Literature retrieval and study characteristics
The study process as shown in Fig. 1. There were 4,930 items in total (1,830 from PubMed, 1,428 from WOS, 501 from Embase, and 1,171 from Cochrane Library). 2,863 items were included in the abstract screening after eliminating duplicates. Then 2,774 pointless studies were excluded. A total of 88 full-text articles were assessed for eligibility. We excluded 33 studies, 22 of which used therapies other than MT or IVT, seven studies were reviews, and four pieces involved RCTs protocol. Finally, 55 studies met our protocol and were qualitatively synthesized and meta-analyzed.
The characteristics of eligible studies were displayed in Table 1. The study evaluated data from 17 nations, including 10 from Europe, four from Asia, two from The North American, and the one from Australia. Nine RCTs and 46 observational studies—29 retrospective (RS), 16 prospective (PS), and one cross-sectional (CS) were included in the analysis. almost all studies used an mRS score ≤ 2 to define FI. Methods to define SR included TICI 2b/3, mTICI 2b/3, and eTICI 2b/3. Additionally, several methods were adopted to assess sICH (ECASS II/III, HBC, and SITS-MOST). A portion of included studies adopted multivariate analysis, multivariate binary logistic regression, and propensity score method (PSM) to adjust the data.
Quality assessment for included studies
According to RoB, most trials were of high quality and possessed a low overall risk of bias. Supplemental Fig. 11 showed the specific details. Due to randomization and blinding items, a trial had a high risk of bias[20]. Additionally, Supplemental Table 4 showed how detailed information from OS were evaluated using the NOS scale. Except for one study[21], which scored only 6 because controls for comparability between the two groups were omitted from the study. All other studies were rated as “high quality”.
Crude analysis
Primary outcomes
The results would be reported separately by RCTs and observational studies. Regarding efficacy outcomes, data from the nine RCTs indicated that MT + IVT group had superior FI than the dMT group (OR:1.27, 95%CI:1.11–1.46, Fig. 2a), with large heterogeneity (I2 = 53%, P = 0.03). About safety outcomes, the results of sICH showed no significant difference between the two groups (OR:1.13, 95%CI:0.86–1.49, Fig. 2b), indicating no heterogeneity (I2 = 0, P = 0.82). Overall, 40 observational studies reported the results for FI, suggesting better results were seen in the MT + IVT group (OR:1.34, 95%CI:1.16–1.33, Fig. 2c), with large heterogeneity (I2 = 70%, P<0.01). Data on sICH was extracted from 36 observational studies and found a 16% higher risk of HT (OR:1.16, 95%CI: 1.11–1.21, Fig. 2d) in the MT + IVT group, with no heterogeneity (I2 = 0, P = 0.80).
Secondary outcomes
On the secondary efficacy outcomes, in nine RCTs, the MT + IVT group outperformed the dMT group for excellent outcomes (mRS score: 0–1) (OR:1.21, 95%CI:1.03–1.43, Fig. 3a) with moderate heterogeneity (I2 = 43%, P = 0.09). Additionally, the MT + IVT group saw 23% more SR than the dMT group (OR:1.23, 95%CI:1.05–1.45, Fig. 3b) in eight RCTs, no heterogeneity accompanied (I2 = 0, P = 0.96). Regarding safety outcomes of aICH from seven RCTs, the MT + IVT group had a 25% higher risk of HT than the dMT group (OR:1.25, 95%CI:1.00–57, Fig. 3c), with low heterogeneity (I2 = 22%, P = 0.26). Mortality at 3-months or hospital discharge from eight RCTs in the MT + IVT group showed a lower mortality compared to the dMT group (OR:0.72, 95%CI:0.54–0.97, Fig. 3d), with large heterogeneity (I2 = 54%, P = 0.03).
For efficacy outcomes, a total of 16 OS reported the excellent outcomes (mRS score: 0–1). Better results were seen in the MT + IVT group (OR:1.30, 95%CI:1.09–1.54, Fig. 4a) with large heterogeneity (I2 = 61%, P<0.01). 38 OS showed SR outcomes, with the MT + IVT group increased the rate of SR (OR:1.23, 95%CI:1.05–1.44, Supplemental Fig. 4b), with large heterogeneity (I2 = 60%, P<0.01). For safety outcomes, the MT + IVT group had higher aICH by 19% than the dMT group (OR:1.24, 95%CI:1.05–1.46, Supplemental Fig. 4c) in 23 observational studies with moderate heterogeneity (I2 = 44%, P = 0.01). Additionally, in 34 investigations, mortality was 30% lower in the MT + IVT group compared to the dMT group (OR:0.70, 95%CI:0.64–0.77, Supplemental Fig. 4d), with moderate heterogeneity (I2 = 42%, P = 0.01).
Subgroup analysis
Given the large heterogeneity of some outcomes, subgroup analysis by study design (RS vs PS) and area (Asia vs Europe vs America) was conducted. Regarding subgroup outcomes by study region in the RCTs, there was significant difference between Europe and Asia group in terms of FI (P = 0.05), Specifically, the MT + IVT group had better outcomes in Europe (OR:1.46, 95%CI:1.07–1.98), whereas there were no significant differences in Asia subgroup between the MT + IVT and the dMT therapy (OR:0.95, 95%CI:0.75–1.21). Moreover, stratifying studies according to mortality showed significant differences (P<0.01,). In Europe, the MT + IVT group reduced mortality risk by 45% (OR:0.55, 95%CI:0.45–0.68), while in Asia there was no significant difference (OR:1.07, 95%CI: 0.78–1.48). There were no significant differences regarding SR (P = 0.73), excellent outcomes (P = 0.14), sICH (P = 0.25), and aICH (P = 0.10). The above details were depicted in Supplemental Table 1. On the basis of the results of study area subgroup in OS, no statistically significant variations regarding FI (P = 0.28), excellent outcomes (P = 0.31), SR (P = 0.93), sICH (P = 0.63), aICH (P = 0.19), and mortality (P = 0.38), of which were detailed in Supplemental Table 2.
The results of the subgroup analysis for observational studies were described in more detail below. As shown in Supplemental Table 3, there was no difference in the outcomes of FI (P = 0.13), excellent outcomes (P = 0.14), SR (P = 0.37), sICH (P = 0.20), aICH (P = 0.70), and mortality (P = 0.92).
Adjusted analysis
Primary outcomes
Results by assessing the adjusted ORs among RCTs between the MT + IVT group and the dMT group were non-significant for both FI (OR:1.17, 95%CI: 0.99–1.38, Fig. 3a) and sICH (OR:1.07, 95%CI:0.79–1.46, Fig. 3b), suggested no heterogeneity (I2 = 0, P = 0.54), and (I2 = 0, P = 0.40), respectively. However, significant better outcomes were seen in the MT + IVT group on FI in observational studies (OR:1.36, 95%CI: 1.21–1.52, Fig. 3c), with moderate heterogeneity (I2 = 48%, P = 0.02). We did not see the significant differences on sICH (OR:0.92, 95%CI:0.76–1.12, Fig. 3d) between groups with low heterogeneity (I2 = 13%, P = 0.32).
Secondary outcomes
Results from RCTs indicated that the MT + IVT group significantly decreased the risk of mortality by 35% (OR:0.65, 95%CI:0.49–0.88, Supplemental Fig. 3d), with large heterogeneity (I2 = 52%, P = 0.07). All other results were non-significant differences between the two groups regarding excellent outcomes (OR:1.11, 95%CI:0.90–1.38, Supplemental Fig. 3a), SR (OR:0.92, 95%CI:0.75–1.13, Fig. 3b), and aICH (OR:0.93,95%CI: 0.75–1.15, Fig. 3c). The heterogeneities of above analyses were none (I2 = 0, P = 0.89), low (I2 = 24%, P = 0.24), and moderate (I2 = 63%, P = 0.04).
About observational studies, better results were seen in the MT + IVT group about the outcomes of excellent outcomes (OR:1.49, 95%CI:1.26–1.75, Supplemental Fig. 4a) with low heterogeneity (I2 = 4%, P = 0.40). We also observed the MT + IVT group reduced the risks of mortality by 27% (OR:0.73, 95%CI: 0.56–0.94, Supplemental Fig. 4d) with large heterogeneity (I2 = 67%, P = 0.40) between two groups. And no significant differences were seen in the outcomes of SR (OR:1.21, 95%CI:0.85–1.74, Supplemental Fig. 4b) with large heterogeneity (I2 = 74%, P<0.01), and aICH (OR:1.06, 95%CI:0.83–1.35, Supplemental Fig. 4c) by large heterogeneity (I2 = 28%, P = 0.22).
Subgroup analysis
Due to the limited number of included RCTs, advanced subgroup analysis was performed solely in observational studies. Among the subgroup of study area, there were no distinguishable differences in the outcomes of FI (P = 0.25), excellent outcomes (P = 0.20), sICH (P = 0.31), and mortality (P = 0.53), except for SR (P = 0.04). Specifically, there was non-significance in Asia between two groups (OR:0.59, 95% CI:0.29–1.21). However, in contrast to the dMT therapy, the MT + IVT therapy raised the rate of SR by 51% in Europe (OR:1.51, 95%CI:1.23–1.86). All details were depicted in Supplemental Table 1.
No discernible differences were observable in outcomes of FI (P = 0.93), excellent outcomes (P = 0.22), SR (P = 0.57), sICH (P = 0.82) and aICH (P = 0.96) within the subgroup of study design between the two groups, except for mortality (P = 0.03). In prospective studies, MT + IVT therapy reduced the risk of mortality by 47% (OR:0.53, 95%CI: 0.43–0.78). Retrospective analyses, however, did not reveal significant differences in the findings (OR:0.95, 95%CI: 0.68–1.34). All details were displayed in Supplemental Table 3.
Sensitivity analysis
The sensitivity analysis of RCTs in crude data showed the effects of sICH (Supplemental Fig. 5b), SR (Supplemental Fig. 5d), and mortality (Supplemental Fig. 5f) were not substantially modified by exclusion of a certain study. The effect size of FI varied (OR:1.15, 95% CI:0.97–1.35, Supplemental Fig. 5a) when one study was excluded[22]. When the trial was eliminated[8], the total effect sizes showed no discernible improvement (OR:1.18, 95%CI: 0.99–1.40) in the excellent outcome of MT + IVT therapy. When this study was excluded[22], a similar outcome (OR:1.07, 95% CI:0.89–1.28) was observed. And the MT + IVT group did not increase the risk of aICH (Supplemental Fig. 5d) while removing the study[23] and the trial[24], the effect sizes were (OR:1.16, 95% CI:0.95–1.41) and (OR: 1.18, 95% CI: 0.98–1.44,), respectively. Similar outcomes were seen in the outcome of excellent outcomes (Supplemental Fig. 5c). As followed by the sensitivity analysis of RCTs in adjusted data, the direction of effect size did not change in the outcomes of our interest (Supplemental Fig. 6b-f) except for the FI. The MT + IVT therapy significantly increased FI (OR:1.23, 95%CI:1.03–1.48, Supplemental Fig. 6a) after eliminating the study[25].
As with the above analyses with observational studies, no significant differences were found in the outcomes of observational studies about crude data (Supplemental Fig. 7a, c-f), with the exception of the sICH (Supplemental Fig. 7b). When excluding the study[26], the effect of direction changed (OR:1.11, 95%CI: 0.96–1.29). Referring to observational studies of adjusted data, there were no discernible variations in the outcomes (Supplemental Fig. 8a-f).
Publication bias
For observational studies of crude data, the inspection of contour-enhanced funnel plots showed evidence of asymmetry of outcomes of FI (Fig. 4a), aICH (Supplemental Fig. 9c), and mortality (Supplemental Fig. 9d). No asymmetry was seen in the outcomes of excellent outcomes (Supplemental Fig. 9a), SR (Supplemental Fig. 9b), and sICH (Fig. 4b). However, there was no evidence in the corresponding Peter's statistical tests for funnel plot asymmetry in terms of the outcomes of FI (P = 0.06), excellent outcomes (P = 0.56), SR (P = 0.83), sICH (P = 0.89), aICH (P = 0.14), and mortality (P = 0.21).
The inspection of contour-enhanced funnel plots for observational studies with adjusted data revealed indications of asymmetries in outcomes of FI and sICH (Fig. 5a-b). There was no asymmetry in the mortality results (Supplemental Fig. 10). Additionally, except for sICH (P = 0.01), there was no indication of funnel plot asymmetry in the appropriate Egger's statistical tests for the outcomes of FI (P = 0.46) or mortality (P = 0.67). We did not run the funnel plot, Peter's, or Egger's statistical tests due to the numerous limitations of including RCTs and other observational studies.