The distribution indices of erythrocytes: which one for acute ischemic stroke?

Introduction

Stroke constitutes the vast majority of cerebrovascular deaths and is ranked fifth among death causes when considered separately from cancer, chronic lower respiratory tract diseases and other cardiovascular diseases such as involuntary injuries/accidents [1]. Ischemic stroke constitutes 87% of all stroke events and it leads to serious neurological disabilitys [2]. Hereby, understanding of the pathophysiological mechanisms of stroke and providing management in risk factors is important.

Increased red blood cell distribution width (RDW) predicted mortality and outcome after acute ischemic stroke (AIS) however, the underlying mechanisms are unclear. Some inflammatory, hematologic, and coagulation related laboratory parameters have been related with poor outcome measures of stroke including lower hemoglobin concentration [3], increased hematocrit [4], high mean platelet volume [5], high C reactive protein [6], high white blood cell (WBC) and neutrophil count [7]. These among useful parameters of the whole blood count, RDW has recently investigated as an independent and strong predictive factor in vary pathologies, including some infections, diabetes, cancer, and acute coroner syndrome [8].

RDW’s high levels make a sign the existence of anisocytosis which is concerned to impairment of erythropoiesis and reduction of erythrocytes [9], finding chronic inflammation and increase of oxidative stress levels, and this may significantly aggravates to development of atherosclerotic diseases [10], [11].

Currently, hematology analyzers give two distinct results with two different calculation methods as the standard deviation of red blood cell distribution width (RDW-SD) (direct measurement) and variation coefficient of red blood cell distribution width (RDW-CV) [12]. The RDW-CV is a calculation based on both the mean cell size and the width of the distribution curve. It is computed by dividing the standard deviation of the mean cell size by the MCV of the red blood cells and multiplying by 100 to turn to a percentage (Figure 1) [13].

Figure 1:

Equation 1.

Histogram based RDW-CV calculation method.

Because it is a calculation, the RDW-CV is dependent not only on the mean corpuscular volume (MCV) of the red cell population but also the width of the distribution curve and may not projected the actual variation in red cell size.

The RDW-SD is real measurement of the red cell distribution curve’s width of in femtoliters. The width of the distribution curve is measured at the point that is 20% level the baseline. Since the RDW-SD is an original measurement, it is not influenced by the MCV and more accurately express the red cell size variance [14].

The main purpose of interest of this study was to investigate whether alteration of the parameters of RDW-CV and RDW-SD of patients who has had an acute ischemic stroke. The second objective is to examine the effect of these RDW parameters on the 3-months survival separately and compare them with each other.

Materials and methods

Results

Initially, a total of 461 patients with acute cerebrovascular events were included to study. After the exclusion criteria, 342 cases were selected in the final analysis as patient group. A control group (n=99) was consisted of age and sex compatible (Figure 2). Finally, 441 patients were included in our analysis.

Figure 2:

Flowchart of patient exclusion criteria.

It was observed that the each RDW parameters (RDW-SD and RDV-CV) in the AIS patients were significantly higher than of mean and median values in non-AIS group (both p<0.001) (Figure 3).

Figure 3:

The status of RDW-CV and RDW-SD parameters according to groups.

The RDW-CV values at admission displayed a diagnostic accuracy for predicting patients with unfavorable outcome after 3 months, demonstrating an AUC of 0.725 (95% CI 0.675–0.775). The RDW-CV diagnostic cut-off was 13.95% with 48% sensitivity and 92% specificity. On the other hand, for RDV-SD values diagnostic threshold, with 52% sensitivity and 80% specificity was 43.55 (AUC 692, 95% CI 0.640–0.743) (Figure 4).

Figure 4:

AUC comparison between RDW-CV and RDW-SD for ischemic stroke.

Table 1 summarizes the main demographic, laboratory, clinical data of patients, presented for the overall population. There were no significant differences in sex, age, random blood glucose, LDL, triglycerides, MCV, MCH, MCHC, neutrophil levels. Also, there were no differences between the comorbidity types of the groups and those of CIRS-G scores. The hemoglobin levels were significantly higher in males than females in both groups (p<0.001) whereas AIS group’s hemoglobin levels were significantly lower than non-AIS group (p<0.001). MCH and MCHC levels were not different in patient and control group. White blood cell counts and lymphocyte percents were also significantly higher in AIS patients (p<0.001). High density lipoprotein (HDL-C) levels were significantly lower in AIS patients (p<0.01), while total cholesterol levels were significantly higher (p<0.05).

The correlations of RDW parameters were showed in Table 2. A statistically positive correlation was found between the RDW-SD and blood pressure in the ischemic stroke patients (p<0.05). The stroke severity score (NIHSS) was showed a significant linear relationship with RDW-CV (r=0.525; p<0.001) and RDW-SD (r=0.798; p<0.001) (Figure 5).

Figure 5:

Acute ischemic stroke survival by RDW- SD and RDW-CV quartiles.

The 90-days mortality of patients was 18.7% (n=64/342) and increased from 3.3% to 29.7% from the first quartile (<12.6) to the last RDW-CV quartile (>14.7). For RDW-SD, 90-days mortality rate was 2.7% in first quartiles and was 34.4% last quartiles (Figure 6 and Tables 3 and 4).

Figure 6:

Linear correlation graphs of RDW-SD and RDW-CV with NIHSS.

RDW-SD values of <42 fL (first quartiles) was significantly different from all quartiles in pairwise Mantel-Cox comparisons. Pairwise comparisons of RDW-SD and RDW-CV quartiles were showed in Table 5.

RDW-CV levels were significantly higher in females than males in the AIS group (p<0.001), while there was no difference between males and females in the AIS group in terms of RDW-SD levels. In the non-AIS group, both RDW parameters did not show any differences among the gender. There was no statistical difference between male and female gender in terms of acute ischemic stroke for RDW-SD and RDW-CV. While RDW-CV is relatively more effective on male gender survival, RDW-SD is more influential on female survival times, but there is no gender effect on total survival for RDW-SD [p=0.572; HR=0.867 (0.528–1.422)] and RDW-CV [p=0.383; HR=0.795 (0.475–1.331)]. The mean RDW-CV is quartiles and age [p=0.049; HR=1.024 (0.528–1.422) per 1 year], with a mean RDW-SD quartiles and age [p=0.038; HR=1.051 (0.522–1.201) per 1 year], and Cox is identified as a potential risk factors for AIS in the proportional hazard model, respectively.

RDW-CV values higher than 14.7% are associated with 11 fold increase in the risk of stroke [p<0.001; HR=11.06 (2.631–46.479)]; and the level of RDW-CV does show a linear relationship with the severity score of stroke. So, RDW-CV values not only predicted the occurrence of ischemic stroke but also with stroke severity. Besides, RDW-SD, the other erythrocyte index, was found to be associated with a 16-fold increase in stroke severity score when >50 fL [p<0.001; HR=16.11 (4.844–53.559)].

Discussion

The RDW expression in the literature has been used to indicate RDV-CV, and the vast majority previous stroke-related work has been done with it. In order to clarify the relationship of the erythrocyte anisocytosis in stroke mechanisms, it may be useful to examine the RDW-CV and RDW-SD separately.

According to the findings of our study, in stroke predictability, RDW-SD was relatively more sensitive and RDW-CV was relatively more specific. The effect of RDW-CV and RDW-SD levels on survival time percentages and hazard ratios due to each other were found to be differently. Also, we found that the increase in the probability of death by RDW-SD is relatively more and the effect of RDW-SD on stroke severity scores (NIHSS) is stronger. In other words, we suggested that the increase in death probability with elevated RDW-SD was greater than with elevated RDV-CV, and that the effect of elevated RDW-SD on stroke severity was stronger than high levels of RDV-CV.

To our best knowledge, our research based on the scope of comparison of these RDW indices has made first study in the literature for AIS patients, because it did not encounter any other similar study.

RDW is commonly measured, inexpensive and widely available and may reflect overall inflammation in AIS. RDW is a robust predictor of the risk of all cause patient mortality in first ever AIS patients. Although the mechanism of RDW-mortality relationship is not clear, the association of RDW with AIS development and subsequent mortality may be related to chronic inflammation and inflammation contributing to the pathophysiology of the disease [21], [22]. RDW is defined as a chronic disease and co-morbidity marker [23] and associated with oxidative stress have been reported [24]. It has been suggested that alterations in erythrocyte morphology are due to oxidative stress and chronic inflammation [25].

It has been suggested that RDW levels are affected by aging, lifestyle, and disease, and associated with cardiovascular health, nutrition, genetic factors, immune state, and oxygen delivery [26].

In healthy individuals, RDW-CV and RDW-SD are parameters that are acting in the same direction and generally associated with compatible each other. These are reported from the same histogram graph, however have different reference intervals (The reference range for RDW-SD is 39–46 fL, and RDW-CV is 11.6–14.6%, in adults) and are calculated with different formula. These parameters vary in parallel in healthy individuals, but due to this calculation difference, the same RDW-CV values in different individuals and in different diseases do not correspond to the same RDW-SD, even if they are of the same age and gender. When considered together with the MCV parameter used in formulation, we can say that the levels of RDW-CV changed depending on the patient’s previously anemia status and age. In the interpretation of chronic disease due to anemia or anemia due to chronic illness, it may need to evaluate RDD-CV in separate categories with RDW-SD. Indeed, hemoglobin and MCV levels were significantly different between our patients and control groups.

Inflammation markers have been associated with the stroke early phase in the activation of immune responses [27]. Platelets have great importance as also an inflammation marker in the pathogenesis of stroke. Anti-platelet agents generally used in practice for ischemic stroke treatment and prevention [28]. In our study, RDW-SD, but not RDW-CV, showed more strong correlation with platelet indices known to be inflammatory markers [29], [30]. This result also suggests the presence of only an anemia unrelated inflammatory response in stroke development. The total comorbidity scores of our patients were not different from the control group, despite the use of a scoring not involving anemia. However, in the control group, no association of the named platelet indices with both RDW parameters was observed. In CIRS-G have was no scoring of the anemia parameters, and perhaps because of this we could not find any difference in terms of total comorbidity between the groups. We therefore recommend that different disease scorings, including the RDW-CV and RDW-SD parameters, be established.

While RDW-CV is more sensitive and effective in detecting anisocytosis at microcytic MCV intervals, it has been shown that the RDW-SD MCV range is more sensitive and effective when it is normocytic [31]. A red cells homogenous population with a narrow distribution curve and low MCV may have an elevated RDW-CV while a red cells heterogeneous population with a broad distribution curve and a high MCV may have a normal RDW-CV [32]. As a matter of fact, the MCV values in the AIS patient group were no different from the non-AIS group. This may support to clarify the relatively stronger effect of RDW-SD on survival and stroke severity.

Hemoglobin levels were significantly lower in AIS patients than in non-AIS patients, both in women and men. AIS patients’ MCV levels were correlated with RDW-CV. We could interpret this result as a consequence of the difference of MCV levels in the stroke and non-stroke groups. Increased RDW values were found in recurrent stroke patients, when compared to persons with first ever stroke and found to be predictive of intensive care unit mortality [33] and of poor outcome in ischemic stroke [34]. The particularly exclusion of recurrent stroke cases known to be associated with higher RDW levels, in this study, may be a determinant of the relatively different hazard ratios, cut-off values and sensitivity values found for RDW-CV.

Infra-tentorial infarction known to have high mortality [15] or known to cause RDW elevation; intra-cerebral hemorrhage [35], serious infection [36], malignancy [37], thalassemia [32] patients were excluded from the study. In fact, RDW levels are elevated when thalassemia is heterozygous. However, RDW levels are in normal ranges in other thalassemia types. But we removed all patients diagnosed with thalassemia before the study because (even if the number of patients with thalassemia is statistically insignificant) there were also patients with whom we could not reach the knowledge of thalassemia typology in the anamnesis patient exclusion process. These exclusions stand out as a solid aspect of our work.

There were main limitations of our study. Firstly; this study was carried out in a single hospital setting including Turkish patients, and results may not be generalized to other cohorts. Secondly; given the intrinsic limitations of the all retrospective studies, other information is the lack of data about other possible determinants of RDW; erythropoietin, transferrin saturation, ferritin, reticulocyte, vitamin B12, folic acid level, iron and iron binding capacity. Data about patient receiving any blood transfusion were not obtained. Also, inflammatory indices related red blood cell life span and iron metabolism; such as erythrocyte sedimentation rate and C-reactive protein could not be achieved.

Increased baseline RDW-CV and higher baseline RDW-SD are independently predicting subsequent risk of poor outcome in with AIS. For these parameters, there should be individual disease risk and specific prognostic definitions and also should be interpreted separately for ischemic stroke disease. It must be pointed out that the RDW-CV and RDW-SD levels may have different cut-off values and different prognostic poor outcomes for interpreting in ischemic stroke events approach. These are known as complementary parameters, but both should be used alone and comparatively with the AIS predictive and prognostic approach. In order to shed light on the underlying mechanism of the differing relationships of these two parameters, it is necessary to further study with different instruments with multi-center and different measurement principles with larger sample numbers.

Compliance with ethical standards

This protocol was approved by the Ethics Committee of Erciyes University.

As this was a retrospective no-risk research, formal patient informed consent was not required.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

All procedures performed in research involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

No potential conflict of interest was declared.