- Email: [email protected]
Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke
Journal of Clinical Neuroscience xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Case study
Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke Yu-Fang Chou, Wei-Chieh Weng, Wen-Yi Huang ⇑ Department of Neurology, Chang-Gung Memorial Hospital, Keelung Branch No. 222, Mai-Jin Road, Keelung 204, Taiwan Department of Medicine, College of Medicine, Chang Gung University, 259 Wen-Hwa 1st Road, Kwei-Shan, Tao-Yuan 333, Taiwan
a r t i c l e
i n f o
Article history: Received 28 February 2017 Accepted 21 June 2017 Available online xxxx Keywords: Gastrointestinal bleeding Stroke outcome Mortality First-ever ischemic stroke
a b s t r a c t The influence of gastrointestinal bleeding on clinical presentation and outcomes of patients with acute ischemic stroke remains controversial. We investigate the effect of gastrointestinal bleeding on the outcomes of patients with acute, first-ever ischemic stroke. We enrolled 934 patients with acute, first-ever ischemic stroke and followed up them for 3 years. Patients were divided into 2 groups according to the presence or absence of gastrointestinal bleeding during acute stroke stage. Clinical presentation, stroke risk factors, laboratory data, co-morbidities, and outcomes were recorded. Seventy-six (8.1%) patients had gastrointestinal bleeding at admission. The prevalence of old age, atrial fibrillation, and previous transient ischemic attack was higher in patients with gastrointestinal bleeding (P < 0.001, P = 0.038, and P = 0.018, respectively). Total anterior circulation syndrome occurred more frequently among patients with gastrointestinal bleeding (P < 0.001). The mean length of acute ward stay, initial impaired consciousness, and stroke in evolution were higher in patients with gastrointestinal bleeding (P < 0.001, P < 0.001, and P < 0.001, respectively). The occurrence of pneumonia and dependent functional outcome were higher in patients with gastrointestinal bleeding (P < 0.001 and P < 0.001, respectively). A multivariate Cox regression analysis revealed that gastrointestinal bleeding is a significant risk factor for 3-year all-cause mortality (hazard ratio = 2.76; 95% confidence interval = 1.61–4.72; P < 0.001). In conclusion, gastrointestinal bleeding is associated with increased risk of 3-year mortality in patients with acute, first-ever ischemic stroke. Prophylactic therapies for gastrointestinal bleeding might improve ischemic stroke outcome. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction Gastrointestinal (GI) bleeding is a well-known complication in acute stroke patients, and it may interfere with the treatment for ischemic stroke, such as antiplatelet or anticoagulant therapies. The incidence of GI bleeding during the acute phase of stroke is about 0.1–8.0%, depending on the stroke subtype [1–5]. The clinical relevance of GI bleeding on stroke outcome remains controversial. A previous study suggested that GI bleeding after stroke is rarely severe and may not contribute significantly to mortality; however, the number of patients enrolled in this study was relatively small [5]. Recently, Rumalla et al. suggested that GI bleeding following acute ischemic stroke may be associated with a poor functional outcome and in-hospital mortality [6]. Another study reported that GI bleeding after acute ischemic stroke is
associated with an increased risk of death and severe dependence; however, the follow-up time in this study was only 6 months [3]. The long-term outcomes of GI bleeding are poorly understood, despite it being the most frequent cause of gastroenterology admission to acute medicine ward. Recently, Crooks et al. have reported that patients who have had a previous nonvariceal upper GI bleeding have higher long-term all-cause mortality rates [7]. To our knowledge, no previous studies have investigated the influence of GI bleeding during the acute stroke stage on the long-term stroke outcome and mortality. The aim of this study is to investigate: (1) the association between GI bleeding in acute stroke stage and clinical presentations, or acute complications, in patients with acute, first-ever stroke and (2) the association between GI bleeding in the acute stroke stage and clinical outcomes, including the functional outcomes and 3-year mortality.
⇑ Corresponding author at: Department of Neurology, Chang-Gung Memorial Hospital, Keelung Branch, No. 222, Mai-Jin Road, Keelung 204, Taiwan. E-mail address: [email protected] (W.-Y. Huang). http://dx.doi.org/10.1016/j.jocn.2017.06.068 0967-5868/Ó 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Chou Y-F et al. Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke. J Clin Neurosci (2017), http://dx.doi.org/10.1016/j.jocn.2017.06.068
2
Y.-F. Chou et al. / Journal of Clinical Neuroscience xxx (2017) xxx–xxx
2. Materials and methods This clinical study followed the Declaration of Helsinki and was approved by the Medical Ethics Committee of Chang Gung Memorial Hospital (CGMH), Taipei, Taiwan. 2.1. Study patients All patients enrolled in this study were recruited from the Stroke Unit of the Department of Neurology at CGMH from January 1, 2001, to December 31, 2003. Only patients with first-ever ischemic stroke were enrolled. The clinical diagnosis of acute ischemic stroke was performed according to the World Health Organization criteria. The diagnosis was further confirmed by brain computed tomography or magnetic resonance imaging (MRI) scan [8]. Patients with symptoms lasting less than 24 h and without evidence of acute cerebral infarction in the MRI were diagnosed with transient ischemic attack (TIA) and were excluded. Co-morbidities were determined after an in-depth review of the medical records. Hypertension was defined as known hypertension diagnosed by a clinician, or systolic blood pressure >160 mmHg and/or diastolic blood pressure >95 mmHg on two different occasions. Diabetes mellitus (DM) was diagnosed in patients with previously treated DM or in patients with fasting plasma glucose levels 126 mg/dl, a 2-h value in the oral glucose tolerance test or a random plasma glucose concentration 200 mg/dl, in the presence of symptoms. Hyperlipidemia was defined as a fasting blood cholesterol level 200 mg/dl and/or a triglyceride level 200 mg/dl. Atrial fibrillation (AF) was diagnosed if it was present on a standard 12-lead electrocardiogram. Coronary artery disease (CAD) was diagnosed if there were past incidences of acute myocardial infarction or angina pectoris. Congestive heart failure (CHF) was present if the patient was previously diagnosed by a cardiologist or confirmed by cardiac echo. 2.2. Definition of gastrointestinal bleeding in acute stroke stage A GI hemorrhage event was defined according to Davenport et al. as any episode of fresh blood or coffee ground material in nasogastric aspirate, hematemesis, melena or bloody stool [9]. The cause and origin of the bleeding were investigated using endoscopy in most of the patients if they could tolerate the procedure. Patients were divided into 2 groups according to the presence or absence of GI bleeding during the acute stroke. 2.3. Definition and clinical subtypes of ischemic stroke Clinical subtypes of ischemic stroke were rated according to the Oxfordshire Community Stroke Project classification. The following subtypes were adopted: Partial Anterior Circulation Syndrome (PACS), Total Anterior Circulation Syndrome (TACS), Posterior Circulation Syndrome (POCS), and Lacunar Syndrome (LACS) [10]. The clinical course of acute stroke stage, mean length of acuteward stay, mortality rates during and following acute ward stay, and frequency of medical complications were monitored. The definition ‘‘stroke in evolution” refers to a neurologic deficit that progresses within 7 days after the stroke onset [11]. Functional outcomes upon discharge were assessed according to the modified Rankin Scale (mRS) [12]. The ‘‘functionally dependent” condition was defined as having an mRS score of 3, 4 or 5. 2.4. Follow-up Patients were followed up for 3 years after the initial assessment. Follow-up consisted of clinical examinations performed 1
and 3 months after the first stroke and then every 3 months. Clinical examinations during follow-up included history taking, physical and neurological examinations, and mRS score assessment. New major medical problems (e.g., death, recurrent cerebral infarction, cerebral hemorrhage, epilepsy, cancer, cardiovascular diseases, head injury, etc.) were recorded during follow-up. The end point was death during 3-year follow up. Every death occurring during the follow-up was reviewed. 2.5. Statistical analysis Continuous variables such as age and laboratory measurement values were expressed as median and interquartile range because the values were not normally distributed, and mean length of stay in the acute ward was expressed as mean ± SD. Categorical variables were expressed as a number, or percentage, for each item. The 2 patient groups were compared using chi-square, MannWhitney U or Student t test. The independent associations between the variables and the probability of having GI bleeding were analyzed using logistic regression. All variables with a P < 0.1 in the univariate logistic regression entered a stepwise, backward multivariate logistic regression. The Cox proportional hazards model was used to determine the significance of each variable in predicting the 3-year all-cause mortality. A univariate Cox model, assessing all previously identified variables, was used to measure hazard ratio (HR) for mortality. A backward, stepwise multivariate Cox regression model was also used to identify the risk factors for the 3-year mortality. All statistical analyses were performed with IBM SPSS statistics 19 for Windows. 3. Results 3.1. Patient characteristics A total of 934 patients were enrolled in the study. The median age was 70 (62–78) years. At baseline, 76 (8.1%) patients had GI bleeding at admission. Age was significantly higher in the group with GI bleeding. Prevalence of AF and previous TIA were significantly higher among patients with GI bleeding. Upon clinical presentation of stroke, TACS occurred more frequently in the group with GI bleeding, whereas LACS occurred more frequently in the group without GI bleeding. Additionally, hemoglobin and estimated glomerular filtration rate (eGFR) (ml/min/1.73 m2) were significantly lower, whereas WBC was higher, in the group with GI bleeding. Among patients with GI bleeding, 37 patients used aspirin, 32 patients use clopidogrel, 2 patients used warfarin, 5 patients did not use antiplatelet or anticoagulant (Table 1). 3.2. Determinants of GI bleeding in patients with acute, first-ever ischemic stroke After adjusting for the potential risk factors (P < 0.1 in univiarate logistic regression) in a backward, stepwise multivariate logistic regression, only anemia, TIA and initial impaired consciousness were positively associated with patients having GI bleeding during acute stroke (Table 2). 3.3. Clinical course in patients with an acute stage of first-ever ischemic stroke The mean length of acute ward stay, initial impaired consciousness, and stroke in evolution were significantly higher in the group with GI bleeding. The occurrence of pneumonia and urinary tract infection (UTI) were significantly higher in patients with GI bleeding. Furthermore, the dependent functional status (mRS score =3)
Please cite this article in press as: Chou Y-F et al. Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke. J Clin Neurosci (2017), http://dx.doi.org/10.1016/j.jocn.2017.06.068
3
Y.-F. Chou et al. / Journal of Clinical Neuroscience xxx (2017) xxx–xxx Table 1 Demographic and clinical characteristics of first-ever ischemic stroke patients, divided according to the presence/absence of GI bleeding. GI bleeding (n = 76)
No GI bleeding (n = 858)
Odds Ratio (95% CI)
P value
Age (years) Female
75.5 (67.5–82.8) 29 (38.2%)
70.0 (63–77) 413 (48.1%)
0.67 (0.41–1.08)
<0.001* 0.006
Risk factors Hypertension Diabetes mellitus Smoking Hyperlipidemia Coronary artery disease Atrial fibrillation Previous transient ischemic attack Congestive heart failure Peripheral vascular disease
57 (75.0%) 36 (47.4%) 27 (35.5%) 23 (30.3%) 6 (7.9%) 17 (22.4%) 9 (11.8%) 2 (2.6%) 0 (0.0%)
616 (71.8%) 323 (37.6%) 278 (32.4%) 347 (40.4%) 53 (6.2%) 119 (13.9%) 42 (4.9%) 17 (2.0%) 4 (0.5%)
1.18 1.48 1.15 0.64 1.30 1.79 2.61 1.34 0.64
(0.69–2.02) (0.93–2.38) (0.70–1.88) (0.39–1.06) (0.54–3.13) (1.01–3.17) (1.22–5.59) (0.30–5.90) (0.39–1.06)
0.326 0.065 0.330 0.051 0.347 0.038* 0.018* 0.467 0.712
Clinical syndromes TACS PACS LACS POCS
27 22 12 15
92 (10.7%) 253 (29.5%) 363 (42.3%) 149 (17.4%)
4.59 0.97 0.26 1.17
(2.74–7.70) (0.58–1.63) (0.14–0.48) (0.65–2.12)
<0.001* 0.519 <0.001* 0.349
Lab data Hemoglobin WBC eGFR
13.0 (11.4–14.8) 8600 (6700–11,400) 66.0 (49.0–75.8)
(35.5%) (28.9%) (15.8%) (19.7%)
0.010* <0.001* <0.001*
13.6 (12.5–14.7) 7100 (5800–9000) 74.1 (57.6–87.9)
CI = confidence interval; GI = gastrointestinal; TACS = total anterior circulation syndrome; PACS = partial anterior circulation syndrome; LACS = lacunar syndromes; POCS = posterior circulation syndrome; MCV = mean corpuscular volume; WBC = white blood cells; eGFR = estimated glomerular filtration rate (ml/min/1.73 m2). Data are presented as median (interquartile range) or n (%). * P < 0.05, Chi-square or Mann-Whitney U test.
Table 2 Logistic regression for predictors of GI bleeding in patients with first-ever ischemic stroke. Variables Age (year) Anemia Hypertension Diabetes mellitus Transient ischemic attack Initial impaired consciousness Hyperlipidemia Coronary artery disease Congestive heart failure Atrial fibrillation Smoking eGFR < 60
Univariate Logistic Regression, Risk Ratio (95% CI)
P value
Stepwise Multivariate Logistic Regression, Risk Ratio (95% CI)
P value
1.99 (1.56–3.44)
0.013y
2.95 (1.30–6.78) 5.32 (3.06–9.23)
0.010y <0.001*
*
1.04 2.56 1.18 1.48 2.61 6.24
(1.02–1.07) (1.57–4.16) (0.69–2.02) (0.93–2.38) (1.22–5.59) (3.78–10.30)
<0.001 <0.001* 0.551 0.101 0.014* <0.001*
0.64 1.30 1.34 1.79 1.15 1.74
(0.39–1.06) (0.54–3.13) (0.30–5.90) (1.01–3.17) (0.70–1.88) (1.07–2.82)
0.084* 0.558 0.701 0.047* 0.578 0.025*
CI = confidence interval; eGFR = estimated glomerular filtration rate (ml/min/1.73 m2); GI = gastrointestinal. P < 0.05 for the univariate logistic regression. y P < 0.05 for the stepwise multivariate logistic regression.
*
Table 3 Clinical course of acute stage, mortality, and stroke recurrence within 3 years after onset of first-ever ischemic stroke, divided according to the presence/absence of GI bleeding. GI bleeding (n = 76)
No GI bleeding (n = 858)
Odds ratio (95% CI)
P value
Mean length of stay in the acute medicine ward (days) Initial impaired consciousness
29.13 ± 22.43 33 (43.4%)
13.36 ± 10.68 94 (11.0%)
15.77 (12.94–18.60) 6.24 (3.38–10.30)
<0.001* <0.001*
Course of acute stroke stage In evolution
37 (48.7%)
196 (22.8%)
3.20 (1.99–5.16)
<0.001*
Complications Pneumonia Urinary tract infection mRS score 3 upon discharge Stroke recurrence Death
37 (48.7%) 22 (28.9%) 71 (93.4%) 7 (9.2%) 18 (23.7%)
63 (7.3%) 95 (11.1%) 530 (61.8%) 145 (16.9%) 63 (7.3%)
11.97 (7.13–20.09) 3.27 (1.91–5.61) 8.79 (3.51–21.99) 0.50 (0.23–1.11) 3.92 (2.18–7.05)
<0.001* <0.001* <0.001* 0.051 <0.001*
GI = gastrointestinal; mRS = modified Rankin scale. Data are presented as mean ± SD or n (%). * P < 0.05, chi-square or Student t tests.
Please cite this article in press as: Chou Y-F et al. Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke. J Clin Neurosci (2017), http://dx.doi.org/10.1016/j.jocn.2017.06.068
4
Y.-F. Chou et al. / Journal of Clinical Neuroscience xxx (2017) xxx–xxx
following discharge was higher in patients with GI bleeding. The rate of mortality within 3 years after stroke onset was significantly higher among patients with GI bleeding (Table 3). 3.4. Kaplan-Meier survival analysis of 3-year mortality in patients with acute, first-ever stroke At the end of the 3-year observation period, 81 patients had died (81/934 = 8.67%), including 18 (23.7%) in the group with GI bleeding and 63 (7.3%) in the group without GI bleeding. Among the patients with GI bleeding, 8 (44.4%) died within acute stroke stage, 3 (16.7%) died of infection, 3 (16.7%) died of cardiovascular disease, 1 (5.6%) died of recurrent stroke, 1 (5.6%) died of complication due to rt-PA infusion, 1 (5.6%) died of sudden apnea with an unknown cause, 1 (5.6%) died of cancer, and 1 (5.6%) died of uremia. The Kaplan-Meier survival analysis revealed that the group with GI bleeding had a higher rate of mortality than the group without GI bleeding (Fig. 1). 3.5. Cox regression multivariate analysis for 3-year mortality in patients with acute, first-ever stroke The univariate Cox regression indicated that age, reduced eGFR, CAD, AF, CHF, lower lipid value, TACS, and GI bleeding were potential risk factors (P < 0.1) for a 3-year mortality. These variables entered the backward, stepwise multivariate Cox proportional hazards model. We found that GI bleeding (HR = 2.76; 95% CI = 1.61– 4.72; P < 0.001) was a significant risk factor for a 3-year all-cause mortality in these patients even after adjusting for these variables (Table 4). 4. Discussion This study shows, for the first time, the association between GI bleeding in the acute stroke stage and increased 3-year mortality in patients with first-ever ischemic stroke. This association was significant even after adjusting for the established clinical predictors of adverse outcomes. To our knowledge, this is the first study to investigate the influence of GI bleeding in the acute stroke stage on long-term mortality.
Fig. 1. Kaplan-Meier estimates of patient survival (all-cause mortality) during the 3-year study period. Symbols are: (—) No GI bleeding; () GI bleeding. Log Rank P < 0.001.
In this study, we found that patients were significantly older and the stroke severity was significantly higher in the group with GI bleeding; furthermore, the prevalence of AF was significantly higher in these patients, which were similar to previous studies [1,3,6,13]. This may be due to the fact that AF is often associated with increased stroke severity [14,15], and a severe stroke increases the risk of stress-derived peptic ulcers. Our results show that the hemoglobin level was significantly lower in the group with GI bleeding. This finding may be due to that small amount of bleeding may frequently occur before the stroke episode in patients with GI bleeding [16]. In this study, the frequency of reduced eGFR is significantly higher in patients with GI bleeding. Previous study showed that patients with chronic kidney disease have higher risk of both upper and lower GI bleeding [17]. Our result also revealed that WBC count is significantly higher in patients with GI bleeding. A previous study has shown that leukocytosis is common in patients with upper GI bleeding, it may reflect the severity of the bleeding episode, and is associated with a complicated course [18]. The mean length of acute ward stay in our study was significantly longer in patients with GI bleeding than in patients without GI bleeding. We also showed that initial impaired consciousness was more frequent in patients with GI bleeding, confirming the findings of a previous study [1,13]. Our study also revealed that the occurrence of pneumonia was significantly higher in patients with GI bleeding. A previous study showed that respiratory complications, such as pneumonia, are common in critically ill medical patients with serious acute upper GI bleeding and are associated with a poor outcome [19]. Another study revealed that stress ulcer prophylactic agents are effective in decreasing bleeding, and the use of them is associated with a lower rate of nosocomial pneumonia and mortality [20]. The relationship between GI bleeding and the neurological course of acute ischemic stroke remains unclear. We observed a significant association between GI bleeding and stroke in evolution. There are several hypotheses regarding the mechanisms for the association between GI bleeding and poor clinical outcomes. Bleeding contributes to hemodynamic insufficiency and leads to the discontinuation of antithrombotic treatment, which may result in a prothrombotic state [21]. These factors may then lead to the deterioration of neurological symptoms and poor functional outcome. Whether treating GI bleeding will improve stroke outcome remains unclear. The use of prophylactic proton pump inhibitors or H2 antagonists may reduce the risk of upper GI bleeding in acute stroke stage and improve the tolerance of antiplatelet or anticoagulant. In addition, treating GI bleeding will prevent hemodynamic instability, which may lead to brain hypoperfusion. Furthermore, previous studies suggested that severe upper GI bleeding increase the risk of respiratory complication (e.g., pneumonia, etc), which is associated with increased mortality [19,22]. Therefore, treating GI bleeding may decrease the occurrence of respiratory complication, which could improve stroke outcome. Further studies are required to elucidate the mechanisms for the associations. This study has several limitations. First, we did not evaluate the stroke severity using the National Institutes of Health Stroke Scale score because a number of patients did not receive this scoring while in admission. Instead, we used the OSCP classification to evaluate stroke severity; LACS was regarded as a less severe and TACS as a more severe stroke. We have included TACS in the Cox regression analysis. Second, we included patients with ischemic stroke only; hence, our results do not reflect the impact of GI bleeding on other types of stroke. However, our study represents a homogenous population. Third, we could not detect all patients with GI bleeding because those who showed a gradual decrease in hemoglobin without hematemesis, or melena, were not
Please cite this article in press as: Chou Y-F et al. Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke. J Clin Neurosci (2017), http://dx.doi.org/10.1016/j.jocn.2017.06.068
5
Y.-F. Chou et al. / Journal of Clinical Neuroscience xxx (2017) xxx–xxx Table 4 Cox regression analysis of patient survival during the 3-year study period. Univariate Cox regression
Age (year-old) Gender eGFR < 60 (Yes) Recurrent stroke (Yes) Hypertension (Yes) Diabetes mellitus (Yes) Coronary artery disease (Yes) Atrial fibrillation (Yes) Congestive heart failure (Yes) Smoking (Yes) Hyperlipidemia (Yes) TACS GI bleeding
Multivariate Cox regression
Hazard Ratio (95% CI)
P value
Hazard Ratio (95% CI)
P value
1.04 0.71 1.99 0.99 1.02 1.23 2.42 2.49 2.82 0.93 0.53 2.87 3.59
<0.001* 0.135 0.002* 0.979 0.945 0.364 0.006* <0.001* 0.044* 0.764 0.011* <0.001* <0.001*
1.03 (1.00–1.05)
0.033y
2.48 (1.30–4.71) 1.94 (1.18–3.21)
0.006y 0.010y
(1.02–1.06) (0.46–1.11) (1.28–3.09) (0.49–2.00) (0.62–1.66) (0.79–1.90) (1.28–4.58) (1.54–4.03) (1.03–7.67) (0.58–1.49) (0.32–0.86) (1.76–4.69) (2.13–6.07)
<0.001y
2.76 (1.61–4.72) 2
CI = confidence interval; TACS = total anterior circulation syndrome; GI = gastrointestinal; eGFR = estimated glomerular filtration rate (ml/min/1.73 m ). P < 0.1 for the univariate Cox regression. y P < 0.05 for the multivariate Cox regression.
*
included. Forth, we did not have the information about the use of prophylaxis for GI bleeding such as H2 antagonists and proton pump inhibitors. 5. Conclusion GI bleeding is associated with an increased risk of 3-year mortality and poor functional outcome in patients with acute, firstever ischemic stroke. Prophylactic therapies for GI bleeding may reduce the risk of GI bleeding in the acute stroke stage and improve the outcomes of ischemic stroke. Conflicts of interest/disclosures
[9] [10]
[11] [12] [13]
[14]
There is no financial or other conflicts of interest in relation to this research and its publication.
[15]
References
[16]
[1] Hsu HL, Lin YH, Huang YC, et al. Gastrointestinal hemorrhage after acute ischemic stroke and its risk factors in Asians. Eur Neurol 2009;62:212–8. [2] Misra UK, Kalita J, Pandey S, et al. Predictors of gastrointestinal bleeding in acute intracerebral haemorrhage. J Neurol Sci 2003;208:25–9. [3] O’Donnell MJ, Kapral MK, Fang J, et al. Gastrointestinal bleeding after acute ischemic stroke. Neurology 2008;71:650–5. [4] Ogata T, Kamouchi M, Matsuo R, et al. Gastrointestinal bleeding in acute ischemic stroke: recent trends from the Fukuoka Stroke Registry. Cerebrovasc Dis Extra 2014;4:156–64. [5] Wijdicks EF, Fulgham JR, Batts KP. Gastrointestinal bleeding in stroke. Stroke 1994;25:2146–8. [6] Rumalla K, Mittal MK. Gastrointestinal bleeding in acute ischemic stroke: a population-based analysis of hospitalizations in the United States. J Stroke Cerebrovasc Dis 2016;25:1728–35. [7] Crooks CJ, Card TR, West J. Excess long-term mortality following non-variceal upper gastrointestinal bleeding: a population-based cohort study. PLoS Med 2013;10:e1001437. [8] Adams Jr HP, del Zoppo G, Alberts MJ, et al. Guidelines for the early management of adults with ischemic stroke: a guideline from the American
[17] [18]
[19]
[20]
[21]
[22]
Heart Association/American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups: the American Academy of Neurology affirms the value of this guideline as an educational tool for neurologists. Stroke 2007;38:1655–711. Davenport RJ, Dennis MS, Warlow CP. Gastrointestinal hemorrhage after acute stroke. Stroke 1996;27:421–4. Wlodek A, Sarzynska-Dlugosz I, Sandercock PA, et al. Agreement between the clinical Oxfordshire Community Stroke Project classification and CT findings in Poland. Eur J Neurol 2004;11:91–6. Gautier JC. Stroke-in-progression. Stroke 1985;16:729–33. Quinn TJ, Dawson J, Walters MR, et al. Reliability of the modified rankin scale: a systematic review. Stroke 2009;40:3393–5. Chen CM, Hsu HC, Chuang YW, et al. Study on factors affecting the occurrence of upper gastrointestinal bleeding in elderly acute stroke patients undergoing rehabilitation. J Nutr Health Aging 2011;15:632–6. Nolte CH, Rossnagel K, Jungehuelsing GJ, et al. Gender differences in knowledge of stroke in patients with atrial fibrillation. Prev Med 2005;41:226–31. Steger C, Pratter A, Martinek-Bregel M, et al. Stroke patients with atrial fibrillation have a worse prognosis than patients without: data from the Austrian Stroke registry. Eur Heart J 2004;25:1734–40. Kim BS, Li BT, Engel A, et al. Diagnosis of gastrointestinal bleeding: a practical guide for clinicians. World J Gastrointest Pathophysiol 2014;5:467–78. Luo JC, Leu HB, Hou MC, et al. Nonpeptic ulcer, nonvariceal gastrointestinal bleeding in hemodialysis patients. Am J Med 2013;126(264):e225–32. Chalasani N, Patel K, Clark WS, et al. The prevalence and significance of leukocytosis in upper gastrointestinal bleeding. Am J Med Sci 1998;315:233–6. Liebler JM, Benner K, Putnam T, et al. Respiratory complications in critically ill medical patients with acute upper gastrointestinal bleeding. Crit Care Med 1991;19:1152–7. Cook DJ. Stress ulcer prophylaxis: gastrointestinal bleeding and nosocomial pneumonia. Best evidence synthesis. Scand J Gastroenterol Suppl 1995;210:48–52. Eikelboom JW, Mehta SR, Anand SS, et al. Adverse impact of bleeding on prognosis in patients with acute coronary syndromes. Circulation 2006;114:774–82. Yu YJ, Weng WC, Su FC, et al. Association between pneumonia in acute stroke stage and 3-year mortality in patients with acute first-ever ischemic stroke. J Clin Neurosci 2016;33:124–8.
Please cite this article in press as: Chou Y-F et al. Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke. J Clin Neurosci (2017), http://dx.doi.org/10.1016/j.jocn.2017.06.068
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Case study
Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke Yu-Fang Chou, Wei-Chieh Weng, Wen-Yi Huang ⇑ Department of Neurology, Chang-Gung Memorial Hospital, Keelung Branch No. 222, Mai-Jin Road, Keelung 204, Taiwan Department of Medicine, College of Medicine, Chang Gung University, 259 Wen-Hwa 1st Road, Kwei-Shan, Tao-Yuan 333, Taiwan
a r t i c l e
i n f o
Article history: Received 28 February 2017 Accepted 21 June 2017 Available online xxxx Keywords: Gastrointestinal bleeding Stroke outcome Mortality First-ever ischemic stroke
a b s t r a c t The influence of gastrointestinal bleeding on clinical presentation and outcomes of patients with acute ischemic stroke remains controversial. We investigate the effect of gastrointestinal bleeding on the outcomes of patients with acute, first-ever ischemic stroke. We enrolled 934 patients with acute, first-ever ischemic stroke and followed up them for 3 years. Patients were divided into 2 groups according to the presence or absence of gastrointestinal bleeding during acute stroke stage. Clinical presentation, stroke risk factors, laboratory data, co-morbidities, and outcomes were recorded. Seventy-six (8.1%) patients had gastrointestinal bleeding at admission. The prevalence of old age, atrial fibrillation, and previous transient ischemic attack was higher in patients with gastrointestinal bleeding (P < 0.001, P = 0.038, and P = 0.018, respectively). Total anterior circulation syndrome occurred more frequently among patients with gastrointestinal bleeding (P < 0.001). The mean length of acute ward stay, initial impaired consciousness, and stroke in evolution were higher in patients with gastrointestinal bleeding (P < 0.001, P < 0.001, and P < 0.001, respectively). The occurrence of pneumonia and dependent functional outcome were higher in patients with gastrointestinal bleeding (P < 0.001 and P < 0.001, respectively). A multivariate Cox regression analysis revealed that gastrointestinal bleeding is a significant risk factor for 3-year all-cause mortality (hazard ratio = 2.76; 95% confidence interval = 1.61–4.72; P < 0.001). In conclusion, gastrointestinal bleeding is associated with increased risk of 3-year mortality in patients with acute, first-ever ischemic stroke. Prophylactic therapies for gastrointestinal bleeding might improve ischemic stroke outcome. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction Gastrointestinal (GI) bleeding is a well-known complication in acute stroke patients, and it may interfere with the treatment for ischemic stroke, such as antiplatelet or anticoagulant therapies. The incidence of GI bleeding during the acute phase of stroke is about 0.1–8.0%, depending on the stroke subtype [1–5]. The clinical relevance of GI bleeding on stroke outcome remains controversial. A previous study suggested that GI bleeding after stroke is rarely severe and may not contribute significantly to mortality; however, the number of patients enrolled in this study was relatively small [5]. Recently, Rumalla et al. suggested that GI bleeding following acute ischemic stroke may be associated with a poor functional outcome and in-hospital mortality [6]. Another study reported that GI bleeding after acute ischemic stroke is
associated with an increased risk of death and severe dependence; however, the follow-up time in this study was only 6 months [3]. The long-term outcomes of GI bleeding are poorly understood, despite it being the most frequent cause of gastroenterology admission to acute medicine ward. Recently, Crooks et al. have reported that patients who have had a previous nonvariceal upper GI bleeding have higher long-term all-cause mortality rates [7]. To our knowledge, no previous studies have investigated the influence of GI bleeding during the acute stroke stage on the long-term stroke outcome and mortality. The aim of this study is to investigate: (1) the association between GI bleeding in acute stroke stage and clinical presentations, or acute complications, in patients with acute, first-ever stroke and (2) the association between GI bleeding in the acute stroke stage and clinical outcomes, including the functional outcomes and 3-year mortality.
⇑ Corresponding author at: Department of Neurology, Chang-Gung Memorial Hospital, Keelung Branch, No. 222, Mai-Jin Road, Keelung 204, Taiwan. E-mail address: [email protected] (W.-Y. Huang). http://dx.doi.org/10.1016/j.jocn.2017.06.068 0967-5868/Ó 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Chou Y-F et al. Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke. J Clin Neurosci (2017), http://dx.doi.org/10.1016/j.jocn.2017.06.068
2
Y.-F. Chou et al. / Journal of Clinical Neuroscience xxx (2017) xxx–xxx
2. Materials and methods This clinical study followed the Declaration of Helsinki and was approved by the Medical Ethics Committee of Chang Gung Memorial Hospital (CGMH), Taipei, Taiwan. 2.1. Study patients All patients enrolled in this study were recruited from the Stroke Unit of the Department of Neurology at CGMH from January 1, 2001, to December 31, 2003. Only patients with first-ever ischemic stroke were enrolled. The clinical diagnosis of acute ischemic stroke was performed according to the World Health Organization criteria. The diagnosis was further confirmed by brain computed tomography or magnetic resonance imaging (MRI) scan [8]. Patients with symptoms lasting less than 24 h and without evidence of acute cerebral infarction in the MRI were diagnosed with transient ischemic attack (TIA) and were excluded. Co-morbidities were determined after an in-depth review of the medical records. Hypertension was defined as known hypertension diagnosed by a clinician, or systolic blood pressure >160 mmHg and/or diastolic blood pressure >95 mmHg on two different occasions. Diabetes mellitus (DM) was diagnosed in patients with previously treated DM or in patients with fasting plasma glucose levels 126 mg/dl, a 2-h value in the oral glucose tolerance test or a random plasma glucose concentration 200 mg/dl, in the presence of symptoms. Hyperlipidemia was defined as a fasting blood cholesterol level 200 mg/dl and/or a triglyceride level 200 mg/dl. Atrial fibrillation (AF) was diagnosed if it was present on a standard 12-lead electrocardiogram. Coronary artery disease (CAD) was diagnosed if there were past incidences of acute myocardial infarction or angina pectoris. Congestive heart failure (CHF) was present if the patient was previously diagnosed by a cardiologist or confirmed by cardiac echo. 2.2. Definition of gastrointestinal bleeding in acute stroke stage A GI hemorrhage event was defined according to Davenport et al. as any episode of fresh blood or coffee ground material in nasogastric aspirate, hematemesis, melena or bloody stool [9]. The cause and origin of the bleeding were investigated using endoscopy in most of the patients if they could tolerate the procedure. Patients were divided into 2 groups according to the presence or absence of GI bleeding during the acute stroke. 2.3. Definition and clinical subtypes of ischemic stroke Clinical subtypes of ischemic stroke were rated according to the Oxfordshire Community Stroke Project classification. The following subtypes were adopted: Partial Anterior Circulation Syndrome (PACS), Total Anterior Circulation Syndrome (TACS), Posterior Circulation Syndrome (POCS), and Lacunar Syndrome (LACS) [10]. The clinical course of acute stroke stage, mean length of acuteward stay, mortality rates during and following acute ward stay, and frequency of medical complications were monitored. The definition ‘‘stroke in evolution” refers to a neurologic deficit that progresses within 7 days after the stroke onset [11]. Functional outcomes upon discharge were assessed according to the modified Rankin Scale (mRS) [12]. The ‘‘functionally dependent” condition was defined as having an mRS score of 3, 4 or 5. 2.4. Follow-up Patients were followed up for 3 years after the initial assessment. Follow-up consisted of clinical examinations performed 1
and 3 months after the first stroke and then every 3 months. Clinical examinations during follow-up included history taking, physical and neurological examinations, and mRS score assessment. New major medical problems (e.g., death, recurrent cerebral infarction, cerebral hemorrhage, epilepsy, cancer, cardiovascular diseases, head injury, etc.) were recorded during follow-up. The end point was death during 3-year follow up. Every death occurring during the follow-up was reviewed. 2.5. Statistical analysis Continuous variables such as age and laboratory measurement values were expressed as median and interquartile range because the values were not normally distributed, and mean length of stay in the acute ward was expressed as mean ± SD. Categorical variables were expressed as a number, or percentage, for each item. The 2 patient groups were compared using chi-square, MannWhitney U or Student t test. The independent associations between the variables and the probability of having GI bleeding were analyzed using logistic regression. All variables with a P < 0.1 in the univariate logistic regression entered a stepwise, backward multivariate logistic regression. The Cox proportional hazards model was used to determine the significance of each variable in predicting the 3-year all-cause mortality. A univariate Cox model, assessing all previously identified variables, was used to measure hazard ratio (HR) for mortality. A backward, stepwise multivariate Cox regression model was also used to identify the risk factors for the 3-year mortality. All statistical analyses were performed with IBM SPSS statistics 19 for Windows. 3. Results 3.1. Patient characteristics A total of 934 patients were enrolled in the study. The median age was 70 (62–78) years. At baseline, 76 (8.1%) patients had GI bleeding at admission. Age was significantly higher in the group with GI bleeding. Prevalence of AF and previous TIA were significantly higher among patients with GI bleeding. Upon clinical presentation of stroke, TACS occurred more frequently in the group with GI bleeding, whereas LACS occurred more frequently in the group without GI bleeding. Additionally, hemoglobin and estimated glomerular filtration rate (eGFR) (ml/min/1.73 m2) were significantly lower, whereas WBC was higher, in the group with GI bleeding. Among patients with GI bleeding, 37 patients used aspirin, 32 patients use clopidogrel, 2 patients used warfarin, 5 patients did not use antiplatelet or anticoagulant (Table 1). 3.2. Determinants of GI bleeding in patients with acute, first-ever ischemic stroke After adjusting for the potential risk factors (P < 0.1 in univiarate logistic regression) in a backward, stepwise multivariate logistic regression, only anemia, TIA and initial impaired consciousness were positively associated with patients having GI bleeding during acute stroke (Table 2). 3.3. Clinical course in patients with an acute stage of first-ever ischemic stroke The mean length of acute ward stay, initial impaired consciousness, and stroke in evolution were significantly higher in the group with GI bleeding. The occurrence of pneumonia and urinary tract infection (UTI) were significantly higher in patients with GI bleeding. Furthermore, the dependent functional status (mRS score =3)
Please cite this article in press as: Chou Y-F et al. Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke. J Clin Neurosci (2017), http://dx.doi.org/10.1016/j.jocn.2017.06.068
3
Y.-F. Chou et al. / Journal of Clinical Neuroscience xxx (2017) xxx–xxx Table 1 Demographic and clinical characteristics of first-ever ischemic stroke patients, divided according to the presence/absence of GI bleeding. GI bleeding (n = 76)
No GI bleeding (n = 858)
Odds Ratio (95% CI)
P value
Age (years) Female
75.5 (67.5–82.8) 29 (38.2%)
70.0 (63–77) 413 (48.1%)
0.67 (0.41–1.08)
<0.001* 0.006
Risk factors Hypertension Diabetes mellitus Smoking Hyperlipidemia Coronary artery disease Atrial fibrillation Previous transient ischemic attack Congestive heart failure Peripheral vascular disease
57 (75.0%) 36 (47.4%) 27 (35.5%) 23 (30.3%) 6 (7.9%) 17 (22.4%) 9 (11.8%) 2 (2.6%) 0 (0.0%)
616 (71.8%) 323 (37.6%) 278 (32.4%) 347 (40.4%) 53 (6.2%) 119 (13.9%) 42 (4.9%) 17 (2.0%) 4 (0.5%)
1.18 1.48 1.15 0.64 1.30 1.79 2.61 1.34 0.64
(0.69–2.02) (0.93–2.38) (0.70–1.88) (0.39–1.06) (0.54–3.13) (1.01–3.17) (1.22–5.59) (0.30–5.90) (0.39–1.06)
0.326 0.065 0.330 0.051 0.347 0.038* 0.018* 0.467 0.712
Clinical syndromes TACS PACS LACS POCS
27 22 12 15
92 (10.7%) 253 (29.5%) 363 (42.3%) 149 (17.4%)
4.59 0.97 0.26 1.17
(2.74–7.70) (0.58–1.63) (0.14–0.48) (0.65–2.12)
<0.001* 0.519 <0.001* 0.349
Lab data Hemoglobin WBC eGFR
13.0 (11.4–14.8) 8600 (6700–11,400) 66.0 (49.0–75.8)
(35.5%) (28.9%) (15.8%) (19.7%)
0.010* <0.001* <0.001*
13.6 (12.5–14.7) 7100 (5800–9000) 74.1 (57.6–87.9)
CI = confidence interval; GI = gastrointestinal; TACS = total anterior circulation syndrome; PACS = partial anterior circulation syndrome; LACS = lacunar syndromes; POCS = posterior circulation syndrome; MCV = mean corpuscular volume; WBC = white blood cells; eGFR = estimated glomerular filtration rate (ml/min/1.73 m2). Data are presented as median (interquartile range) or n (%). * P < 0.05, Chi-square or Mann-Whitney U test.
Table 2 Logistic regression for predictors of GI bleeding in patients with first-ever ischemic stroke. Variables Age (year) Anemia Hypertension Diabetes mellitus Transient ischemic attack Initial impaired consciousness Hyperlipidemia Coronary artery disease Congestive heart failure Atrial fibrillation Smoking eGFR < 60
Univariate Logistic Regression, Risk Ratio (95% CI)
P value
Stepwise Multivariate Logistic Regression, Risk Ratio (95% CI)
P value
1.99 (1.56–3.44)
0.013y
2.95 (1.30–6.78) 5.32 (3.06–9.23)
0.010y <0.001*
*
1.04 2.56 1.18 1.48 2.61 6.24
(1.02–1.07) (1.57–4.16) (0.69–2.02) (0.93–2.38) (1.22–5.59) (3.78–10.30)
<0.001 <0.001* 0.551 0.101 0.014* <0.001*
0.64 1.30 1.34 1.79 1.15 1.74
(0.39–1.06) (0.54–3.13) (0.30–5.90) (1.01–3.17) (0.70–1.88) (1.07–2.82)
0.084* 0.558 0.701 0.047* 0.578 0.025*
CI = confidence interval; eGFR = estimated glomerular filtration rate (ml/min/1.73 m2); GI = gastrointestinal. P < 0.05 for the univariate logistic regression. y P < 0.05 for the stepwise multivariate logistic regression.
*
Table 3 Clinical course of acute stage, mortality, and stroke recurrence within 3 years after onset of first-ever ischemic stroke, divided according to the presence/absence of GI bleeding. GI bleeding (n = 76)
No GI bleeding (n = 858)
Odds ratio (95% CI)
P value
Mean length of stay in the acute medicine ward (days) Initial impaired consciousness
29.13 ± 22.43 33 (43.4%)
13.36 ± 10.68 94 (11.0%)
15.77 (12.94–18.60) 6.24 (3.38–10.30)
<0.001* <0.001*
Course of acute stroke stage In evolution
37 (48.7%)
196 (22.8%)
3.20 (1.99–5.16)
<0.001*
Complications Pneumonia Urinary tract infection mRS score 3 upon discharge Stroke recurrence Death
37 (48.7%) 22 (28.9%) 71 (93.4%) 7 (9.2%) 18 (23.7%)
63 (7.3%) 95 (11.1%) 530 (61.8%) 145 (16.9%) 63 (7.3%)
11.97 (7.13–20.09) 3.27 (1.91–5.61) 8.79 (3.51–21.99) 0.50 (0.23–1.11) 3.92 (2.18–7.05)
<0.001* <0.001* <0.001* 0.051 <0.001*
GI = gastrointestinal; mRS = modified Rankin scale. Data are presented as mean ± SD or n (%). * P < 0.05, chi-square or Student t tests.
Please cite this article in press as: Chou Y-F et al. Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke. J Clin Neurosci (2017), http://dx.doi.org/10.1016/j.jocn.2017.06.068
4
Y.-F. Chou et al. / Journal of Clinical Neuroscience xxx (2017) xxx–xxx
following discharge was higher in patients with GI bleeding. The rate of mortality within 3 years after stroke onset was significantly higher among patients with GI bleeding (Table 3). 3.4. Kaplan-Meier survival analysis of 3-year mortality in patients with acute, first-ever stroke At the end of the 3-year observation period, 81 patients had died (81/934 = 8.67%), including 18 (23.7%) in the group with GI bleeding and 63 (7.3%) in the group without GI bleeding. Among the patients with GI bleeding, 8 (44.4%) died within acute stroke stage, 3 (16.7%) died of infection, 3 (16.7%) died of cardiovascular disease, 1 (5.6%) died of recurrent stroke, 1 (5.6%) died of complication due to rt-PA infusion, 1 (5.6%) died of sudden apnea with an unknown cause, 1 (5.6%) died of cancer, and 1 (5.6%) died of uremia. The Kaplan-Meier survival analysis revealed that the group with GI bleeding had a higher rate of mortality than the group without GI bleeding (Fig. 1). 3.5. Cox regression multivariate analysis for 3-year mortality in patients with acute, first-ever stroke The univariate Cox regression indicated that age, reduced eGFR, CAD, AF, CHF, lower lipid value, TACS, and GI bleeding were potential risk factors (P < 0.1) for a 3-year mortality. These variables entered the backward, stepwise multivariate Cox proportional hazards model. We found that GI bleeding (HR = 2.76; 95% CI = 1.61– 4.72; P < 0.001) was a significant risk factor for a 3-year all-cause mortality in these patients even after adjusting for these variables (Table 4). 4. Discussion This study shows, for the first time, the association between GI bleeding in the acute stroke stage and increased 3-year mortality in patients with first-ever ischemic stroke. This association was significant even after adjusting for the established clinical predictors of adverse outcomes. To our knowledge, this is the first study to investigate the influence of GI bleeding in the acute stroke stage on long-term mortality.
Fig. 1. Kaplan-Meier estimates of patient survival (all-cause mortality) during the 3-year study period. Symbols are: (—) No GI bleeding; () GI bleeding. Log Rank P < 0.001.
In this study, we found that patients were significantly older and the stroke severity was significantly higher in the group with GI bleeding; furthermore, the prevalence of AF was significantly higher in these patients, which were similar to previous studies [1,3,6,13]. This may be due to the fact that AF is often associated with increased stroke severity [14,15], and a severe stroke increases the risk of stress-derived peptic ulcers. Our results show that the hemoglobin level was significantly lower in the group with GI bleeding. This finding may be due to that small amount of bleeding may frequently occur before the stroke episode in patients with GI bleeding [16]. In this study, the frequency of reduced eGFR is significantly higher in patients with GI bleeding. Previous study showed that patients with chronic kidney disease have higher risk of both upper and lower GI bleeding [17]. Our result also revealed that WBC count is significantly higher in patients with GI bleeding. A previous study has shown that leukocytosis is common in patients with upper GI bleeding, it may reflect the severity of the bleeding episode, and is associated with a complicated course [18]. The mean length of acute ward stay in our study was significantly longer in patients with GI bleeding than in patients without GI bleeding. We also showed that initial impaired consciousness was more frequent in patients with GI bleeding, confirming the findings of a previous study [1,13]. Our study also revealed that the occurrence of pneumonia was significantly higher in patients with GI bleeding. A previous study showed that respiratory complications, such as pneumonia, are common in critically ill medical patients with serious acute upper GI bleeding and are associated with a poor outcome [19]. Another study revealed that stress ulcer prophylactic agents are effective in decreasing bleeding, and the use of them is associated with a lower rate of nosocomial pneumonia and mortality [20]. The relationship between GI bleeding and the neurological course of acute ischemic stroke remains unclear. We observed a significant association between GI bleeding and stroke in evolution. There are several hypotheses regarding the mechanisms for the association between GI bleeding and poor clinical outcomes. Bleeding contributes to hemodynamic insufficiency and leads to the discontinuation of antithrombotic treatment, which may result in a prothrombotic state [21]. These factors may then lead to the deterioration of neurological symptoms and poor functional outcome. Whether treating GI bleeding will improve stroke outcome remains unclear. The use of prophylactic proton pump inhibitors or H2 antagonists may reduce the risk of upper GI bleeding in acute stroke stage and improve the tolerance of antiplatelet or anticoagulant. In addition, treating GI bleeding will prevent hemodynamic instability, which may lead to brain hypoperfusion. Furthermore, previous studies suggested that severe upper GI bleeding increase the risk of respiratory complication (e.g., pneumonia, etc), which is associated with increased mortality [19,22]. Therefore, treating GI bleeding may decrease the occurrence of respiratory complication, which could improve stroke outcome. Further studies are required to elucidate the mechanisms for the associations. This study has several limitations. First, we did not evaluate the stroke severity using the National Institutes of Health Stroke Scale score because a number of patients did not receive this scoring while in admission. Instead, we used the OSCP classification to evaluate stroke severity; LACS was regarded as a less severe and TACS as a more severe stroke. We have included TACS in the Cox regression analysis. Second, we included patients with ischemic stroke only; hence, our results do not reflect the impact of GI bleeding on other types of stroke. However, our study represents a homogenous population. Third, we could not detect all patients with GI bleeding because those who showed a gradual decrease in hemoglobin without hematemesis, or melena, were not
Please cite this article in press as: Chou Y-F et al. Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke. J Clin Neurosci (2017), http://dx.doi.org/10.1016/j.jocn.2017.06.068
5
Y.-F. Chou et al. / Journal of Clinical Neuroscience xxx (2017) xxx–xxx Table 4 Cox regression analysis of patient survival during the 3-year study period. Univariate Cox regression
Age (year-old) Gender eGFR < 60 (Yes) Recurrent stroke (Yes) Hypertension (Yes) Diabetes mellitus (Yes) Coronary artery disease (Yes) Atrial fibrillation (Yes) Congestive heart failure (Yes) Smoking (Yes) Hyperlipidemia (Yes) TACS GI bleeding
Multivariate Cox regression
Hazard Ratio (95% CI)
P value
Hazard Ratio (95% CI)
P value
1.04 0.71 1.99 0.99 1.02 1.23 2.42 2.49 2.82 0.93 0.53 2.87 3.59
<0.001* 0.135 0.002* 0.979 0.945 0.364 0.006* <0.001* 0.044* 0.764 0.011* <0.001* <0.001*
1.03 (1.00–1.05)
0.033y
2.48 (1.30–4.71) 1.94 (1.18–3.21)
0.006y 0.010y
(1.02–1.06) (0.46–1.11) (1.28–3.09) (0.49–2.00) (0.62–1.66) (0.79–1.90) (1.28–4.58) (1.54–4.03) (1.03–7.67) (0.58–1.49) (0.32–0.86) (1.76–4.69) (2.13–6.07)
<0.001y
2.76 (1.61–4.72) 2
CI = confidence interval; TACS = total anterior circulation syndrome; GI = gastrointestinal; eGFR = estimated glomerular filtration rate (ml/min/1.73 m ). P < 0.1 for the univariate Cox regression. y P < 0.05 for the multivariate Cox regression.
*
included. Forth, we did not have the information about the use of prophylaxis for GI bleeding such as H2 antagonists and proton pump inhibitors. 5. Conclusion GI bleeding is associated with an increased risk of 3-year mortality and poor functional outcome in patients with acute, firstever ischemic stroke. Prophylactic therapies for GI bleeding may reduce the risk of GI bleeding in the acute stroke stage and improve the outcomes of ischemic stroke. Conflicts of interest/disclosures
[9] [10]
[11] [12] [13]
[14]
There is no financial or other conflicts of interest in relation to this research and its publication.
[15]
References
[16]
[1] Hsu HL, Lin YH, Huang YC, et al. Gastrointestinal hemorrhage after acute ischemic stroke and its risk factors in Asians. Eur Neurol 2009;62:212–8. [2] Misra UK, Kalita J, Pandey S, et al. Predictors of gastrointestinal bleeding in acute intracerebral haemorrhage. J Neurol Sci 2003;208:25–9. [3] O’Donnell MJ, Kapral MK, Fang J, et al. Gastrointestinal bleeding after acute ischemic stroke. Neurology 2008;71:650–5. [4] Ogata T, Kamouchi M, Matsuo R, et al. Gastrointestinal bleeding in acute ischemic stroke: recent trends from the Fukuoka Stroke Registry. Cerebrovasc Dis Extra 2014;4:156–64. [5] Wijdicks EF, Fulgham JR, Batts KP. Gastrointestinal bleeding in stroke. Stroke 1994;25:2146–8. [6] Rumalla K, Mittal MK. Gastrointestinal bleeding in acute ischemic stroke: a population-based analysis of hospitalizations in the United States. J Stroke Cerebrovasc Dis 2016;25:1728–35. [7] Crooks CJ, Card TR, West J. Excess long-term mortality following non-variceal upper gastrointestinal bleeding: a population-based cohort study. PLoS Med 2013;10:e1001437. [8] Adams Jr HP, del Zoppo G, Alberts MJ, et al. Guidelines for the early management of adults with ischemic stroke: a guideline from the American
[17] [18]
[19]
[20]
[21]
[22]
Heart Association/American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups: the American Academy of Neurology affirms the value of this guideline as an educational tool for neurologists. Stroke 2007;38:1655–711. Davenport RJ, Dennis MS, Warlow CP. Gastrointestinal hemorrhage after acute stroke. Stroke 1996;27:421–4. Wlodek A, Sarzynska-Dlugosz I, Sandercock PA, et al. Agreement between the clinical Oxfordshire Community Stroke Project classification and CT findings in Poland. Eur J Neurol 2004;11:91–6. Gautier JC. Stroke-in-progression. Stroke 1985;16:729–33. Quinn TJ, Dawson J, Walters MR, et al. Reliability of the modified rankin scale: a systematic review. Stroke 2009;40:3393–5. Chen CM, Hsu HC, Chuang YW, et al. Study on factors affecting the occurrence of upper gastrointestinal bleeding in elderly acute stroke patients undergoing rehabilitation. J Nutr Health Aging 2011;15:632–6. Nolte CH, Rossnagel K, Jungehuelsing GJ, et al. Gender differences in knowledge of stroke in patients with atrial fibrillation. Prev Med 2005;41:226–31. Steger C, Pratter A, Martinek-Bregel M, et al. Stroke patients with atrial fibrillation have a worse prognosis than patients without: data from the Austrian Stroke registry. Eur Heart J 2004;25:1734–40. Kim BS, Li BT, Engel A, et al. Diagnosis of gastrointestinal bleeding: a practical guide for clinicians. World J Gastrointest Pathophysiol 2014;5:467–78. Luo JC, Leu HB, Hou MC, et al. Nonpeptic ulcer, nonvariceal gastrointestinal bleeding in hemodialysis patients. Am J Med 2013;126(264):e225–32. Chalasani N, Patel K, Clark WS, et al. The prevalence and significance of leukocytosis in upper gastrointestinal bleeding. Am J Med Sci 1998;315:233–6. Liebler JM, Benner K, Putnam T, et al. Respiratory complications in critically ill medical patients with acute upper gastrointestinal bleeding. Crit Care Med 1991;19:1152–7. Cook DJ. Stress ulcer prophylaxis: gastrointestinal bleeding and nosocomial pneumonia. Best evidence synthesis. Scand J Gastroenterol Suppl 1995;210:48–52. Eikelboom JW, Mehta SR, Anand SS, et al. Adverse impact of bleeding on prognosis in patients with acute coronary syndromes. Circulation 2006;114:774–82. Yu YJ, Weng WC, Su FC, et al. Association between pneumonia in acute stroke stage and 3-year mortality in patients with acute first-ever ischemic stroke. J Clin Neurosci 2016;33:124–8.
Please cite this article in press as: Chou Y-F et al. Association between gastrointestinal bleeding and 3-year mortality in patients with acute, first-ever ischemic stroke. J Clin Neurosci (2017), http://dx.doi.org/10.1016/j.jocn.2017.06.068