METHOD

This is a cross-sectional and retrospective study with prospective data collection from medical records in a computerized date registry. The study group consisted of patients with stroke in the MCA territory, including cortical, subcortical and deep lesions, who underwent IVTT from 1^st March 2010 to 31^st December 2011 at two Brazilian Stroke Centres (Hospital de Clínicas, Universidade Federal do Paraná in Curitiba, Paraná; and the Hospital Municipal São José in Joinville, Santa Catarina). Stroke management, including the standardized IVTT, was based on current guidelines^10,11,12,13. Exclusion criteria were the following: the presence of bilateral MCA stroke during the same admission, the presence of an infarction outside of the MCA territory during the same admission, intra-arterial rescue therapy, and incomplete data about the territory of infarction and the modified Rankin score (mRS) at 3 months. Although, among all ischemic stroke subtypes, lacunar strokes have been considered the most benign, this subtype was not excluded because reviewed studies show that thrombolysis is an effective treatment in those cases and the prognosis can be better or without significant trend for better or worse outcome compared with others etiologies^{10,14,15,16,17,18}. The ethics committees of both hospitals had previously evaluated and authorized the study.

Neurologists certified in the use of the NIHSS examination evaluated all of the patients at both centres using the validated Portuguese version of the NIHSS¹⁹. The NIHSS was stratified in three different severity groups based on total punctuation: mild ranging from 0 to 3; moderate: 4 - 19; and severe: 20 - 42⁹.

The main outcome measure was mortality during the three months after stroke onset. The other outcome variables evaluated were hemorrhagic transformation (HT), symptomatic HT (SHT), functional independency as defined as a mRS ≤ 2 on discharge, mRS ≤ 2 after three months and death during admission.

HT was defined as a hemorrhagic infarction (HI) in which a petechial bleeding inside the infarcted area was demonstrated on control neuroimaging, but without a space-occupying effect. A parenchymal haemorrhage (PH) was defined as a haemorrhage with a space-occupying effect^20,21. A SHT was defined as brain imaging evidence of HT with clinical worsening, which was indicated by an increase of at least four points in the NIHSS score¹⁵.

The statistical analyses were performed with the Statistica 8.0 software: statistical significance was assessed by either the Student’s t-test or the Mann-Whitney’s test for continuous variables and the χ²-test or Fisher’s exact test for categorical variables. Statistical significance was set at a p < 0.05 value with a 95% confidence interval.

RESULTS

During the study period, 112 MCA IS patients were analysed: 45 (40.2%) RMCAs and 67 (59.8%) LMCAs. Table 1 shows the demographic and admission variables of both groups. On the one hand patients with LMCA IS had atrial fibrillation more often when compared to the RMCA IS group (34.3% vs. 15.6%, p = 0.031), while on the other hand patients with RMCA IS were more prone to have had a history of previous IS as compared to the LMCA IS ones (6% vs. 20%, p = 0.034). Regardless of laterality, patients with a history of previous IS on admission had a lower punctuation in the NIHSS scale if compared to those without (7 ± 7 vs. 15 ± 5.8, p = 0.027).

In LMCA IS, the etiological distribution was: 14 (20.9%) had an atherothrombotic cause, 26 (38.8%) were cardioembolic, two (3%) had lacunar strokes and the remaining 25 (37.3%) had an undetermined mechanism. In RMCA IS, the etiological classification was: 14 (31.1%) were atherothrombotic, 17 (37.8%) cardioembolic, two (4.4%) lacunar and 12 (26.7%) had an undetermined cause (p = 0.529).

Both sides of unilateral MCA compromise had similar median NIHSS on admission (14 vs. 14, p = 0.456). Nevertheless, those patients that had a more severe compromise, the severe NIHSS subgroup, were more frequent in the LMCA IS group than in the RMCA IS (20 vs. 4, p = 0.009) (Table 1).

The outcome findings of each group are presented in Table 2. There was no statistically significant difference between the RMCA IS and LMCA IS groups for any of the outcome measures.

DISCUSSION

The present study failed to demonstrate that there was a difference in the mortality after 3 months between RMCA IS and LMCA IS. However, there is still some controversy in the literature regarding whether the side of the MCA IS can have any influence in its prognosis. Some authors suggested that patients with RMCA IS could have a longer symptom-to-needle time and a lower score in neurological scales because these scores are predominantly focused on language items, therefore favouring LMCA IS, and thus having a negative impact in the treatment and a worse outcome when compared to LMCA IS^1,2,4,5,6. In the present study, the symptom-to-needle time and the median NIHSS were similar in both groups. Additionally, the LMCA IS group had a greater number of severely compromised patients as compared to the RMCA IS group, and this could have influenced the results obtained. Previous studies have shown that the influence of the dominant hemisphere occurs predominantly in those patients within the lower total punctuation of NIHSS score subgroup, because the lateralization of hemispheric compromise has a significant influence on the cortical items, mostly on those related to assessment of language and comprehension, an effect which disappears in patients with greater compromise and higher scores, becoming similar for either side².

It was observed that in the RMCA stroke patients with HT (n = 5), most of them had SHT (n = 4), but this did not change the patients outcome and also it was not statistically significant when compared to LMCA stroke patients (HT = 14, SHT = 6). When analyzed all SHT in the current study (n = 10, 8.9%), this was similar when compared to previous large trials^10,20 and real world studies¹³.

There are some limitations in the present study. First, only those patients with MCA strokes that underwent intravenous thrombolysis were analysed; this restricted the validation of the current data with other topographies and with non-thrombolysis treated patients. The retrospective method excluded patients without complete data and this could also have influenced the observed results. In addition to the previous validation of NIHSS in the Portuguese-speaking Brazilian population¹⁹, patients included in this study could have different cultural and educational levels, predominantly in regards to the language items, and this could have affected the observed results. Language dominance with neuropsychological testing was not evaluated in the present model. Finally, this study did not access the mortality in the different severity groups based on NIHSS. Nevertheless, some studies suggest that the influence of the dominant hemisphere occurs predominantly when there is lower total punctuation of NIHSS score² and in our study there is no patient with mild NIHSS in any of the groups. However, this is the first study that includes only patients who underwent IVTT and uses a differentiated analysis of NIHSS.

In conclusion, the side of MCA IS did not influence the patient’s outcome in the present study and therefore should not be used as a prognostic marker. Future studies should emphasize other factors that could be used to establish the prognosis of MCA IS in a more reliable manner.