DOI: https://doi.org/10.21203/rs.3.rs-90058/v1
Introduction: The role of Platelet-to-lymphocyte ratio (PLR) in outcomes of acute ischemic stroke, especially before and after intravenous thrombolysis with recombinant tissue plasminogen activator (rtPA), has not been elucidated. Thus, the aim of this study was to evaluate the effect of PLR before and after rtPA on clinical outcomes.
Methods: A total of 582 consecutive patients who had acute ischemic stroke diagnosed and received intravenous thrombolysis with rtPA were included in this study. We collected demographics, vascular risk factors, previous history of drugs and other clinical information for all patients. Specifically, blood samples for PLR values were collected on admission and at 24 hours after stroke. Multivariate logistic regression analysis was used to assess the association between PLR with the risk of poor outcome (mRS≥3), death and hemorrhagic transformation (HT).
Results: Of 582 patients, 191 (32.8%) had a poor outcome, 40 (6.9%) died and 82 (14.1%) had HT. After adjustment for potential confounders, multivariate logistic regression analysis showed that higher PLR at 24h after rtPA was independently associated with an increased risk of poor outcome (OR=1.004; 95% CI:1.001-1.007; P=0.009) and the occurrence of death (OR=1.009; 95% CI:1.004-1.013; P<0.001), but not associated with the risk of HT (OR=1.003; 95% CI:0.999-1.007; P=0.165). In addition, PLR on admission was not associated with the risk of poor outcome, death and HT (all P>0.05).
Conclusions: We found that PLR at 24h after rtPA can predict the risk of poor outcome and death in acute ischemic stroke patients, but PLR on admission cannot.
Stroke is one of the most common causes of disability and death worldwide and ischemic stroke accounts for 80% of all cases[1, 2]. Intravenous thrombolysis with recombinant tissue plasminogen activator (rtPA) is the most effective therapy and has been widely used in acute ischemic stroke patients within 4.5 hours[3]. However, it also may cause some complications, such as hemorrhagic transformation (HT), that affect the functional outcomes[4, 5]. Therefore, it is important to understand what clinical indicator can predict patients’ prognosis and do what is possible to control the risk factors at an earlier stage.
Previous research suggested that inflammatory response plays an important role in the whole process of acute ischemic stroke[6]. Platelet-to-lymphocyte ratio (PLR), as a combination of platelets and lymphocytes, is a new biomarker to reflect the thrombus formation pathway and inflammation pathway. Several pieces of evidence have confirmed that PLR was a potential marker to predict increased inflammation, the risk of myocardial infarction and ischemic stroke[7–12]. However, the predictive value of PLR on the outcomes of ischemic stroke with rtPA treatment are largely unknown. Thus, the aim of this study was to evaluate the relationship between the PLR before and after thrombolysis treatment and poor outcome, death and HT in acute ischemic stroke patients.
Study Population
This retrospective study included patients who had acute ischemic stroke diagnosed and received intravenous thrombolysis with rtPA in our department from April 2015 to March 2019. The inclusion criteria were diagnosis of acute ischemic stroke and treatment with standard-dose rtPA (0.9-mg/kg) within 4.5 hours. The exclusion criteria were missing clinical data or failure of follow-up. A flow chart of the study is given in Figure 1. The study was approved by the Ethics Review Committee of the First Hospital of Jilin University, and written informed consent was obtained from the patient or an appropriate surrogate.
Data Collection
Baseline clinical data were collected for all patients, including demographic information (age, gender), vascular risk factors (smoking, alcohol consumption, hypertension, diabetes, previous stroke), previous history of drugs (antihypertensive drugs, hypoglycemic agents, antiplatelet agents), as well as other clinical information. Patients were all treated with standard-dose thrombolysis (0.9-mg/kg). Hypertension was defined as systolic blood pressure of 140-mmHg, a diastolic blood pressure of 90-mmHg, the reported use of antihypertensive medications, or a history of diagnosed hypertension[13]. Diabetes was defined as fasting serum glucose of ≥126-mg/dL (7-mmol/L), a non-fasting glucose of ≥200-mg/dL (11.1-mmol/L), use of diabetic medications, or a previously established diagnosis[13]. Blood samples for platelet and lymphocyte counts of all patients were collected on admission and at 24 hours after thrombolysis. PLR was calculated as the ratio of platelet count to lymphocyte count.
Outcome
In this study, we included three clinical outcome indicators to assess the predictive value of PLR on the outcomes of ischemic stroke with rtPA treatment.
Modified Rankin Scale (mRS): the mRS was assessed at three months with a score ranging from 0 (no symptoms) to 6 (death)[14]. We defined mRS >2 as a poor outcome and mRS ≤2 as a favorable outcome.
Death: was defined as death associated with the stroke within three months of thrombolysis.
HT: was defined as any visible hemorrhage on brain CT or MRI 22–36h after thrombolysis[15].
Statistical Analysis
All statistical analyses were performed using IBM SPSS Statistics 20. Distribution normality was tested by the Kolmogorov-Smirnov test. We found all continuous variables were nonparametrically distributed variables. So the Mann-Whitney U-test was used for the comparison of continuous variables. The chi-squared test was used for comparison of categorical variables. Median with interquartile range (IQR) and percentages were used to describe the distribution of continuous and categorical variables respectively. Multivariate analysis was performed using logistic regression analysis which adjusted for other variables. The area under the curve of receiver operating characteristics curve (ROC) was produced to determine accuracy and threshold of PLR values for poor outcome, death and HT. P<0.05 was set as the statistical significance level.
Baseline characteristics of patients
In our study, a total of 616 patients treated with rtPA were included according to the inclusion criteria. After 34 patients were excluded according to exclusion criteria, 582 patients were finally included and the baseline clinical characteristics of participants are presented in Table 1 and Figure 1. The mean age of patients was 61 (interquartile range, 53-69) and 158 (27.1%) were women. The NIHSS score was 9 (interquartile range, 5–13). The time to treat was 180min (interquartile range, 141–230min). Of 582 patients, 191 patients (32.8%) had poor outcome, 40 patients (6.9%) died and 82 patients (14.1%) had HT.
Patients with poor outcome were more likely to have a higher age, a higher prevalence of hypertension, higher systolic blood pressure, higher NIHSS scores and a higher level of PLR at 24h after rtPA (all P<0.05) than those with a favorable outcome (mRS≤2). Patients with death after stroke within three months were older, more likely to suffer from diabetes and atrial fibrillation, and had a higher proportion of antiplatelet agents, higher NIHSS scores and a higher level of PLR at 24h after rtPA (all P<0.05). Patients who had HT or non-HT were similar in most characteristics except for age, diabetes, atrial fibrillation, antiplatelet agents and NIHSS scores (all P<0.05). There was no significant difference in other listed variables between these groups (Table 1 and 2).
Association between PLR at 24h after thrombolysis and outcomes
After adjustment for potential confounders, multivariate logistic regression analysis disclosed that higher PLR at 24h after thrombolysis was independently associated with an increased risk of poor outcome (OR=1.004; 95% CI:1.001-1.007; P=0.009) and the occurrence of death (OR=1.009; 95% CI:1.004-1.013; P<0.001), but not associated with the risk of HT (OR=1.003; 95% CI:0.999-1.007; P=0.165) (Table 3).
ROC curve analysis was performed and the results showed that the ability of the areas under the curve (AUC) for the PLR at 24h after thrombolysis to predict clinical outcome was 0.617 with a sensitivity of 55.0% and a specificity of 64.2%, while the ability to predict death was 0.699 with a sensitivity of 77.5% and a specificity of 55.9%. But no significant threshold of PLR at 24h after thrombolysis was found to predict HT (Table 4).
Association between PLR on admission and outcomes
After adjustment for potential confounders, multivariate logistic regression analysis disclosed that PLR on admission was not associated with the risk of poor outcome (OR=1.001; 95% CI:0.998-1.003; P=0.617), death (OR=0.999; 95% CI:0.993-1.004; P=0.658) and HT (OR=1.000; 95% CI:0.999-1.001; P=0.740) (Table 3). In addition, no significant threshold of PLR on admission was found to predict these outcomes (Table 4).
This study attempted to evaluate the effect of PLR before and after thrombolysis treatment on the outcomes in acute ischemic stroke patients. We found that PLR at 24 h after rtPA treatment, but not on admission, was an independent risk factor for poor outcome and death. In addition, we found PLR had no significant interaction with the risk of HT, no matter whether the index was calculated before or after thrombolysis.
Previous studies have shown that PLR has a strong relationship with ischemic events in general. For example, in a recent study, researchers suggested that PLR can be used in clinical practices for prediction of no reflow in patients with acute ST-segment elevation myocardial infarction after primary percutaneous intervention[8]. Simultaneously, the study by Tekesin A et al. [9] showed a significant association between PLR and cerebral vein thrombosis in suspected patients. Another study by Altintas O et al.[10] found that, in patients with acute ischemic stroke who underwent endovascular therapy, low-PLR values were shown to be correlated with better clinical outcome (mRS ≤ 2). In addition, in a retrospective study of 56 patients who had acute ischemic stroke diagnosed and underwent mechanical thrombectomy, researchers found that PLR had no relationship with clinical outcome, but PLR was lower in the group with a dramatic improvement in the 24th hour[11]. Compared with previous studies, the sample size of our study is relatively large and we further found that the patients with high-PLR values at 24 h after thrombolysis had poorer outcome (mRS > 2) and a higher occurrence of death compared with the patients with low-PLR values at the same time.
Platelets play an important role in prothrombotic status. A high platelet count may represent a higher tendency to thrombosis, leading to poor prognosis[16]. Moreover, platelets can release a variety of inflammatory factors and recruit leukocytes to the sites of inflammation and injury, which leads to the aggravation of inflammatory reaction in thrombus sites[17]. In a previous study, it was suggested that the increase of platelet count in patients with thrombus may be due to multiple inflammatory mediators that stimulate megakaryocyte proliferation and thus produce more platelets[18]. In contrast, lymphocytes are known to control inflammatory pathways[19, 20]. A decreased lymphocyte count may aggravate the injury of cerebral infarction and neurological deficits[21]. On the one hand, regulatory T cells, specific subtypes of lymphocytes, play a key role in eliminating inflammatory response, and act as the main brain protective immunomodulator in the process of acute stroke[22, 23]. On the other hand, a decrease in lymphocyte count may be an index to reflect the stress level of the body, which can reflect the production of cortisol-induced stress response and the activation of the renin–angiotensin system which further increases the production of pro-inflammatory cytokines promoting ischemic injury[21, 24, 25]. In general, lymphocytes have an important role in the healing and repairing inflammation[26].
PLR, as a combination of platelets and lymphocytes, has the advantage of reflecting not only the thrombus formation pathway, but also the inflammation pathway[27]. There are two main reasons why PLR may be superior to absolute platelet or lymphocyte count in predicting prognosis. First, PLR represents two inversely related predictors and immunologic pathways[20]. Second, previous studies have suggested that PLR is more stable compared with single platelet and lymphocyte counts, because platelet or lymphocyte counts could be altered by many physiologic and pathologic conditions[28–30]. Therefore, the PLR is more useful, rational and reliable for predicting the risk of poor outcome and death in patients. However, we did not find a significant association between PLR and HT in acute ischemic stroke patients after thrombolysis in this study. We think the association between PLR and HT needs further investigation in the future.
Also, the AUC for the PLR after 24 h appeared to be greater than that at baseline, and the mechanisms have not been well established. We suspect that there may be two reasons to explain this phenomenon. One reason is that, within 24 hours of stroke, the lymphocytes accumulate in cerebral vessels at a later time [31]. So we speculated that the PLR values after 24 h was more valuable than the PLR at baseline for reflecting the patients’ conditions. The other reason is that the recent study suggested that the lymphocyte count after ischemic stroke exhibited a significant temporal variation which was characterized by “dynamic” changes[22, 32]. Moreover, thrombolysis also has a significant impact on the PLR. Thus, the PLR values on admission may be unable to dynamically and comprehensively reflect the patients’ conditions.
Our study has some limitations. First, the study was retrospective and the data included patients only treated at one hospital, which might lead to selection bias. Second, we only examined PLR at two time points, which was before thrombolysis and 24 h after thrombolysis. The dynamic change of PLR during the whole process of ischemic injury and its definite mechanisms have not been fully elucidated. Further investigations are needed to explore the relevant mechanisms of this study. Lastly, some other inflammatory markers such as neutrophils, CRP, and IL-2 were not taken into consideration.
In conclusion, we found that the PLR at 24 h after rtPA thrombolysis, rather than before thrombolysis, was associated with the risk of poor outcome and death in acute ischemic stroke patients, which can be used as a simple, novel and inexpensive method for predicting patients’ prognosis.
Acknowledgments:
The authors thank the patients and their families and appreciate the study participants for their assistance in this study.
Funding:
This work was supported by the National Natural Science Foundation of China (Grant No. 81971105) to Zhen-Ni Guo, the Program for JLUSTIRT (2017TD-12), and the Jilin Provincial Key Laboratory (20190901005JC) to Yi Yang.
Competing interest:
There are no competing interests for all the other authors of this manuscript.
Consent for publication:
For each subject, an informed consent form was obtained. The document included consent for the use of individual data for academic purposes and publications.
Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Authors' contributions:
Yi Yang and Zhen-Ni Guo were responsible for study design. Material preparation, data collection and analysis were performed by Ying-Ying Sun, Mei-Qi Wang and Yan Wang. The first draft of the manuscript was written by Ying-Ying Sun and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Ethics approval and consent to participate:
This study was approved by the Ethics Review Committee of the First Hospital of Jilin University. Written informed consent was obtained from all study participants. The privacy rights of all study participants always be observed. All patients consent to participate.
Table 1 Baseline characteristics of patients according to presence/absence of poor outcome and death
|
Total |
Favorable outcome |
Poor outcome |
P |
Alive |
Dead |
P |
N |
582 |
391 |
191 |
|
542 |
40 |
|
Age, years, median (IQR) |
61(53-69) |
60(52-68) |
63(54-71) |
0.010 |
61(53-69) |
68(58-75) |
0.003 |
Females, % |
158(27.1) |
108(27.6) |
50(26.2) |
0.713 |
147(27.1) |
11(27.5) |
0.959 |
Smoking, % |
327(56.2) |
226(57.8) |
101(52.9) |
0.261 |
302(55.7) |
25(62.5) |
0.404 |
Alcohol consumption, % |
254(43.6) |
171(43.7) |
83(43.5) |
0..949 |
235(43.4) |
19(47.5) |
0.610 |
Hypertension, % |
302(51.9) |
190(48.6) |
112(58.6) |
0.023 |
283(52.2) |
19(47.5) |
0.565 |
Diabetes, % |
189(32.5) |
117(29.9) |
72(37.7) |
0.06 |
170(31.4) |
19(47.5) |
0.035 |
Coronary artery disease, % |
116(19.9) |
75(19.2) |
41(21.5) |
0.517 |
110(20.3) |
6(15.0) |
0.419 |
Atrial fibrillation, % |
27(4.6) |
20(5.1) |
7(3.7) |
0.435 |
22(4.1) |
5(12.5) |
0.014 |
Previous stroke, % |
88(15.1) |
52(13.3) |
36(18.8) |
0.079 |
82(15.1) |
6(15.0) |
0.982 |
Antihypertensive drugs, % |
200(34.4) |
124(31.7) |
76(39.8) |
0.054 |
188(34.7) |
12(30.0) |
0.547 |
Hypoglycemic agents, % |
102(17.5) |
65(16.6) |
37(19.4) |
0.413 |
93(17.2) |
9(22.5) |
0.391 |
Antiplatelet agents, % |
73(12.5) |
45(11.5) |
28(14.7) |
0.281 |
64(11.8) |
9(22.5) |
0.049 |
SBP, mmHg, median (IQR) |
154(138-165) |
152(138-164) |
157(142-169) |
0.012 |
154(138-165) |
155(135-169) |
0.881 |
DBP, mmHg, median (IQR) |
89(80-98) |
89(80-98) |
89(81-98) |
0.596 |
89(80-98) |
89(80-99) |
0.758 |
Blood glucose, mmol/L, median (IQR) |
6.92(6.13-8.65) |
6.8(6.0-8.6) |
7.1(6.3-8.7) |
0.156 |
6.9(6.1-8.6) |
7.0(6.3-9.0) |
0.636 |
Time to treat, min, median (IQR) |
180(141-230) |
182(144-231) |
175(135-230) |
0.165 |
180(140-230) |
180(158-236) |
0.405 |
Baseline NIHSS score, median (IQR) |
9(5-13) |
8(4-11) |
12(8-15) |
<0.001 |
9(5-12) |
13(10-16) |
<0.001 |
PLR on admission |
120.7(91.6-155.7) |
122.7(92.0-155.5) |
118(90.4-156.2) |
0.874 |
120.7(91.6-155.7) |
127.2(90.3-154.3) |
0.803 |
PLR at 24h after rtPA |
116.9(92.0-154.6) |
110.7(88.4-147.9) |
134(103.3-172.8) |
<0.001 |
115.3(91.0-152.0) |
148.8(121.3-208.7) |
<0.001 |
IQR, interquartile range; SBP, systolic blood pressure; DBP, diastolic blood pressure; PLR, platelet-to-lymphocyte ratio; mRS, modified Rankin Scale scores.
Table 2 Baseline characteristics of patients according to presence/absence of HT
|
Total |
No HT |
HT |
P |
N |
582 |
500 |
82 |
|
Age, years, median (IQR) |
61(53-69) |
62(54-70) |
57(50-66) |
0.006 |
Females, % |
158(27.1) |
141(28.2) |
17(20.7) |
0.159 |
Smoking, % |
327(56.2) |
278(55.6) |
49(59.8) |
0.482 |
Alcohol consumption, % |
254(43.6) |
213(42.6) |
41(50.0) |
0.210 |
Hypertension, % |
302(51.9) |
260(52.0) |
42(51.2) |
0.896 |
Diabetes, % |
189(32.5) |
153(30.6) |
36(43.9) |
0.017 |
Coronary artery disease, % |
116(19.9) |
103(20.6) |
13(15.6) |
0.319 |
Atrial fibrillation, % |
27(4.6) |
16(3.2) |
11(13.4) |
<0.001 |
Previous stroke, % |
88(15.1) |
75(15.0) |
13(15.9) |
0.841 |
Antihypertensive drugs, % |
200(34.4) |
172(34.4) |
28(34.1) |
0.964 |
Hypoglycemic agents, % |
102(17.5) |
83(16.6) |
19(23.2) |
0.147 |
Antiplatelet agents, % |
73(12.5) |
56(11.2) |
17(20.7) |
0.016 |
SBP, mmHg, median (IQR) |
154(138-165) |
154(138-165) |
154(140-165) |
0.895 |
DBP, mmHg, median (IQR) |
89(80-98) |
89(80-98) |
89(82-98) |
0.391 |
Blood glucose, mmol/L, median (IQR) |
6.92(6.13-8.65) |
6.92(6.09-8.62) |
6.91(6.18-9.56) |
0.444 |
Time to treat, min, median (IQR) |
180(141-230) |
179(138-227) |
203(154-236) |
0.115 |
Baseline NIHSS score, median (IQR) |
9(5-13) |
8(5-12) |
12(8-15) |
<0.001 |
PLR on admission |
120.7(91.6-155.7) |
118.8(90.8-154.6) |
130.8(96.1-163.7) |
0.153 |
PLR at 24h after rtPA |
116.9(92.0-154.6) |
115.8(92.0-153.4) |
127.2(94.5-183.4) |
0.131 |
IQR, interquartile range; SBP, systolic blood pressure; DBP, diastolic blood pressure; PLR, platelet-to-lymphocyte ratio; mRS, modified Rankin Scale scores; HT, hemorrhagic transformation.
Table 3 Multivariable logistic regression analysis of PLR associated with the poor outcome, death and HT
|
OR |
95% confidence interval |
Pa |
|
Lower bound |
Upper bound |
|||
On admission |
||||
Poor outcome |
1.001 |
0.998 |
1.003 |
0.617 |
Death |
0.999 |
0.993 |
1.004 |
0.658 |
HT |
1.000 |
0.999 |
1.001 |
0.740 |
At 24h after rtPA |
||||
Poor outcome |
1.004 |
1.001 |
1.007 |
0.009 |
Death |
1.009 |
1.004 |
1.013 |
<0.001 |
HT |
1.003 |
0.999 |
1.007 |
0.165 |
a Adjusting for age, sex, smoking, alcohol drinking, hypertension, diabetes, coronary artery disease, atrial fibrillation, previous stroke, antihypertensive drugs, hypoglycemic agents, antiplatelet agents, systolic blood pressure, diastolic blood pressure, blood glucose, time to treat and NIHSS score at baseline. HT, hemorrhagic transformation; rtPA, recombinant tissue plasminogen activator.
Table 4 Accuracy of PLR in prediction of outcomes
|
Threshold |
AUC (95% CI) |
Sensitivity, % |
Specificity, % |
P |
On admission |
|||||
Poor outcome |
234.883 |
0.496(0.446-0.546) |
9.9 |
93.9 |
0.874 |
Death |
133.895 |
0.512(0.417-0.607) |
47.5 |
63.5 |
0.802 |
HT |
130.526 |
0.549(0.481-0.617) |
51.2 |
61.6 |
0.153 |
At 24h after rtPA |
|||||
Poor outcome |
126.995 |
0.617(0.568-0.666) |
55.5 |
64.2 |
<0.001 |
Death |
120.920 |
0.699(0.620-0.779) |
77.5 |
55.9 |
<0.001 |
HT |
199.194 |
0.552(0.482-0.622) |
22.0 |
89.6 |
0.131 |
HT, hemorrhagic transformation; rtPA, recombinant tissue plasminogen activator.