On 2015 Dec 24, Andrea Messori commented:

Meta-analysis on thrombectomy in acute ischemic stroke: is the degree of heterogeneity influenced by the type of device?

In the overall analysis of 8 randomized controlled trials (RCTs), the meta-analytic results reported by Badhiwala et al (1) for the end-point of functional independence at 90 days contained a significant degree of heterogeneity (I² = 75.4%, p-value for heterogeneity <0.01 according to the results shown in Figure 1 published on page 1839 by these authors; data from 8 RCTs). Since the presence of heterogeneity is considered a negative factor in terms of reliability of the meta-analytic results (2-5), this finding represents an important limitation of the study by Badhiwala et al.; in their article, these authors have correctly pointed out this limitation. Likewise, the meta-analysis recently published by Elgendy et al. (6) on the same topic shared the same limitation and in fact contained a significant degree of heterogeneity (I² = 54.0%, p-value for heterogeneity= 0.021 according to the results shown in Figure 1 published on page 2501 by these authors; data from 9 RCT).

In their papers, Badhiwala et al. (1) and Elgendy et al.(6) have discussed which factors could be responsible for this significant heterogeneity. For example, the time window for the thrombectomy procedure, the type of device, and the use of functional perfusion imaging before randomization were thought to be implicated. In this context, the type of device is likely to play an important role as shown in previous studies (7).

For this reason, given that the Solitaire device was the one most commonly employed in this clinical material, we identified which trials among those examined Badhiwala et al. and by Elgendy et al. were based the comparison between a group receiving the Solitaire device (treatment group) and the control. group. The following four trials were found to be based on this comparison: ESCAPE, EXTEND-A, SWIFT-PRIME, and REVASCAT. In these four trials, the crude rates for the end-point of functional independence at 90 days were the following (comparison: Solitaire group vs control group): ESCAPE, 43 /147 vs. 87 /164; EXTEND-A, 14 /35 vs. 25 /35; SWIFT-PRIME, 33 /83 vs. 59 /98; REVASCAT, 29 /103 vs. 45 /103 (data shown in Figure 1 published on page 1839 by Badhiwala et al).

We have analyzed these data by application of the random-effect model (DerSimonian and Laird method as implemented in the Open Meta-analysis software) and we have obtained the following meta-analytic results: odds-ratio = 2.47 (95% confidence interval[CI], 1.84 to 3.33; I² = 0%, p-value for heterogeneity = 0.689); relative risk = 1.66 (95%CI, 1.40 to 1.97; I² = 0%, p-value for heterogeneity = 0.821). This re-analysis has some interest because, after the selection of a single device, the degree of heterogeneity was markedly reduced and changed from a statistically significant level in the overall analysis to 0%. In conclusion, our re-analysis indicates that the type of device can have an important role influencing the results of these two meta-analyses.

References

1) Badhiwala JH, Nassiri F, Alhazzani W, Selim MH, Farrokhyar F, Spears J, Kulkarni AV, Singh S, Alqahtani A, Rochwerg B, Alshahrani M, Murty NK, Alhazzani A, Yarascavitch B, Reddy K, Zaidat OO, Almenawer SA. Endovascular Thrombectomy for Acute Ischemic Stroke: A Meta-analysis. JAMA. 2015 Nov 3;314(17):1832-43.

2) Gagnier JJ, Morgenstern H, Altman DG, Berlin J, Chang S, McCulloch P, Sun X, Moher D; Ann Arbor Clinical Heterogeneity Consensus Group. Consensus-based recommendations for investigating clinical heterogeneity in systematic reviews. BMC Med Res Methodol. 2013 Aug 30;13:106.

3) Pigott T, Shepperd S. Identifying, documenting, and examining heterogeneity in systematic reviews of complex interventions. J Clin Epidemiol. 2013 Nov;66(11):1244-50.

4) Laurin D, Carmichael PH. Combining or not combining published results in the presence of heterogeneity? Am J Clin Nutr. 2010 Sep;92(3):669-70,

5) Bollen CW, Uiterwaal CS, van Vught AJ. Pooling of trials is not appropriate in the case of heterogeneity. Arch Dis Child Fetal Neonatal Ed. 2006 May;91(3):F233-4.

6) Elgendy IY, Kumbhani DJ, Mahmoud A, Bhatt DL, Bavry AA. Mechanical Thrombectomy for Acute Ischemic Stroke: A Meta-Analysis of Randomized Trials. J Am Coll Cardiol. 2015 Dec 8;66(22):2498-505.

7) Messori A, Fadda V, Maratea D, Trippoli S. New endovascular devices for acute ischemic stroke: summarizing evidence by multiple treatment comparison meta-analysis. Ann Vasc Surg. 2013 Apr;27(3):395-6.

On 2015 Dec 24, Andrea Messori commented:

Meta-analysis on thrombectomy in acute ischemic stroke: is the degree of heterogeneity influenced by the type of device?

In the overall analysis of 8 randomized controlled trials (RCTs), the meta-analytic results reported by Badhiwala et al (1) for the end-point of functional independence at 90 days contained a significant degree of heterogeneity (I² = 75.4%, p-value for heterogeneity <0.01 according to the results shown in Figure 1 published on page 1839 by these authors; data from 8 RCTs). Since the presence of heterogeneity is considered a negative factor in terms of reliability of the meta-analytic results (2-5), this finding represents an important limitation of the study by Badhiwala et al.; in their article, these authors have correctly pointed out this limitation. Likewise, the meta-analysis recently published by Elgendy et al. (6) on the same topic shared the same limitation and in fact contained a significant degree of heterogeneity (I² = 54.0%, p-value for heterogeneity= 0.021 according to the results shown in Figure 1 published on page 2501 by these authors; data from 9 RCT).

In their papers, Badhiwala et al. (1) and Elgendy et al.(6) have discussed which factors could be responsible for this significant heterogeneity. For example, the time window for the thrombectomy procedure, the type of device, and the use of functional perfusion imaging before randomization were thought to be implicated. In this context, the type of device is likely to play an important role as shown in previous studies (7).

For this reason, given that the Solitaire device was the one most commonly employed in this clinical material, we identified which trials among those examined Badhiwala et al. and by Elgendy et al. were based the comparison between a group receiving the Solitaire device (treatment group) and the control. group. The following four trials were found to be based on this comparison: ESCAPE, EXTEND-A, SWIFT-PRIME, and REVASCAT. In these four trials, the crude rates for the end-point of functional independence at 90 days were the following (comparison: Solitaire group vs control group): ESCAPE, 43 /147 vs. 87 /164; EXTEND-A, 14 /35 vs. 25 /35; SWIFT-PRIME, 33 /83 vs. 59 /98; REVASCAT, 29 /103 vs. 45 /103 (data shown in Figure 1 published on page 1839 by Badhiwala et al).

We have analyzed these data by application of the random-effect model (DerSimonian and Laird method as implemented in the Open Meta-analysis software) and we have obtained the following meta-analytic results: odds-ratio = 2.47 (95% confidence interval[CI], 1.84 to 3.33; I² = 0%, p-value for heterogeneity = 0.689); relative risk = 1.66 (95%CI, 1.40 to 1.97; I² = 0%, p-value for heterogeneity = 0.821). This re-analysis has some interest because, after the selection of a single device, the degree of heterogeneity was markedly reduced and changed from a statistically significant level in the overall analysis to 0%. In conclusion, our re-analysis indicates that the type of device can have an important role influencing the results of these two meta-analyses.

References

1) Badhiwala JH, Nassiri F, Alhazzani W, Selim MH, Farrokhyar F, Spears J, Kulkarni AV, Singh S, Alqahtani A, Rochwerg B, Alshahrani M, Murty NK, Alhazzani A, Yarascavitch B, Reddy K, Zaidat OO, Almenawer SA. Endovascular Thrombectomy for Acute Ischemic Stroke: A Meta-analysis. JAMA. 2015 Nov 3;314(17):1832-43.

2) Gagnier JJ, Morgenstern H, Altman DG, Berlin J, Chang S, McCulloch P, Sun X, Moher D; Ann Arbor Clinical Heterogeneity Consensus Group. Consensus-based recommendations for investigating clinical heterogeneity in systematic reviews. BMC Med Res Methodol. 2013 Aug 30;13:106.

3) Pigott T, Shepperd S. Identifying, documenting, and examining heterogeneity in systematic reviews of complex interventions. J Clin Epidemiol. 2013 Nov;66(11):1244-50.

4) Laurin D, Carmichael PH. Combining or not combining published results in the presence of heterogeneity? Am J Clin Nutr. 2010 Sep;92(3):669-70,

5) Bollen CW, Uiterwaal CS, van Vught AJ. Pooling of trials is not appropriate in the case of heterogeneity. Arch Dis Child Fetal Neonatal Ed. 2006 May;91(3):F233-4.

6) Elgendy IY, Kumbhani DJ, Mahmoud A, Bhatt DL, Bavry AA. Mechanical Thrombectomy for Acute Ischemic Stroke: A Meta-Analysis of Randomized Trials. J Am Coll Cardiol. 2015 Dec 8;66(22):2498-505.

7) Messori A, Fadda V, Maratea D, Trippoli S. New endovascular devices for acute ischemic stroke: summarizing evidence by multiple treatment comparison meta-analysis. Ann Vasc Surg. 2013 Apr;27(3):395-6.