2023 Volume 30 Issue 11 Pages 1703-1714
Aim: Studies investigating the relationship between pulse pressure (PP) and prognosis in acute ischemic stroke remain limited. Thus, in this study, we aim to determine whether changes in PP in the early phase of ischemic stroke are associated with neurological deterioration or stroke recurrence.
Methods: Patients who participated in the Acute Aspirin Plus Cilostazol Dual Therapy for Non-cardiogenic Stroke Patients Within 48 Hours of Symptom Onset (ADS) trial were included in this study. We then divided the patients into four groups (low-low, low-high, high-low, high-high) according to low or high PP both on admission and 24 h after admission. The threshold PP calculated by receiver operating characteristic curve analysis of PP on admission for neurological deterioration within 14 days and recurrent ischemic stroke/transient ischemic attack (TIA) within 3 months was 69 mmHg.
Results: Neurological deterioration within 14 days was observed in 118 patients (10.6%), whereas recurrent ischemic stroke/TIA within 3 months was noted in 34 patients (3.2%). Among these four groups, both neurological deterioration within 14 days (odds ratio [OR] 2.09, 95% confidence interval [CI] 1.12–3.91; p=0.0209) and recurrent ischemic stroke/TIA within 3 months (OR 4.80; 95% CI 1.62–14.86; p=0.0064) were significantly more frequent in the high-high group than in the low-low group as per the results of our multivariate analysis. In addition, neurological deterioration within 14 days was significantly higher in the high-low group than that in the low-low group (OR 2.70; 95% CI 1.44–5.05; p=0.0019).
Conclusions: High PP during the acute phase of ischemic stroke appears to be associated with ischemic stroke recurrence and neurological deterioration, particularly if PP is elevated both on admission and 24 h later after admission.
Abbreviations: ADS, Acute Aspirin Plus Cilostazol Dual Therapy for Non-cardiogenic Stroke Patients Within 48 Hours of Symptom Onset; BAD, branch atheromatous disease; CI, confidence interval; IMT, intima–media thickness; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio; PP, pulse pressure; ROC, receiver operating characteristic; TEE, transesophageal echocardiography; TIA, transient ischemic attack; TOAST, Trial of Org 10172 in Acute Stroke Treatment.
Blood pressure is commonly noted to increase during the acute phase of ischemic stroke1-3). Moreover, patients with high systolic blood pressure in the early stage of acute ischemic stroke were determined to have a higher risk of stroke recurrence and poor prognosis2, 4). Pulse pressure (PP), which is the focus of this present study, has been considerably less reported than systolic blood pressure in terms of its association with ischemic stroke. Thus, in this study, we have undertaken a post-hoc analysis of the Acute Aspirin Plus Cilostazol Dual Therapy for Non-cardiogenic Stroke Patients Within 48 Hours of Symptom Onset (ADS) trial5). As per the findings of the ADS trial, no difference was noted in terms of short-term neurological deterioration between patients receiving dual antiplatelet agents (aspirin and cilostazol) or aspirin alone as acute antiplatelet therapy for acute non-cardiogenic ischemic stroke. On the other hand, a sub-analysis showed that tachycardia was associated with neurological deterioration6). Hemodynamics may therefore affect neurological deterioration and poor prognosis in patients with non-cardiogenic ischemic stroke independent of antithrombotic therapy. PP can vary depending on one’s atherosclerotic status; moreover, an increase in PP can occur when arteries become narrowed or stiff7). Our hypothesis was that higher PP is associated with poor hemodynamics and thus reflects worse regional cerebral vasculopathy in patients with atherothrombotic cerebral infarction or other types of ischemic stroke associated with atherosclerotic risk factors. PP of hypertensive patients serves as an independent risk factor for overall and cardiovascular mortality8). On the other hand, few studies have reported an association between acute ischemic stroke and PP. These reports have primarily examined the association of high PP with poor prognosis and recurrent stroke9, 10). Two other reports have demonstrated a J-shaped or reverse J-shaped curve for the relationship between PP and stroke recurrence11, 12). We predicted that high PP would be associated with neurological deterioration in the acute phase and poor short-term prognosis. However, the findings with regard to the relationship between PP and prognosis in acute stroke have remained inconsistent between studies, and there are no reports clarifying the effect of fluctuations in PP through the early days of ischemic stroke. Therefore, in this study, we investigated whether persistent high PP in the early stages of non-cardiogenic ischemic stroke is associated with neurological deterioration or ischemic stroke recurrence; this is a post-hoc analysis of the ADS trial.
In this study, patients enrolled in the ADS trial were subjected to further analysis. The details of ADS study and its main results have already been reported5, 13).
The ADS trial was conducted at 34 stroke centers in Japan between May 2011 and June 2017. In the study, the inclusion criteria for patients were as follows: age ≥ 18 years; presentation with non-cardioembolic stroke within 48 h of symptom onset; neurological deficits with National Institutes of Health Stroke Scale (NIHSS) score ＜20; and a premorbid modified Rankin Scale (mRS) score of 0–2. Meanwhile, the exclusion criteria were as follows: cardioembolic stroke with a high-risk source as defined by the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria; current medication with antiplatelet agents, cilostazol, aspirin ≥ 200 mg/day, clopidogrel, ticlopidine, or any anticoagulants before stroke onset; having undergone or planning to undergo thrombectomy; bleeding or tendency to bleed; pregnant or lactating; congestive heart failure; gastrointestinal ulcers; malignancy within the past 5 years; allergies to salicylic acid or cilostazol; or judged ineligible for the study by an investigator. Patients were randomly assigned to receive dual antiplatelets (aspirin plus cilostazol) or aspirin alone as acute antiplatelet therapy. The primary efficacy evaluation included neurological deterioration and transient ischemic attacks (TIAs) or stroke recurrence within 14 days. Neurological deterioration was considered positive when NIHSS score increased by ≥ 2. The primary safety result was intracerebral or subarachnoid hemorrhage. Secondary outcomes included an mRS score of 0–1 at 3 months after symptom onset. Another evaluation has confirmed the recurrence of symptomatic stroke (including TIA) within 3 months, and we verified this result in this study. One of the main selection criteria for the ADS trial was that patients should be admitted within 48 h of symptom onset. The dual therapy with cilostazol and aspirin lasted for 2 weeks. However, this treatment was not able to reduce the rate of short-term neurological deterioration but may have improved clinical outcomes 3 months after symptom onset.
This study was a sub-analysis of the ADS trial, which enrolled 1208 patients (Fig.1). However, seven patients who withdrew their consent were excluded from this analysis. After excluding patients whose blood pressure could not be confirmed either on admission or 24 h after admission, 1190 patients were followed up until 14 days, whereas 1116 patients were followed up 3 months later. This study has focused on non-cardiogenic stroke as atherothrombotic stroke, or another type of stroke mainly associated with atherosclerotic risk factors, as we hypothesized that PP reflects atherosclerotic status, primarily in the form of arterial changes. Patients with a final diagnosis of TIA or cardiogenic cerebral embolism were excluded. In total, 1117 patients were analyzed at 14 days after symptom onset, whereas 1051 patients were analyzed at 3 months after symptom onset. The local institutional ethics committee approved this study. All patients provided informed consent to participate in the initial ADS trial before enrollment.
Patient selection procedure
In each patient, blood pressure was measured on admission and at 24 h, 48 h, 7 days, and 14 days after admission. The endpoint of this present sub-analysis was to determine the association of PP fluctuations to neurological deterioration within the duration of 14 days after admission. Blood pressure values on admission and 24 h after admission were therefore used to evaluate PP. PP is calculated as the difference between systolic and diastolic blood pressures.
The following clinical information were also obtained and analyzed: (1) age and sex; (2) vascular risk factors (hypertension, hyperlipidemia, and diabetes mellitus); (3) pharmacotherapies before stroke and previous history of stroke; and (4) stroke classification. Neurological deficits were assessed using NIHSS score on admission and 24 h, 48 h, 7 days, and 14 days (or at discharge) after admission. The final diagnosis was determined at discharge. Stroke classification is divided into large-artery atherosclerotic disease, lacunar disease, branch atheromatous disease (BAD), other determining infarcts (paradoxical embolism, cerebrovascular dissection, aortogenic embolism, or others), undetermined infarct, cardioembolic stroke, or TIA. Large-artery atherosclerosis was defined as a major arterial lesion with ≥ 50% stenosis. Internal carotid artery stenosis and occlusion were evaluated using cervical ultrasonography and conventional cerebral angiography as deemed appropriate. Patients who underwent cervical ultrasonography also had their maximum intima–media thickness (IMT) of the common carotid artery recorded. For cases with possible embolic mechanisms, transesophageal echocardiography (TEE) was performed at the discretion of the attending physician. BAD was defined as a disease involving the perforating artery territory with an infarct diameter of ≥ 1.5 cm without major arterial lesions. At 3 months after onset, mRS was evaluated at each institution.
Statistical AnalysisFirst, clinical characteristics of patients were analyzed for neurological deterioration within 14 days and recurrent ischemic stroke/TIA within 3 months. To evaluate trends, those 1117 patients who were evaluable after 14 days and 1051 patients who were evaluable after 3 months were divided into four groups according to their quartiles of PP on admission (thresholds: 60, 71, and 86 mmHg). Associations between PP grade and neurological deterioration within 14 days or recurrent ischemic stroke/TIA within 3 months were thereafter assessed.
Next, we investigated the prognostic association of variability in acute-phase PP using PP thresholds during admission and 24 h after admission for neurological deterioration within 14 days and recurrent ischemic stroke/TIA within 3 months, respectively. Thresholds were calculated using receiver operating characteristic (ROC) curve analysis. All patients were divided into four groups according to their PP threshold values on admission: low PP on admission＋low PP at 24 h after admission (low-low group); low PP on admission＋high PP at 24 h after admission (low-high group); high PP on admission＋low PP at 24 h after admission (high-low group); and high PP on admission＋high PP at 24 h after admission (high-high group). Group comparisons were conducted to determine differences in clinical background factors among the four groups. Univariate analysis was thereafter performed for neurological deterioration within 14 days and recurrent ischemic stroke/TIA within 3 months. Finally, multivariate analyses were performed on the four groups, with clinical background factors exhibiting values of p＜0.1 for neurological deterioration within 14 days and recurrent ischemic stroke/TIA within 3 months. Then, χ2 test, non-paired t-test, and Wilcoxon test were used for univariate analyses. For multivariate analysis, logistic regression analysis was performed. Two-tailed probabilities of p＜0.05 were considered indicative of statistical significance. Statistical analysis was performed using JMP Pro version 16.1 software (SAS Institute, Cary, NC).
Neurological deterioration assessments within 14 days after admission were available for 1117 patients, whereas evaluations for recurrent ischemic stroke/TIA within 3 months after admission were available for 1051 patients. Neurological deterioration within 14 days was observed in 118 patients (10.6%), while recurrent ischemic stroke/TIA within 3 months was observed in 34 patients (3.2%).
First, we examined the characteristics of patients with neurological deterioration within 14 days and recurrent ischemic stroke/TIA within 3 months (Tables 1, 2). It can be noted that PP was significantly higher on admission (p=0.0002) and at 24 h after admission (p=0.0084) for patients showing neurological deterioration within 14 days. In addition, significant differences in terms of age, sex, mRS score, NIHSS score, prior cerebral hemorrhage, concomitant anticoagulation therapy, lacunar disease, and BAD were noted, in addition to PP (Table 1). Patients exhibiting recurrent ischemic stroke/TIA within 3 months likewise exhibited significantly higher PP on both admission (p=0.0115) and at 24 h after admission (p=0.0130). Moreover, significant differences were observed for previous aspirin use, previous ischemic stroke, large-artery atherosclerosis, and lacunar disease (Table 2). For patients who underwent carotid ultrasonography, maximum IMT of the common carotid artery showed no significant difference between patients with and without neurological deterioration within 14 days (p=0.2344). On the other hand, maximum IMT was significantly greater in patients with recurrent ischemic stroke/TIA within 3 months than in those without (p=0.0083) (Tables 1, 2). TEE was also performed in 291 (26.5%) patients with evaluated 14-day neurological deterioration. Similarly, TEE was performed in 277 patients (26.4%) with evaluated recurrent ischemic stroke/TIA within 3 months.
| Variables | Present (n= 118) | Absent (n= 999) | p value |
|---|---|---|---|
| Pulse pressure on admission, mmHg (mean±SD) | 80.56±22.16 | 72.87±20.95 | 0.0002 |
| Pulse pressure 24 h after admission, mmHg (mean±SD) | 69.92±17.03 | 65.41±17.59 | 0.0084 |
| Age, years (mean±SD) | 71.92±11.86 | 67.64±11.75 | 0.0002 |
| Male, n (%) | 68 (57.63) | 677 (67.77) | 0.0271 |
| Premorbid modified Rankin Scale, median (IQR) | 0 (0–0) | 0 (0–0) | 0.0099 |
| NIHSS score, median (IQR) | 3 (2-5) | 2 (1-4) | ＜0.0001 |
| Vascular risk factor | |||
| Hypertension, n (%) | 97 (82.20) | 775 (77.58) | 0.2508 |
| Diabetes mellitus, n (%) | 46 (38.98) | 314 (31.43) | 0.0969 |
| Hyperlipidemia, n (%) | 50 (42.37) | 490 (49.05) | 0.1699 |
| Past aspirin therapy, n (%) | 12 (10.17) | 95 (9.51) | 0.8178 |
| Past statin therapy, n (%) | 16 (13.56) | 149 (14.91) | 0.6947 |
| Past ischemic stroke, n (%) | 10 (8.47) | 98 (9.81) | 0.6425 |
| Past intracerebral hemorrhage, n (%) | 5 (4.24) | 11 (1.10) | 0.0067 |
| Treatment | |||
| Dual antiplatelet therapy, n (%) | 61 (51.69) | 489 (48.95) | 0.5726 |
| Aspirin dosage: 162–200 mg, n (%) | 98 (83.05) | 798 (79.88) | 0.4135 |
| Concomitant therapy (argatroban or heparin), n (%) | 85 (72.03) | 277 (27.73) | ＜0.0001 |
| Statin, n (%) | 73 (61.86) | 579 (57.96) | 0.4156 |
| Antihypertensive, n (%) | 56 (47.46) | 526 (52.65) | 0.2854 |
| Clinical subtype | |||
| Large artery atherosclerosis, n (%) | 19 (16.10) | 147 (14.71) | 0.6887 |
| Lacunar, n (%) | 19 (16.10) | 511 (51.15) | ＜0.0001 |
| Branch atheromatous disease, n (%) | 56 (47.46) | 115 (11.51) | ＜0.0001 |
| Other determined etiology, n (%) | 12 (9.16) | 106 (9.49) | 0.578 |
| Undetermined etiology, n (%) | 12 (10.08) | 106 (10.62) | 0.857 |
| CCA maximum IMT, mm (mean±SD) | 1.40±0.81 (n= 108) | 1.31±0.75 (n= 900) | 0.2344 |
SD, standard deviation; IQR, interquartile range; NIHSS, National Institutes of Health Stroke Scale; IMT, intima–media thickness; CCA, common carotid artery.
| Variables | Present (n= 34) | Absent (n= 1017) | p value |
|---|---|---|---|
| Pulse pressure on admission, mmHg (mean±SD) | 82.53±20.35 | 73.18±21.22 | 0.0115 |
| Pulse pressure 24 h after admission, mmHg (mean±SD) | 72.97±18.83 | 65.33±17.58 | 0.0130 |
| Age, years (mean±SD) | 71.38±10.33 | 67.98±11.89 | 0.0998 |
| Male, n (%) | 25 (73.53) | 676 (66.47) | 0.3902 |
| Premorbid modified Rankin Scale, median (IQR) | 0 (0–0.25) | 0 (0–0) | 0.1796 |
| NIHSS score, median (IQR) | 3 (1–4.25) | 2 (1–4) | 0.6672 |
| Vascular risk factor | |||
| Hypertension, n (%) | 27 (79.41) | 795 (78.17) | 0.8631 |
| Diabetes mellitus, n (%) | 14 (41.18) | 324 (31.86) | 0.2525 |
| Hyperlipidemia, n (%) | 16 (47.06) | 494 (48.57) | 0.8619 |
| Past aspirin therapy, n (%) | 7 (20.59) | 93 (9.14) | 0.0253 |
| Past statin therapy, n (%) | 5 (14.71) | 152 (14.95) | 0.9692 |
| Past ischemic stroke, n (%) | 7 (20.59) | 93 (9.14) | 0.0253 |
| Past intracerebral hemorrhage, n (%) | 0 (0.00) | 14 (1.38) | 0.4910 |
| Treatment | |||
| Dual antiplatelet therapy, n (%) | 15 (44.12) | 499 (49.07) | 0.5702 |
| Aspirin dosage: 162–200 mg, n (%) | 29 (85.29) | 812 (79.84) | 0.4342 |
| Concomitant therapy (argatroban or heparin), n (%) | 13 (38.24) | 327 (32.15) | 0.4558 |
| Statin, n (%) | 20 (58.82) | 596 (58.60) | 0.9796 |
| Antihypertensive, n (%) | 14 (41.18) | 490 (48.18) | 0.4213 |
| Clinical subtype | |||
| Large artery atherosclerosis, n (%) | 12 (35.29) | 146 (14.36) | 0.0008 |
| Lacunar, n (%) | 9 (26.47) | 500 (49.16) | 0.0092 |
| Branch atheromatous disease, n (%) | 2 (5.88) | 151 (14.85) | 0.1448 |
| Other determined etiology, n (%) | 5 (14.71) | 115 (11.31) | 0.5400 |
| Undetermined etiology, n (%) | 6 (17.65) | 105 (10.32) | 0.1718 |
| CCA maximum IMT, mm (mean±SD) | 1.69±1.11 (n= 28) | 1.31±0.74 (n= 923) | 0.0083 |
TIA, transient ischemic attack; SD, standard deviation; IQR, interquartile range; NIHSS, National Institutes of Health Stroke Scale; IMT, intima– media thickness; CCA, common carotid artery.
Next, to evaluate trends, we compared the PP and frequency of neurological deterioration within 14 days or recurrent ischemic stroke/TIA within 3 months in four groups divided by quartiles of PP on admission (＜61 mmHg, 61–71 mmHg, 72–86 mmHg, and ≥ 87 mmHg). In univariate analyses, compared to the lowest-quartile PP group, the highest-quartile PP group has exhibited significantly greater frequencies of neurological deterioration both within 14 days of admission (odds ratio [OR] 2.49; 95% confidence interval [CI] 1.36–4.54; p=0.0030) and at 24 h after admission (OR 2.00; 95% CI 1.12–3.59; p=0.0193) (Table 3). Recurrent ischemic stroke/TIA within 3 months was also noted to be significantly more frequent in the highest-quartile PP group than in the lowest-quartile PP group on admission (OR 3.23; 95% CI 1.01–10.26; p=0.0437) and at 24 h after admission (OR 4.20; 95% CI 1.49–11.82; p=0.0065).
| n (%) | OR | 95%CI | p value | |
|---|---|---|---|---|
| PP on admission | ||||
| Neurological deterioration within 14 days according to PP on admission | ||||
| ＜61 | 17 (5.67) | 1 | ||
| 61-71 | 22 (8.49) | 1.55 | 0.80-2.98 | 0.1935 |
| 72-86 | 43 (15.30) | 3.01 | 1.67-5.41 | 0.0002 |
| ≥ 87 | 36 (13.00) | 2.49 | 1.36-4.54 | 0.003 |
| Recurrent ischemic stroke/TIA within 3 months according to PP on admission | ||||
| ＜61 | 3 (1.03) | 1 | ||
| 61-71 | 7 (2.65) | 2.16 | 0.62-7.47 | 0.224 |
| 72-86 | 11 (4.60) | 3.31 | 1.06-10.40 | 0.0401 |
| ≥ 87 | 13 (5.08) | 3.23 | 1.01-10.26 | 0.0473 |
| PP at 24 h | ||||
| Neurological deterioration within 14 days according to PP at 24 h after admission | ||||
| ＜61 | 36 (7.68) | 1 | ||
| 61-71 | 33 (11.91) | 1.63 | 0.99-2.68 | 0.0554 |
| 72-86 | 29 (12.55) | 1.73 | 1.03-2.90 | 0.0383 |
| ≥ 87 | 20 (14.29) | 2 | 1.12-3.59 | 0.0193 |
| Recurrent ischemic stroke/TIA within 3 months according to PP at 24 h after admission | ||||
| ＜61 | 7 (1.55) | 1 | ||
| 61-71 | 11 (4.26) | 2.83 | 1.08-7.40 | 0.0337 |
| 72-86 | 8 (3.77) | 2.49 | 0.89-6.97 | 0.0815 |
| ≥ 87 | 8 (5.20) | 4.2 | 1.49-11.82 | 0.0065 |
OR, odds ratio; TIA, transient ischemic attack; PP, pulse pressure.
Thereafter, we calculated the threshold values of PP on admission for neurological deterioration within 14 days and recurrent ischemic stroke/TIA within 3 months using ROC curve analyses. Thresholds of PP on admission were calculated as 69 mmHg for both neurological deterioration within 14 days and recurrent ischemic stroke/TIA within 3 months. Areas under the ROC curve were determined as follows: 0.6068 for neurological deterioration within 14 days and 0.6275 for recurrent ischemic stroke/TIA within 3 months.
To assess the effects of changes in PP over time, the background characteristics of patients with neurological deterioration within 14 days and recurrent ischemic stroke/TIA within 3 months were evaluated for the four groups (low-low, low-high, high-low, and high-high group) using the PP thresholds of 69 mmHg on admission and 24 h after admission (Tables 4, 5). For neurological deterioration within 14 days, a significant difference was observed between the four groups (p=0.0242). Significant differences were also noted in terms of age, sex, hypertension, dual antiplatelet therapy, aspirin dosage of 162–200 mg/day, concomitant anticoagulant therapy, antihypertensive therapy, and BAD (Table 4). Likewise for recurrent ischemic stroke/TIA within 3 months, a significant difference was observed between the four groups (p＜0.0001). Significant differences were also evident as regards age, sex, hypertension, dual antiplatelet therapy, aspirin dosage of 162–200 mg/day, concomitant anticoagulant therapy, antihypertensive therapy, and BAD (Table 5). For patients who underwent carotid ultrasonography, maximum IMT of the common carotid artery showed significant differences among the four groups in both neurological deterioration within 14 days (p=0.0012) and recurrent ischemic stroke/TIA within 3 months (p=0.0022). Maximum IMT tended to be greater in the high-high group than in the low-low group (Tables 4, 5).
| Variables | low-low (n= 377) | low-high (n= 103) | high-low (n= 306) | high-high (n= 331) | p value |
|---|---|---|---|---|---|
| Neurological deterioration within 14 days | 18 (4.77) | 11 (10.68) | 43 (14.05) | 46 (13.90) | ＜0.0001 |
| Age, years (mean±SD) | 64.77±12.22 | 70.16±12.26 | 69.37±11.36 | 70.06±10.87 | ＜0.0001 |
| Male, n (%) | 286 (75.86) | 66 (64.08) | 197 (64.38) | 196 (59.21) | ＜0.0001 |
| Premorbid mRS, median (IQR) | 0 (0–0) | 0 (0–0) | 0 (0–0) | 0 (0–0) | 0.3246 |
| NIHSS score, median (IQR) | 2 (1–4) | 2 (1–4) | 2 (1–4) | 3 (1–4) | 0.2091 |
| Vascular risk factor | |||||
| Hypertension, n (%) | 265 (70.29) | 81 (78.64) | 243 (79.41) | 283 (85.50) | ＜0.0001 |
| Diabetes mellitus, n (%) | 108 (28.65) | 36 (34.95) | 95 (31.05) | 121 (36.56) | 0.1329 |
| Hyperlipidemia, n (%) | 188 (49.87) | 50 (48.54) | 139 (45.42) | 163 (49.24) | 0.6814 |
| Past aspirin therapy, n (%) | 34 (9.02) | 12 (11.65) | 35 (11.44) | 26 (7.85) | 0.3911 |
| Past statin therapy, n (%) | 61 (16.18) | 21 (20.39) | 45 (14.71) | 38 (11.48) | 0.1104 |
| Past ischemic stroke, n (%) | 36 (9.55) | 13 (12.62) | 32 (10.46) | 27 (8.16) | 0.5482 |
| Past intracerebral hemorrhage, n (%) | 6 (1.59) | 3 (2.91) | 5 (1.63) | 2 (0.60) | 0.3391 |
| Treatment | |||||
| Dual antiplatelet therapy, n (%) | 185 (49.07) | 50 (48.54) | 174 (56.86) | 141 (42.60) | 0.0047 |
| Aspirin dosage: 162–200 mg, n (%) | 284 (75.33) | 82 (79.61) | 262 (85.62) | 268 (80.97) | 0.0096 |
| Concomitant therapy (argatroban or heparin), n (%) | 98 (25.99) | 38 (36.89) | 94 (30.72) | 132 (39.88) | 0.0008 |
| Statin, n (%) | 207 (54.91) | 59 (57.28) | 176 (57.52) | 210 (63.44) | 0.1381 |
| Antihypertensive, n (%) | 161 (42.71) | 62 (60.19) | 148 (48.37) | 211 (63.75) | ＜0.0001 |
| Clinical subtype | |||||
| Large artery atherosclerosis, n (%) | 56 (14.85) | 15 (14.56) | 42 (13.73) | 53 (16.01) | 0.8812 |
| Lacunar, n (%) | 192 (50.93) | 49 (47.57) | 148 (48.37) | 141 (42.60) | 0.1676 |
| Branch atheromatous disease, n (%) | 40 (10.61) | 64 (19.34) | 15 (14.56) | 64 (19.34) | 0.0103 |
| Other determined etiology, n (%) | 53 (14.06) | 13 (12.62) | 30 (9.80) | 35 (10.57) | 0.3109 |
| Undetermined etiology, n (%) | 36 (9.55) | 11 (10.68) | 34 (11.11) | 38 (11.48) | 0.8523 |
| CCA maximum IMT, mm (mean±SD) | 1.19±0.74 | 1.31±0.63 | 1.36±0.84 | 1.42±0.71 | 0.0012 |
| (n= 333) | (n= 91) | (n= 282) | (n= 302) |
SD, standard deviation; IQR, interquartile range; mRS, modified Rankin scale; NIHSS, National Institutes of Health Stroke Scale; IMT, intima– media thickness; CCA, common carotid artery.
| Variables | low-low (n= 359) | low-high (n= 95) | high-low (n= 292) | high-high (n= 305) | p value |
|---|---|---|---|---|---|
| Ischemic stroke/TIA recurrence within 3 months | 4 (1.11) | 4 (4.21) | 10 (3.42) | 16 (5.25) | 0.0242 |
| Age (years±SD) | 64.76±12.28 | 70.75±12.04 | 69.41±11.96 | 69.92±10.91 | ＜0.0001 |
| Male, n (%) | 272 (75.77) | 59 (62.11) | 190 (65.07) | 180 (59.02) | ＜0.0001 |
| Premorbid mRS, median (IQR) | 0 (0–0) | 0 (0–0) | 0 (0–0) | 0 (0–0) | 0.2326 |
| NIHSS score, median (IQR) | 2 (1–4) | 2 (1–4) | 2 (1–4) | 3 (1–4) | 0.1739 |
| Vascular risk factor | |||||
| Hypertension, n (%) | 254 (70.75) | 74 (77.89) | 232 (79.45) | 262 (85.90) | ＜0.0001 |
| Diabetes mellitus, n (%) | 101 (28.13) | 35 (36.84) | 91 (31.16) | 111 (36.39) | 0.0996 |
| Hyperlipidemia, n (%) | 180 (50.14) | 45 (47.37) | 133 (45.55) | 152 (49.84) | 0.6433 |
| Past aspirin therapy, n (%) | 33 (9.19) | 10 (10.53) | 34 (11.64) | 23 (7.54) | 0.3804 |
| Past statin therapy, n (%) | 59 (16.43) | 19 (20.00) | 43 (14.73) | 36 (11.80) | 0.1779 |
| Past ischemic stroke, n (%) | 35 (9.75) | 12 (12.63) | 31 (10.62) | 22 (7.21) | 0.3362 |
| Past intracerebral hemorrhage, n (%) | 6 (1.67) | 2 (2.11) | 5 (1.71) | 1 (0.33) | 0.3329 |
| Treatment | |||||
| Dual antiplatelet therapy, n (%) | 175 (48.75) | 46 (48.42) | 168 (57.53) | 125 (40.98) | 0.001 |
| Aspirin dosage: 162–200 mg, n (%) | 270 (75.21) | 76 (80.00) | 250 (85.62) | 245 (80.33) | 0.0121 |
| Concomitant therapy (argatroban or heparin), n (%) | 95 (26.46) | 36 (37.89) | 89 (30.48) | 120 (39.34) | 0.0025 |
| Statin, n (%) | 201 (55.99) | 55 (57.89) | 167 (57.19) | 193 (63.28) | 0.2594 |
| Antihypertensive, n (%) | 153 (42.62) | 57 (60.00) | 142 (48.63) | 195 (63.93) | ＜0.0001 |
| Clinical subtype | |||||
| Large artery atherosclerosis, n (%) | 53 (14.76) | 14 (14.74) | 40 (13.70) | 51 (16.72) | 0.7735 |
| Lacunar, n (%) | 185 (51.53) | 44 (46.32) | 144 (49.32) | 136 (44.59) | 0.3279 |
| Branch atheromatous disease, n (%) | 36 (10.03) | 13 (13.68) | 49 (16.78) | 55 (18.03) | 0.0177 |
| Other determined etiology, n (%) | 49 (13.65) | 13 (13.68) | 28 (9.59) | 30 (9.84) | 0.2648 |
| Undetermined etiology, n (%) | 36 (10.03) | 11 (11.58) | 31 (10.62) | 33 (10.82) | 0.9718 |
| CCA maximum IMT, mm (mean±SD) | 1.20±0.75 | 1.34±0.64 | 1.34±0.80 | 1.43±0.73 | 0.0022 |
| (n= 320) | (n= 83) | (n= 270) | (n= 278) |
TIA, transient ischemic attack; SD, standard deviation; IQR, interquartile range; mRS, modified Rankin scale; NIHSS, National Institutes of Health Stroke Scale; IMT, intima–media thickness; CCA, common carotid artery.
In univariate analyses for the four groups (low-low, low-high, high-low, and high-high group) using the PP thresholds of 69 mmHg on admission and 24 h after admission (Table 6), as compared with the low-low group, the high-high group showed a significantly higher risk of neurological deterioration within 14 days (OR 3.22, 95% CI 1.83–5.67; p＜0.0001). Meanwhile, the low-high group (OR 2.38; 95% CI 1.09–5.22; p=0.0299) and high-low group (OR 3.26; 95% CI 1.84–5.78; p＜0.0001) also showed significantly higher risks of neurological deterioration within 14 days. The risk of recurrent ischemic stroke/TIA within 3 months was also significantly higher in the high-high group than in the low-low group (OR 4.91; 95% CI 1.62–14.86; p=0.0048). The low-high and high-low groups tended to show higher recurrence rates than in the low-low group, but the differences were deemed insignificant.
| PP on admission＊ | PP at 24 h＊ | Present patients, n (%) | OR | 95%CI | p value | |
|---|---|---|---|---|---|---|
| Neurological deterioration within 14 days | ||||||
| low-low | ＜69 | ＜69 | 18 (4.77) | 1 | ||
| low-high | ＜69 | ≥ 69 | 11 (10.68) | 2.38 | 1.09–5.22 | 0.0299 |
| high-low | ≥ 69 | ＜69 | 43 (14.05) | 3.26 | 1.84–5.78 | ＜0.0001 |
| high-high | ≥ 69 | ≥ 69 | 46 (13.90) | 3.22 | 1.83–5.67 | ＜0.0001 |
| Recurrent ischemic stroke/TIA within 3 months | ||||||
| low-low | ＜69 | ＜69 | 4 (1.11) | 1 | ||
| low-high | ＜69 | ≥ 69 | 4 (4.21) | 3.9 | 0.96–15.90 | 0.0576 |
| high-low | ≥ 69 | ＜69 | 10 (3.42) | 3.15 | 0.98–10.14 | 0.0548 |
| high-high | ≥ 69 | ≥ 69 | 16 (5.25) | 4.91 | 1.62–14.86 | 0.0048 |
TIA, transient ischemic attack; CI, confidence interval; OR, odds ratio; PP, pulse pressure.
＊The threshold PP was 69 mmHg, calculated from pulse pressure on admission using receiver operating characteristic curves.
Finally, we performed a multivariate analysis including clinical factors that exhibited values of p＜0.1 in univariate analyses for patient characteristics. Clinical factors associated with neurological deterioration within 14 days included age; sex; premorbid mRS; NIHSS on admission; diabetes mellitus; past cerebral hemorrhage; treatment, concomitant therapy (argatroban or heparin); and clinical subtype (lacunar disease and BAD). Factors associated with recurrent ischemic stroke/TIA within 3 months included age, past aspirin therapy, past ischemic stroke, and clinical subtype (large-artery atherosclerosis and lacunar disease). In the multivariate analysis, in comparison to the low-low group, the high-high group has exhibited a significantly higher risk of neurological deterioration within 14 days (OR 2.09, 95% CI 1.12–3.91; p=0.0209) (Table 7). Compared with the low-low group, the high-high group has exhibited a significantly higher risk of recurrent ischemic stroke/TIA within 3 months (OR 4.80, 95% CI 1.62–14.86; p=0.0064). In addition, the risk of neurological deterioration within 14 days was significantly higher in the high-low group than that in the low-low group (OR 2.70; 95% CI 1.44–5.05; p=0.0019), although no significant difference was evident between the low-high and low-low groups.
| PP on admission＊ | PP at 24 h＊ | OR | 95%CI | p value | |
|---|---|---|---|---|---|
| Neurological deterioration within 14 days | |||||
| low-low | ＜69 | ＜69 | 1 | ||
| low-high | ＜69 | ≥ 69 | 1.59 | 0.67–3.78 | 0.2920 |
| high-low | ≥ 69 | ＜69 | 2.70 | 1.44–5.05 | 0.0019 |
| high-high | ≥ 69 | ≥ 69 | 2.09 | 1.12–3.91 | 0.0209 |
| Recurrent ischemic stroke/TIA within 3 months | |||||
| low-low | ＜69 | ＜69 | 1 | ||
| low-high | ＜69 | ≥ 69 | 3.65 | 0.96–15.90 | 0.0743 |
| high-low | ≥ 69 | ＜69 | 3.02 | 0.98–10.14 | 0.0678 |
| high-high | ≥ 69 | ≥ 69 | 4.8 | 1.62–14.86 | 0.0064 |
TIA, transient ischemic attack; CI, confidence interval; OR, odds ratio; PP, pulse pressure.
＊The threshold pulse pressure was 69 mmHg, calculated from pulse pressure on admission using receiver operating characteristic curves.
In this study, we have demonstrated that in patients with non-cardiogenic ischemic stroke admitted within 48 h after onset, high PP in the acute phase was associated with both neurological deterioration within 14 days and recurrent ischemic stroke/TIA within 3 months. In particular, high PP extending for 24 h after admission has showed significant associations with 14-day neurological deterioration and 3-month recurrent ischemic stroke/TIA.
Several epidemiological studies have already investigated the relationships between PP and stroke occurrence. With regard to the primary prevention of cerebrovascular diseases, one meta-analysis of seven studies found that a 10 mmHg increase in PP was associated with a six-fold increase in total mortality and a sevenfold increase in cardiovascular mortality8). In a follow-up study by Bruce et al. (mean follow-up, 6.7 years), an association between baseline PP and stroke was determined. That specific study revealed that higher baseline PP and systolic and diastolic blood pressures were all associated with greater risk of stroke14). With regard to the chronic phase of ischemic stroke, another study reported that high PP was associated with stroke recurrence and death within 3 months only among patients ≥ 60 years of age, but not among patients ＜60 years of age15). Although these studies suggested high PP as an important predictor of stroke occurrence in both primary and secondary prevention, particularly among the elderly, the clinical impact of high PP in the acute phase of stroke has not yet been sufficiently clarified.
A few reports have already examined PP in the acute phase of stroke. One prospective study showed that high baseline PP in patients with ischemic stroke admitted within 48 h was associated with death or neurological deterioration within 10 days4). In a retrospective study, it was found that mean PP during the first 7 days among patients with acute stroke was associated with both poor prognosis at discharge and death within 30 days16). Other studies have observed similar tendencies for neurological deterioration or recurrent strok9, 10). In another retrospective study, it was reported that PP measured within 3 days after the onset of acute ischemic stroke was associated with cardiovascular events and stroke recurrence, showing a J-shaped relationship for this association11). These results were similar to several studies that reported J- or U-shaped associations between post-stroke blood pressure and clinical outcomes17-19). Another investigation reported a reverse J-shaped relationship between PP and unfavorable outcomes12). As acute PP can be highly variable over time, prognosis can vary depending on the date of PP measurement. Our sub-analysis of the prospective ADS trial differed from previous studies in that we investigated PP changes in the acute phase for the first time. In our study, higher PP tended to be associated with higher risks of neurological deterioration within 14 days and recurrent ischemic stroke/TIA within 3 months. The key finding of this report was that high PP both on admission and at 24 h after admission correlated with higher risks of neurological deterioration and ischemic stroke recurrence. For neurological deterioration within 14 days, multivariate analysis showed the high-low group, in addition to the high-high group, has significantly higher rates than the low-low group. These results can be helpful in predicting the risk of neurological deterioration in the early stages of ischemic stroke when PP is noted high (≥ 69 mmHg) earlier in ischemic stroke onset, in addition to persistently high PP. With recurrent ischemic stroke/TIA within 3 months, the groups without persistently high PP (high-low and low-high groups) also showed worsening tendencies, but the differences were deemed insignificant. These results suggest that only persistently high PP (≥ 69 mmHg) in the early phase of stroke onset may represent a useful predictor of stroke recurrence.
The efficacy and optimal PP or blood pressure to use as thresholds for starting antihypertensive treatment in the acute phase of stroke are yet to be established1). However, a subgroup analysis of the China Antihypertensive Trial in Acute Ischemic Stroke trial showed that acute ischemic stroke patients who received antihypertensive medications between 24 and 48 h after symptom onset may experience reduced rates of death and major disability, recurrent stroke, or vascular events at 3 months after onset20). As per the results of our study, patients with low PP in the acute phase also exhibited better prognosis and less neurological deterioration, but we could not confirm the efficacy of antihypertensive treatment.
Regarding the mechanisms by which increased PP is associated with neurological deterioration and stroke recurrence, one study measured 24-h blood pressure during the acute phase of stroke and showed that systolic blood pressure, PP, and heart rate were high in patients with cerebral edema21). As for the mechanisms underlying short-term cerebral edema, one study using an animal model found that an increase in blood pressure may cause breakdown of the blood–brain barrier22). In our study, no difference in terms of neurological deterioration within 14 days was evident between the low-high and low-low groups. Still, the frequency of neurological deterioration within 14 days was significantly greater in both the high-low and high-high groups than that in the low-low group. This result may indicate that acute reactive high PP is involved in the short-term neurological deterioration by exacerbating cerebral edema.
On the other hand, chronic increases in PP have been attributed to the following: (1) progressive stiffening of central vessels resulting from changes in endothelial cells and vessel wall structure and (2) excessive wave reflection from high-resistance peripheral vessels23). As a result, chronic elevation of PP may lead to dysfunction of the blood–brain barrier, neurodegeneration, and dementia. Concerning cerebral arteries, another report suggested that 24-h blood pressure fluctuations in ischemic stroke patients are associated with intracranial intraplaque hemorrhage24). In our study, carotid artery plaque was noted to be thicker in the high-high group than in the low-low group (Tables 4, 5). In addition, the high-high group included a greater proportion of patients with hypertension and BAD; this only suggests that chronic arteriosclerosis may be a prominent feature among such patients. Patients with acutely high PP may thus present with chronic brain damage and greater susceptibility to recurrent stroke.
Even among patients with the same subtype, such as atherothrombotic stroke, and BAD, some may experience exacerbation, and some may not. This fact suggests that local brain pathology and degree of arteriosclerosis may differ even within the same disease subtype; PP may be one of the data points contributing to such differences. The risk of exacerbation may thus be amenable to prediction using acute-phase PP.
This study has several limitations. First, no distinction between small and large vessel diseases was considered. Reactive increases in blood pressure in the acute phase are reportedly marked in small vessel diseases25). Second, aortic, and carotid plaques are often used as indicators of arteriosclerosis. In this analysis, carotid artery occlusion, or stenosis was assessed in all patients. Moreover, at least one of the following should be performed: carotid ultrasonography, magnetic resonance angiography, or cerebral angiography. However, not all patients underwent assessment of IMT via carotid ultrasonography5). TEE was also used to diagnose the type of stroke and assessment of aortic plaque, but only a relatively small proportion of cases were evaluated. Third, the effects of prior or newly initiated antihypertensive therapy in the acute phase and the effects of interventions for PP or blood pressure were not considered. Fourth, the results of this post-hoc analysis have focused on antithrombotic drugs, and blood pressure measurement conditions were not strictly standardized. Therefore, the PP threshold of 69 mmHg may be suboptimal for general application as a prognostic threshold in all ischemic stroke patients. However, high PP in the acute phase was significantly associated with neurological deterioration and recurrence; thus, persistently high PP in the acute phase of stroke should be further evaluated in terms of neurological deterioration and recurrent ischemic stroke/TIA. Future studies with standardized methods of measuring blood pressure are deemed warranted. Fifth, our study evaluated only short-term outcomes at 3 months. Finally, as noted in other studies, systolic, and diastolic blood pressures fluctuated similar to PP, but blood pressure components other than PP that may also affect prognosis and recurrences were not evaluated in this study.
High PP both on admission and 24 h after admission in the acute phase of non-cardiogenic ischemic stroke may predict the risk of stroke recurrence and short-term neurological deterioration. However, optimal blood pressure and PP control in the acute phase of non-cardiogenic stroke remain controversial. In this study, we examined the relationship between PP and the clinical course of non-cardiogenic ischemic stroke. As other reports have shown associations between PP and ischemic stroke recurrence and prognosis regardless of ischemic stroke subtype9-12), the inclusion of cardiogenic cerebral embolism in future studies may be worthwhile.
The authors thank all clinical staff who assisted with the study.
This study received funding from Kawasaki Medical School and Nippon Medical School and an endowment from Otsuka Pharmaceutical Corporation (Tokyo, Japan), which markets cilostazol (Pletaal). Otsuka Pharmaceutical Corporation was blinded in the study design, data collection, and data analysis. This was stated on the informed consent form approved by each institutional review board, and potential conflicts of interest were appropriately managed at all participating institutions.
Dr. Fujimoto reports having received lecture fees and speaker’s fees from Nippon Boehringer Ingelheim, Co., Bayer Yakuhin, Pfizer Japan Inc., Daiichi Sankyo Co., Eisai Co., and Bristol-Myers Squibb Co. Dr. Kimura reports having received lecture fees and speaker’s fees from Otsuka Pharmaceutical Co., Bayer Yakuhin, and Sanofi K.K.
