a, Manolo Lafita^a, Santos Monica^a, Jesus Pablo Saez de Lafuente^a, Enrique Molinero^b and Jose Antonio Iriarte^a
aFIDEC, Department of Atherosclesosis and Thrombosis, Gurtubay s/n, 48012, Bilbao, Vizcaya, Spain
^bCardiology Service, Hospital de Basurto, Bilbao, Spain
Received 18 July 2006;  revised 18 December 2006;  accepted 28 December 2006.  Available online 28 February 2007.

Aspirin protects from cardiovascular events. However, a number of patients who take this drug suffer events, probably due to aspirin resistance. The role of certain biologic variables that may affect resistance is still uncertain.

To determine the prevalence of aspirin resistance in patients taking this drug and to test if resistance is related to haemostatic, inflammatory and lipidic variables.

Platelet function measured with PFA-100 was studied in 268 patients (185 men) with stable coronary disease who took aspirin (100 to 300 mg/day). Aspirin resistance was defined when epinephrine closure time < 174 s. Results of lipoprotein(a) are expressed in median (interquartile range).

Aspirin resistance was found in 16% of cases. Patients with aspirin resistance had higher levels of Apolipoprotein B (109.27 ± 27.65 vs 100.92 ± 23.77 mg/dl; p < 0.05), lipoprotein(a) [20.37 (4.83–36.72) vs 10.02 (1.88–25.41); p < 0.01], Platelet Count (241.42 ± 75.35 vs 213.94 ± 56.74 mm³; p < 0.05) and fibrinogen (388.93 ± 107.27 vs 354.33 ± 89.35 mg/dl; p < 0.05). We used the logistic regression analysis to detect the independent predictors of aspirin resistance. Lipoprotein(a) was found to be the only independent risk factor to identify aspirin resistance (p < 0.05; OR: 1.302; CI 95%: 1.003–1.688).

Although the potential mechanisms of aspirin resistance still remains uncertain, we found that platelet responsiveness to aspirin is reduced in patients with high levels of Apolipoprotein B and lipoprotein(a). Our work demonstrate that lipoprotein(a) is an independent risk factor for aspirin resistance possibly due to the interaction of Apolipoprotein(a) with human platelets.

Aspirin resistance; Coronary artery; PFA-100; Epinephrine closure time; Lipoproteins; Fibrinogen

Baseline clinical characteristics according to aspirin sensitive status

	ASA-R (n = 44)	ASA-S (n = 224)	P value
Age (years)	64.09 ± 12.69	64.36 ± 11.34	0.896
Females n (%)	13 (29.5)	70 (31.3)	0.48
BMI (kg/m2)	26.99 ± 3.64	28.01 ± 3.93	0.102
Smokers n (%)	12 (27.9)	55 (25.1)	0.88
Diabetes (%)	10 (26.3)	50 (23.8)	0.837
Hypertension (%)	19 (43.2)	112 (50.0)	0.416
Statins (%)	25 (56.8)	116 (51.8)	0.33
Aspirin (mg/day)	119.23 ± 65.51	136.64 ± 84.83	0.150
Fibrinogen (mg/dl)	388.93 ± 107.27	354.33 ± 89.35	0.024
Platelet (mm3)	241.42 ± 75.35	213.94 ± 56.74	0.018
Leukocytes (mm3)	8.19 ± 2.43	7.68 ± 2.11	0.22
Hemoglobin (mg/dl)	13.56 ± 1.96	13.83 ± 2.0	0.48
Apo A (mg/dl)	128.95 ± 27.69	128.03 ± 25.58	0.83
Apo B 100 (mg/dl)	109.27 ± 27.65	100.92 ± 23.77	0.039
LDL-c (mg/dl)	118.95 ± 37.10	111.21 ± 32.26	0.19
Lg Lp(a)	2.65 ± 1.41	1.86 ± 1.77
Lp(a) (mg/dl)	20.37 (4.83–36.72)	10.02 (1.88–25.41)	0.004
Lg D-Dimer (ng/ml)	6.25 ± 0.81	6.18 ± 0.75
D-Dimer (ng/ml)	432.0 (270.75–980.50)	430.0 (283.0–780.25)	0.79
Lg CRP (mg/L)	1.41 ± 1.37	1.54 ± 1.37
CRP (mg/L)	3.26 (1.36–11.72)	4.28 (1.53–13.01)	0.54

ASA-R = aspirin resistant; ASA-S = aspirin sensitive; BMI; Body Mass Index.
Apo A1: Apolipoprotein A1.
Apo B 100: Apolipoprotein B100.
Lp(a): Lipoprotein(a).
Lg Lp(a): Logarithm Lipoprotein(a).
CRP: C reactive protein.

Table 2.
Univariate and multivariate analysis of predictors of aspirin resistance

	P value	O. R.	95% C. I.
Univariate predictors
Apo B 100	0.041	1.013	1.001–1.026
Fibrinogen	0.027	1.004	1.000–1.007
Platelets	0.021	1.007	1.001–1.013
Lp(a)	0.008	1.349	1.083–1.681
Multivariate predictors
Lp(a)	0.047	1.302	1.003–1.688

Thrombosis Research
Volume 120, Issue 5, 2007, Pages 671-677

---

Serial Changes in von Willebrand Factor-Cleaving Protease (ADAMTS13) and Prognosis After Acute Myocardial Infarction

Masakazu Matsukawa MD^a, Koichi Kaikita MD^{a, ,} , Kenji Soejima PhD^b, Shunichiro Fuchigami MD^a, Yoshinori Nakamura MD^c, Tsuyoshi Honda MD^a, Kenichi Tsujita MD^a, Yasuhiro Nagayoshi MD^a, Sunao Kojima MD^a, Hideki Shimomura MD^c, Seigo Sugiyama MD^a, Kazuteru Fujimoto MD^d, Michihiro Yoshimura MD^a, Tomohiro Nakagaki PhD^b and Hisao Ogawa MD^a
aDepartment of Cardiovascular Medicine, Graduate School of Medical Sciences, Kumamoto University, Kumamoto, Japan
^bFirst Research Department, The Chemo-Sero-Therapeutic Research Institute, Kumamoto, Japan
^cDepartment of Cardiovascular Medicine, Fukuoka Tokushukai Medical Center, Kumamoto, Japan
^dDepartment of Cardiovascular Medicine, National Hospital Organization Kumamoto Medical Center, Kumamoto, Japan.
Received 8 February 2007;  revised 29 March 2007;  accepted 29 March 2007.  Available online 12 June 2007.

Von Willebrand factor (VWF), a cofactor in platelet adhesion and aggregation, increases hemostasis and thrombosis. Recently, a metalloprotease that cleaves VWF multimers has been identified, namely ADAMTS13. The aim of this study was to investigate the relation between serial changes in plasma VWF and ADAMTS13 and the prognosis after acute myocardial infarction (AMI). We measured serial changes of plasma VWF and ADAMTS13 antigen levels in 92 patients with AMI and 40 control subjects. VWF levels were significantly higher in patients with AMI compared with controls (p <0.01) on admission, peaked 3 days after admission, and remained high for 14 days. In contrast, on admission, ADAMTS13 levels were significantly lower in patients with AMI compared with controls (p <0.0001), with minimum antigen levels reached after 3 days, and remained lower for 14 days. The ratio of VWF/ADAMTS13 antigen levels was higher in patients with AMI compared with controls throughout the time course. Cox hazards analysis revealed that the early increase of VWF and VWF/ADAMTS13 ratio levels and the early decrease of ADAMTS13 levels were significant predictors of future thrombotic events during the 1-year follow-up period. Kaplan-Meier analysis demonstrated that patients with major decreases of ADAMTS13 levels and high increases of VWF/ADAMTS13 levels had significantly greater probabilities for development of thrombotic events (p = 0.0104 and 0.0209, respectively). In conclusion, these findings suggest that monitoring the changes of VWF and ADAMTS13 antigen levels in the early phase might be valuable for predicting and preventing thrombosis during 1-year follow-up in patients with AMI.

Figure 1. Serial changes in plasma VWF (A) and ADAMTS13 (B) antigen levels (mean ± SEM, units per milliliter) and VWF/ADAMTS13 ratios (C) in 92 patients with AMI. *p <0.01 versus CTL group; ^†p <0.01 versus admission; ^‡p <0.0001 versus CTL group; ^§p <0.01 versus admission.

Figure 2. Kaplan-Meier cumulative 1-year event-free survival curves in 2 groups divided according to cut-off values of ΔVWF (A), ΔADAMTS13 (B), and ΔVWF/ADAMTS13 (C). Each cut-off point was defined by a receiver-operating characteristic curve. Patients with high ΔVWF/ADAMTS13 and low ΔADAMTS13 values had significantly higher probabilities of developing thrombosis.

Patient characteristics

Variable	AMI (n = 92)	CTL (n = 40)	p Value
Age (yrs)	65 ± 1	65 ± 1	0.95
Men/women	67/25	25/15	0.24
Hypertension	58 (63%)	19 (48%)	0.08
Diabetes mellitus	22 (24%)	9 (23%)	0.83
Active smokers	48 (52%)	11 (28%)	0.0006
Body mass index (kg/m2)	23.8 ± 0.3	23.9 ± 0.5	0.90
Total cholesterol (mg/dl)	199 ± 4	185 ± 5	0.06
High-density lipoprotein cholesterol (mg/dl)	45 ± 1	49 ± 1	0.15
Low-density lipoprotein cholesterol (mg/dl)	122 ± 3	113 ± 4	0.14
Triglyceride (mg/dl)	133 (98–178)	99 (76–132)	0.002
Uric acid (mg/dl)	5.8 ± 0.1	5.6 ± 0.2	0.48
C-reactive protein (mg/dl)	0.16 (0.05–0.42)	0.05 (0.02–0.14)	0.001

Table 2.
Univariate Cox hazards analysis of clinical events during one-year follow-up

Variable	HF + Thrombosis (n = 17, 18.4%)		HF (n = 6, 6.5%)		Thrombosis (n = 13, 14.1%)
	OR (95% CI)	p Value	OR (95% CI)	p Value	OR (95% CI)	p Value
VWF day 0	0.94 (0.62–1.42)	0.79	0.72 (0.26–2.02)	0.54	1.11 (0.63–1.96)	0.70
VWF day 3	1.13 (0.86–1.48)	0.38	1.13 (0.68–1.88)	0.61	1.46 (1.06–2.01)	0.019
ΔVWF day 3 − day 0	1.17 (0.91–1.51)	0.20	1.23 (0.80–1.87)	0.33	1.41 (1.03–1.95)	0.03
ADAMTS13 day 0	0.47 (0.07–3.03)	0.42	0.20 (0.003–15.18)	0.46	1.92 (0.16–21.9)	0.59
ADAMTS13 day 3	0.08 (0.008–0.09)	0.04	0.19 (0.001–25.8)	0.51	0.06 (0.002–2.12)	0.12
ΔADAMTS13 day 3 − day 0	0.12 (0.006–2.33)	0.16	3.36 (0.006–1752)	0.70	0.001 (0.0001–0.08)	0.002
VWF/ADAMTS13 day 0	0.98 (0.81–1.18)	0.87	0.97 (0.63–1.49)	0.90	0.92 (0.68–1.26)	0.63
VWF/ADAMTS13 day 3	1.20 (1.03–1.40)	0.015	1.16 (0.87–1.54)	0.28	1.41 (1.14–1.75)	0.001
ΔVWF/ADAMTS13 day 3 − day 0	1.21 (1.04–1.40)	0.011	1.17 (0.89–1.53)	0.25	1.49 (1.21–1.82)	0.0001

CI = confidence interval; OR = odds ratio.

Table 3.
Comparison of drugs administered during the follow-up period in patients with and without one-year thrombosis

Variable	With 1-yr Thrombosis (n = 13)	Without 1-yr Thrombosis (n = 79)	p Value
Aspirin	12 (92%)	77 (97%)	0.39
Ticlopidine	8 (62%)	59 (75%)	0.33
Calcium antagonists	3 (23%)	36 (46%)	0.11
Long-acting nitrates	4 (31%)	11 (14%)	0.15
β Blockers	3 (23%)	16 (20%)	0.81
Angiotensin-converting enzyme inhibitors	9 (69%)	35 (44%)	0.09
Angiotensin II type I receptor blockers	4 (31%)	39 (49%)	0.20
Statins	4 (31%)	34 (43%)	0.39

Table 4.
Cox hazards analysis of risk factors for thrombosis during one-year follow-up in patients with acute myocardial infarction

Variable	Univariate Analysis			Multivariate Analysis
	OR	95% CI	p Value	OR	95% CI	p Value
Age	1.05	0.99–1.10	0.056	1.10	0.99–1.22	0.057
Men	0.83	0.25–2.71	0.76	—	—	—
Hypertension	1.45	0.48–4.33	0.50	—	—	—
Diabetes mellitus	0.48	0.15–1.49	0.20	—	—	—
Active smokers	2.66	0.82–8.66	0.10	—	—	—
Body mass index	0.93	0.78–1.11	0.44	—	—	—
Family history of coronary artery disease	0.56	0.15–2.04	0.38	—	—	—
Total cholesterol	0.99	0.98–1.00	0.55	—	—	—
High-density lipoprotein cholesterol	0.95	0.89–1.00	0.08	—	—	—
Low-density lipoprotein cholesterol	1.00	0.98–1.01	0.99	—	—	—
Log triglyceride	0.13	0.09–2.16	0.15	—	—	—
Uric acid	1.16	0.80–1.68	0.41	—	—	—
Log C-reactive protein	0.76	0.31–1.86	0.55	—	—	—
Log B-type natriuretic peptide (pg/ml)	3.93	0.84–18.40	0.08	0.71	0.09–5.41	0.74
Log maximum creatine phosphokinase	2.28	0.51–10.16	0.27	—	—	—
Log maximum creatine