Aspirin Versus Clopidogrel Monotherapy for the Secondary Prevention of Recurrent Cerebrovascular Attack Following Previous Ischemic Stroke in Patients with Type 2 Diabetes Mellitus: A Systematic Review and Meta-Analysis

Online medical databases including Web of Science, MEDLINE, Cochrane central, EMBASE and http://​www.​ClinicalTrials.​com were searched for published articles that satisfied the inclusion and exclusion criteria of this study. Reference lists of suitable articles were also verified for relevant articles.

Specific search terms were used during the retrieval of English publications:

Studies were included if:

Criteria for exclusion were:

Data were carefully extracted by seven authors, including the number of participants with T2DM; the number of participants assigned to aspirin and clopidogrel monotherapy, respectively; the number of T2DM participants with ischemic stroke, recurrent stroke, fatal stroke, cerebral hemorrhage and myocardial infarction; the total number of deaths; the baseline characteristics of the participants; the type of study; and the relevant follow-up time periods.

Any disagreement among the authors was further discussed, and a decision was made by the corresponding author.

The quality assessment for the trials was carried out based on the recommendations by the Cochrane Collaboration [8], and the assessment of the observational studies was carried out by the criteria of the Newcastle-Ottawa Scale (NOS) [9]. Grades were allotted: grade ‘A’ implied low risk of bias, grade ‘B’ was moderate risk of bias, and grade ‘C’ was high risk of bias.

Recurrent cerebrovascular attack was defined as the occurrence of a new stroke in a patient who already had had a previous stroke.

The following outcomes were assessed during a mean follow-up time period between 1 and 5 years:

RevMan 5.3 software was used to statistically analyze the data representing each subgroup. Heterogeneity was assessed by two methods: the Q statistic test, whereby a P value ≤ 0.05 was considered statistically significant, and the I² statistic test, whereby heterogeneity increased with a rising I² value. In addition, a fixed statistical model (I² < 50%) or a random statistical model (I² > 50%) was used based on this I² value. Risk ratios (RRs) with 95% confidence intervals (CIs) were used to represent the results following analysis.

Sensitivity analysis was carried out by a method of exclusion whereby studies were eliminated one by one, and a new analysis was carried out each time to observe any deviation from the main result.

Publication bias was visually observed through funnel plots.

This meta-analysis is based on previously conducted studies and does not contain any studies with human participants or animals performed by any of the authors.

A total of 945 publications were found through the literature search following the PRISMA guideline [10]. The abstracts and titles were carefully assessed by the authors, and an initial elimination was carried out whereby 878 publications were eliminated since they were not linked or related to the scope of the current title of this publication. Sixty-seven (67) full-text articles were assessed for eligibility.

A second set of eliminations was carried out but this time following assessment of the full text articles. This elimination was due to publications that:

Finally, only six studies [11‐16] were selected for this analysis (Fig. 1) (Table 1).

A total of 9218 participants with T2DM who were previously affected by ischemic stroke were included in this analysis, whereby 4917 were assigned to aspirin and 4301 to clopidogrel, as shown in Table 2.

A methodologic assessment was carried out on the studies. Grade ‘B’ was allotted to each of the studies showing moderate risk of bias.

Table 3 lists the baseline features of the T2DM participants with previous ischemic stroke. Male participants who were on aspirin varied from 47.5% to 66.5%, whereas male patients who were on clopidogrel varied from 49.3% to 74.8%. Mean age of the participants varied between 64.3 and 75.3 years; 42.3% to 79.2% of the participants were hypertensive, 20.3% to 59.8% of the participants suffered from dyslipidemia, and 22.0% to 46.0% of the participants were current smokers.

This current analysis showed that there was no significant difference in recurrent stroke (RR: 0.79, 95% CI: 0.61–1.02; P = 0.07) observed with aspirin versus clopidogrel monotherapy for the secondary prevention of recurrent cerebrovascular attack following previous ischemic stroke in patients with T2DM as shown in Fig. 2. Our analysis also showed that the risk of fatal stroke (RR: 0.88, 95% CI: 0.39–1.98; P = 0.76), cerebral hemorrhage (RR: 0.65, 95% CI: 0.38–1.11; P = 0.12), MI (RR: 0.88, 95% CI: 0.43–1.79; P = 0.71) and mortality (RR: 1.07, 95% CI: 0.90–1.27; P = 0.44) were also similar with aspirin versus clopidogrel monotherapy in these patients with T2DM as shown in Fig. 3.

The results have been summarized in Table 4.

Consistent results were obtained throughout when sensitivity analysis was carried out.

Visual assessment of the funnel plot also showed minimal evidence of publication bias across the studies that were involved in assessing the cerebrovascular and other clinical outcomes among these patients with T2DM as shown in Fig. 4.

Several limitations were observed in this study. First, the total number of participants might not be sufficiently large to draw a strong and powerful conclusion. The total number of studies selected for this analysis was limited since there have seldom been research articles based on aspirin versus clopidogrel monotherapy in T2DM participants with previous cerebrovascular attack. Hence, subgroups assessing fatal stroke and MI involved only two studies. The other studies did not report these two outcomes and hence could not be included during subgroup analyses for these specific outcomes. Several other important outcomes and adverse drug events including gastrointestinal bleeding, abdominal pain, melena and other bleeding tendencies could not be assessed since they were not reported in the original studies. Moreover, another limitation of this analysis might be the fact that data from randomized trials and observational studies were combined and analyzed. This might have affected the result of this analysis. However, the total number of studies were limited, and separating randomized trials and observational cohorts and carrying out an analysis would not have been possible.