Editorial Comment—Mechanical Embolus Removal
A New Day Dawning
We must applaud the resourceful inventors, manufacturer, and investigators of the Merci Retriever device for completing the decade-long odyssey of directing the device through the design, development, study, and Food and Drug Administration (FDA) approval process. A number of devices have failed to climb that hill, falling back in pilot trials despite enthusiasm for their potential. However, previous editorials regarding approval of devices in general, and the potential Pandora’s box of approval of the Merci Retriever in particular,1,2,3⇓⇓dictate that we evaluate the data that gained its approval as an embolectomy and revascularization device, as finally presented in this journal.
The device is intended to restore blood flow in the neurovasculature by removing thrombus in patients experiencing ischemic stroke.4 The manuscript gives no confirmation that flow restoration in this study is due to thrombus removal (embolectomy), as opposed to clot disruption with proximal revascularization and distal embolization. Aggressive clot manipulation more quickly opens primary occlusions with fibrinolytic agent.5,6⇓ No substantive data on wire-based clot manipulation without lytic therapy exists. An abstract from the February 2005 issue of Stroke suggested 50% subject clot removal with use of the Merci device.7 It was suggested that the 50% who might have a clot removed will be the same 50% that recanalize, and therefore that those who improve are among the 50% with clot removal. Nevertheless, even at that level of effectiveness, it is unclear that the device is doing what it is approved to do, which is to remove clots dependably. This embolectomy and/or revascularization issue seems a regulatory Scylla-and-Charybdis issue that might be of lesser import if clinical trial outcomes had been more impressive.
The primary Mechanical Embolus Removal in Cerebral Ischemia (MERCI) study outcome was the rate of recanalization of the terminal internal carotid artery (ICA-T), M1, M2, or basilar arteries, compared with 18% spontaneous recanalization of M1 or M2 in the Prolyse in Acute Cerebral Thromboembolism II (PROACT II) trial.8 PROACT II had a central core laboratory for determining TIMI recanalization outcomes. The MERCI study reports Thrombolysis in Myocardial Infarction (TIMI) scores as determined by the operator. Unfortunately, interobserver variability in determination of recanalization efficacy hasn’t been validated. An interim report from Interventional Management of Stroke II (IMS II) indicates a 41% discrepency in central laboratory versus operator in ascribing TIMI flow, with the operator usually overscoring.
We still lack a conventional, validated revascularization analysis method, and this is an obstacle to comparing results obtained with one treatment paradigm to those with another.9 The MERCI manuscript lacks a specific definition of TIMI recanalization as applied in this study, including recanalization of the basilar artery. It seems clear that the authors evaluated recanalization of the primary occlusions in the ICA, M1, M2, and basilar arteries, and did not report distal perfusion or more distal MCA emboli.
The term “TIMI recanalization” or “TIMI perfusion/reperfusion” has been used with different meanings in different case series and studies. It has even been applied to magnetic resonance angiography (MRA) analysis. Recanalization does not equal reperfusion. Arteriographic demonstration of flow restoration, or revascularization, has two components: recanalization of the arterial occlusive lesion and subsequent proximal branches, and reperfusion into the distal arterial bed of the occluded vessel, including terminal branches with tissue staining, as applied in the original TIMI definition as an angiographic perfusion measure. Recanalization of the primary occlusion may be complete, but with variable distal patency and reperfusion because of preexisting emboli or emboli released by the revascularization procedure itself. The effect of distal emboli is not well understood in the setting of intra-arterial thrombolytic therapy, but is generally thought less deleterious than an unrecanalized proximal occlusion. Conversely, recanalization may be incomplete, with or without distal patency and perfusion, predisposing to reocclusion with clinical deterioration. The 2 components of revascularization have been analyzed in IMS I, and both (re)perfusion (taking distal flow and emboli limiting distal antegrade perfusion into account) and recanalization of the primary arterial occlusive lesion, have been associated with higher Rankin 0 to 2 outcome (P. Khatri, unpublished data, 2005). Other treatment paradigms might not achieve that same result.
The MERCI study fails to confirm that the recanalization definition and its application are comparable to those of PROACT II. In the absence of a guarantee of a comparable, reproducible definition of TIMI 2 to 3 reperfusion and its application, comparing the outcomes is difficult. Confirmatory post hoc analysis by the same core laboratory that reviewed arteriograms in PROACT II, using the same criteria, would be comforting. However, the PROACT core laboratory would need a convention for scoring recanalization and flow in revascularization of the carotid T occlusion and basilar artery occlusion, not evaluated in PROACT II.
The major complication rates reported are probably acceptable for patients with large deficits and poor predictable outcomes. Rates of symptomatic intracerebral hemorrhage (ICH) (the Achilles’ heel of thrombolytic therapy), as well as asymptomatic ICH, seem acceptable. Vessel ruptures/perforations occurred, as they have elsewhere, including 3 of 64 (4.7%) patients in IMS I with no definite, direct, deleterious effect.10 Distal emboli are a threat during a revascularization procedure, and 3 emboli to the anterior cerebral artery (ACA) during middle cerebral artery (MCA) revascularization are reported in the MERCI study. However, that same phenomenon also occurs during MCA microcatheter thrombolysis: of 101 MCA occlusions in IMS I, and a local registry, 4 new ACA occlusions occurred during thrombolysis. Analysis of ACA emboli during 55 ICA-T occlusion thrombolysis treatments indicates 25% of patients had demonstrable occlusion beyond the A1 segment (A2-4) branches prior to therapy. Therapy led to 15% new, distal occlusions, not previously demonstrated, after therapy. In most T occlusions, ipsilateral ACA flow is maintained via the opposite carotid artery through the anterior communicating artery. Distal ACA emboli may lead to reduced collateral flow with MCA occlusion, which may be critical when MCA recanalization does not occur. New occlusions do occur during therapy, and we must find a way to identify and measure the effect of not only those in previously uninvolved distributions such as the ACA, but also more distal emboli in the MCA. Using both a vessel recanalization score, a perfusion score, a combination of both, or some other perfusion measure, may give us that insight. One treatment paradigm or device may differ from another in creating, or negating, secondary emboli.11
It is hard to digest the results so heavily weighted to historical comparison to observations from PROACT II, but with the primary and secondary efficacy outcome measures reversed. Certainly a drug that achieved 66% vessel recanalization seems as worthy of approval as a device with 50% recanalization. If the PROACT II study’s primary efficacy outcome were based on the same recanalization measure, it succeeded more admirably than MERCI. Any revascularization study (performed in a reasonable time window) that examines Rankin 0 to 2 outcomes in recanalizers versus nonrecanalizers should prove effective! Recanalization has been shown to be associated with favorable outcome in numerous case series and studies. If the PROACT II study’s secondary efficacy outcome were Rankin 0 to 2 outcomes in recanalizers versus nonrecanalizers, it may have been as or more successful than MERCI. Favorable secondary historical comparisons of the MERCI MCA group to the control group of PROACT II is tantalizing, but may be a siren’s song to be avoided. Other variables may be confounding.
The mortality expected in the target group may be overstated according to other historical data. In the National Institute of Neurological Disorders and Stroke (NINDS) trial, mortality in the control group for patients of National Institutes of Health Stroke Scale (NIHSS) score >20, where the likelihood of a major occlusion must have been high, was 33%, compared with 43.5% overall mortality in the MERCI treated group. Mortality was 54.2% in the MERCI nonrecanalized group, increasing to 71% where adjuvant therapy was further applied unsuccessfully. This might be interpreted that the therapy attempt can do harm, and the more you do, the more harm can be done. The goal to reopen the vessel must first be tempered with recognition that 13% of the control group of NINDS with NIHSS score >20, and 25% of the control group of PROACT II, achieved independence.
Technical improvements are being made in the device. The device is appropriately being studied further in the Multi-MERCI registry and MR-Rescue Trial. The device will be included for use in the forthcoming IMS III Trial, where 900 patients will be randomized to standard-dose IV alteplase versus reduced-dose alteplase followed by arterial intervention. Multi-Merci and IMS III may show that the device is more effective when applied at 3 to 4 hours after unsuccessful lytic therapy, rather than at 6 hours as in the MERCI trial. MR-Rescue should show that the device can be applied effectively up to 8 hours in subjects selected on the basis of diffusion/perfusion mismatch. Adjuvant therapies must be controlled and limited in these studies so that the issues of efficacy of the device aren’t clouded by safety issues introduced by superimposed therapies with untested safety allowances, and limited further predictable benefit.
Certainly we have seen the rosy dawning of a new day in ischemic stroke therapy study. However, the MERCI device should be viewed as a stop along the path, and not the end of the journey itself.
Furlan AJ, Fischer M. Devices, drugs, and the food and drug administration. Increasing implications for ischemic stroke. Stroke. 2005; 36: 398–399.
Becker KJ, Brott TG. Approval of the MERCI clot retriever. A critical review. Stroke. 2005; 36: 400–403.
Felten RP, Pena C, Provost MC, Schlosser MJ, Witten CM: The Food and Drug Administration medical device review process: clearance of a clot retriever for use in ischemic stroke. Stroke. 2005; 36: 404–406.
Neurological Device Panel Meeting, February 23, 2004. http://www.fda.gov/ohrms/dockets/ac/04/transcripts/4022t1.htm. Accessed June 6, 2005.
Noser EA, Shaltoni HM, Hall CE, Alexandrov AV, Garami Z, Cacayorin ED, Song JK, Grotta JC, Campbell MS 3rd: Aggressive mechanical clot disruption: a safe adjunct to thrombolytic therapy in acute stroke? Stroke. 2005; 36: 292–296.
Qureshi AI, Siddiqui AM, Suri MF, Kim SH, Ali Z, Yahia AM, Lopes DK, Boulos AS, Ringer AJ, Saad M, Guterman LR, Hopkins LN: Aggressive mechanical clot disruption and low-dose intra-arterial third-generation thrombolytic agent for ischemic stroke: a prospective study. Neurosurgery. 2002; 51: 1319–1329.
Marder V, Chute D, Starkman S, Abolian A; Kidwell C, Ovbiagele B, Vinuela F, Duckwiler G, Jahan R, Rajajee V, Selco S, Saver JL: Histology of thrombi retrieved from acute ischemic stroke patients by endovascular embolectomy. Stroke. 2005; 36: 516.[Astract.]
Furlan A, Higashida R, Wechsler L, Gent M, Rowley H, Kase C, Pessin M, Ahuja A, Callahan F, Clark WM, Silver F, Rivera F. Intra-arterial prourokinase for acute ischemic stroke. The PROACT II study: a randomized controlled trial. Prolyse in Acute Cerebral Thromboembolism. JAMA. 1999; 282: 2003–2011.
Higashida RT, Furlan AJ, Roberts H, Tomsick T, Connors B, Barr J, Dillon W, Warach S, Broderick J, Tilley B, Sacks D; Technology Assessment Committee of the American Society of Interventional and Therapeutic Neuroradiology; Technology Assessment Committee of the Society of Interventional Radiology: Trial design and reporting standards for intra-arterial cerebral thrombolysis for acute ischemic stroke. Stroke. 2003; 34: e109–e137.
IMS Study Investigators. Combined intravenous and intra-arterial recanalization for acute ischemic stroke: the Interventional Management of Stroke (IMS) study. Stroke. 2004; 35: 904–912.
Yoneyama T, Nakano S, Kawano H, Iseda T, Ikeda T, Goya T, Wakisaka S: Combined direct percutaneous transluminal angioplasty and low-dose native tissue plasminogen activator therapy for acute embolic middle cerebral artery trunk occlusion. AJNR Am J Neuroradiol. 2002; 23: 277–281.