Safety and Efficacy of Endovascular Thrombectomy in Patients With Abnormal Hemostasis
Pooled Analysis of the MERCI and Multi MERCI Trials
Background and Purpose— Patients with abnormal hemostasis are not considered candidates for thrombolysis. We analyzed the MERCI/Multi MERCI cohort as an attempt to establish the risks and benefits of thrombectomy in this patient population.
Methods— Two patient groups were identified: Group 1 (n=35): patients with INR >1.7 or PTT >45 seconds or platelet count <100 000/μL; Group 2 (n=270): patients with INR ≤1.7, PTT ≤45 seconds, and platelet count ≥100 000/μL. Clinical, radiographic, and revascularization outcomes were subsequently compared.
Results— In Group 1, 20 patients had INR >1.7 (mean: 2.4; range: 1.8 to 4.9), 11 had PTT >45 seconds (mean: 95; range: 46 to 190), and 6 had platelets <100 000/μL (mean: 63 400; range: 16 000 to 94 000). Two patients had both INR >1.7 and PTT >45 seconds. The two groups did not significantly differ in terms of age, gender, baseline NIHSS scores, intraarterial thrombolytic use/dosage, or occlusion site. Time-to-treatment was slightly earlier in Group 1. There was no significant difference in the rates of revascularization (TIMI 2 to 3: 60% versus 65%), mortality (40% versus 38%), or major symptomatic intracranial hemorrhage (SICH; 8.6% versus 8.5%). Group 2 had higher rates of good clinical outcomes (9% versus 35%; P=0.002). This was likely related to a lower prestroke health status in Group 1 patients. In Group 1, successful revascularization was associated with improved outcomes (P=0.015) and lower mortality (24% versus 64%; P=0.033).
Conclusion— Patients with abnormal hemostasis who undergo thrombectomy do not appear to be at a higher risk for SICH but have lower rates of good outcomes. In this patient group, successful revascularization appears to be associated with improved clinical outcomes and lower mortality.
- acute stroke
- endovascular treatment
- interventional neuroradiology intracerebral hemorrhage
- intracranial hemorrhage
Patients with significantly abnormal hemostasis at stroke onset including international normalized ratio (INR) greater than 1.7, elevated partial-thromboplastin time (PTT), or platelet count less than 100 000/μL are not considered candidates for thrombolysis with intravenous rt-PA.1 Similar criteria have been arbitrarily adopted by studies evaluating intraarterial (IA) thrombolysis, including the PROACT I and II as well as the IMS I and II trials.2–5 Therefore, the risks and benefits of endovascular treatment for acute stroke in patients with abnormal hemostasis remain unknown.
Endovascular thrombectomy with the Merci Retrieval System (Concentric Medical Inc) may obviate or lessen the use of thrombolytic drugs. Consequently, different inclusion criteria for baseline hemostasis were adopted during the MERCI and Multi MERCI trials.6,7 These trials enrolled patients who were on anticoagulation as long as their INR was less than 3.0 and their PTT was less than 2-times the control. In addition, the platelet count thresholds used in these trials were substantially lower than what had been stipulated in previous trials (>50 000/μL in MERCI-Part I and >30 000/μL in MERCI-Part II and Multi MERCI). We thereby performed a retrospective analysis of the pooled MERCI and Multi MERCI cohorts to compare the periprocedural rates of intracranial hemorrhage (ICH) and other hemorrhagic complications, vessel recanalization, mortality, and clinical outcomes between patients with and without “significantly abnormal hemostasis.”
In the current analysis, the entire patient cohort from the MERCI (Mechanical Embolus Removal in Cerebral Ischemia) and Multi MERCI trials were combined in a single dataset totalizing 305 patients. Two patient groups were identified. Group 1 (n=35) included patients with INR >1.7 or PTT >45 seconds or platelet count <100 000/μL. Group 2 (n=270) included patients with INR ≤1.7, PTT ≤45 seconds and platelet count ≥100 000/μL.
Patients and Techniques
The MERCI trial was a prospective single-arm multicenter trial designed to test the safety and efficacy of the Merci Retriever to restore the patency of intracranial arteries in patients ineligible for IV rt-PA during the first 8 hours of an acute stroke. The occlusion sites were the intracranial vertebral artery, basilar artery, internal carotid artery (ICA), or proximal middle cerebral artery branches (M1 or M2 segments).6 The Multi MERCI trial was an international multicenter single-arm trial with 3 objectives: to gain greater experience with the first-generation Merci Retrieval devices (X5 and X6) in patients ineligible for IV rt-PA; to explore the safety and technical efficacy of the Merci Retriever in patients treated with IV rt-PA who failed to recanalize; and to collect safety and technical efficacy data on a second-generation retriever (L5).7 Inclusion criteria and techniques were otherwise similar to the ones used in the MERCI trial.
The technical details regarding the procedures used in MERCI and Multi MERCI have been previously described.6,7 Successful recanalization was defined as achieving TIMI (Thrombolysis In Myocardial Infarction) II or III flow in all treatable vessels.8 Intraarterial thrombolytics were used in cases of device failure after 6 passes or to treat distal emboli not accessible to the device after successful proximal thrombectomy. Angioplasty or stenting were not allowed.
Clinical Variables and Measurement of Outcome
The baseline clinical variables and the clinical, angiographic, and radiographic outcome measurements were similar in the MERCI and Multi MERCI trials. In both trials, postprocedural head computed tomography (CT) was performed at 24 hours or for any deterioration in neurological status. Symptomatic intracranial hemorrhage (SICH) was defined as a 4 or more point worsening of the NIHSS score within 24 hours with any blood products identified on head CT scan, or any ICH in which no further NIHSS scores were available beyond baseline and the patient died. Asymptomatic hemorrhage was defined as evidence of any blood on the CT or MRI scans at 24 hours with less than a 4-point decline in the NIHSS score. All head CT images were reviewed in a core laboratory. Hemorrhages were adjudicated by the Data Safety Monitoring Board (DSMB). SICHs were further categorized in hemorrhagic infarction (HI) type 1 and type 2 and parenchymal hematoma (PH) type 1 and type 2, according to previously published criteria.9
Primary outcomes included the vascular recanalization rate and the rate of procedure-related complications. Procedure-related adverse events were defined as vascular perforation, arterial dissection, embolization of a previously uninvolved territory, symptomatic hemorrhage adjudicated as procedure-related, and access site complications requiring surgery or transfusion. Clinically significant procedural complications were defined as a procedure complication with decline in NIHSS of ≥4 points or death, groin complication requiring surgery, or blood transfusion. Secondary outcomes included clinical outcomes, as measured by the modified Rankin Scale (mRS) at 90 days, and 90-day mortality. Good neurological outcome was prespecified as mRS ≤2. In this study, we also analyzed the cut-point of mRS ≤3 (“acceptable outcome”) because patients with abnormal hemostasis were expected to have a poorer health status at baseline.
Comparisons were made to determine whether any differences existed in terms of predictors of stroke outcome among the two groups. These included the following variables: age, gender, baseline NIHSS score, time to treatment, use and dosage of intraarterial (IA) thrombolytics, and location of the occlusive thrombus. The outcome variables were then compared among the 2 groups. Categorical data in contingency tables were analyzed by the Fisher exact test. Normally distributed continuous data were analyzed by 2-sample t test. Continuous data not approximately normal in distribution were compared using the Wilcoxon rank-sum test. Ordinal data were analyzed by the Cochran-Mantel-Haenszel test. Means, standard deviations, medians, and ranges were calculated for all continuous data. Relative risks were calculated along with their respective 95% confidence intervals for dichotomous data. Given all analyses are posthoc, probability values are provided for descriptive purposes. All computations were performed with the aid of SAS software (version 8.2; SAS Institute Inc).
The data on the 35 patients in Group 1 is summarized in Table 1⇓. Twenty of these patients had INR greater than 1.7 (mean: 2.4; range: 1.8 to 4.9), 11 had PTT greater than 45 seconds (mean: 95; range: 46 to 190), and 6 had platelet count less than 100 000/μL (mean: 63,400; range: 16 000 to 94 000). Two patients had both INR greater than 1.7 and PTT greater than 45 seconds. All patients with elevated INR or PTT had been treated with warfarin or heparin for the indications summarized in Table 1⇓. The etiologies of thrombocytopenia were more diverse and included chemotherapy and sepsis (n=1), sepsis alone (n=1), malignancy-related disseminated intravascular coagulopathy (DIC; n=1), drug reaction (n=1), and recent cardiac surgery with extracorporeal circulation±heparin-induced thrombocytopenia (n=2). Thus, iatrogenic/therapeutic coagulopathies accounted for 82.8% (29/35) of the abnormal hemostasis cases.
The two groups did not significantly differ in terms of age (mean: 67.5±15.4 versus 67.6±15.8 years; P=0.97), gender (female: 60% versus 51%; P=0.37), baseline NIHSS scores (mean: 20.9±7.8 versus 19.5±6.3; P=0.25), or site of occlusion (ICA: 34% versus 32%; MCA: 57% versus 59%; Vertebrobasilar: 9% versus 9%). Fourteen patients with “significantly abnormal hemostasis” received concomitant IA thrombolytics and both the frequency of IA thrombolytic use (40% versus 31%; P=0.33), and IA rt-PA dosage (mean: 9.7±9.8 mg versus 11.5±7.6 mg; P=0.11) did not differ between groups. Time-to-treatment was slightly earlier in Group 1 (mean: 3.74±1.64 versus 4.43±1.76 hours; P=0.027). Forty-eight (17.8%) patients in Group 2 and none of the patients in Group 1 received intravenous rt-PA. Group 1 patients had a higher incidence of both atrial fibrillation (66% versus 39%; P=0.004) and diabetes mellitus (34% versus 18%; P=0.04) than group 2 patients. No statistically significant differences were found in the incidences of coronary artery disease (47% versus 40%), congestive heart failure (26% versus 18%), dyslipidemia (41% versus 33%), peripheral vascular disease (17% versus 12%), hypertension (69% versus 72%), or smoking (21% versus 23%) between the 2 groups.
No formal protocol was adopted to manage abnormal hemostasis during the MERCI and Multi MERCI trials—this obviously led to a heterogeneous approach. However, none of the patients with elevated INR or PTT had their coagulopathy corrected before the endovascular procedure. Two patients with abnormal INR who were treated under protocol violation (INR: 4.7 and 4.9) were treated with fresh frozen plasma (FFP) and vitamin K postprocedure. One patient with preprocedural INR of 1.8 received FFP 1 day postprocedure to reverse his increasing INR in the setting of progression of his stroke and marked cerebral edema. One patient with preprocedural INR of 2.4 and PTT of 35 seconds who received 3000 U of intravenous heparin bolus at beginning of the procedure was treated with 20 mg of protamine sulfate at end of procedure. One patient with preprocedural PTT of 150 seconds in the setting of heparinization during a cardiac ablation was reversed with protamine sulfate at the end of the thrombectomy procedure. Two patients with advanced cancer and transfusion-refractory thrombocytopenia received periprocedural platelet transfusion.
The outcomes for both groups are summarized in Table 2. There was no significant difference in terms of overall revascularization (TIMI 2 to 3: 60% versus 65%; P=0.58) or mortality rates (40% versus 38%; P=0.85) between the 2 groups. There were higher rates of good clinical outcomes (mRS ≤2) at 90 days in Group 2 (9% versus 35%; P=0.002) but no significant difference was seen in terms of “acceptable outcomes” (mRS ≤3) at 90 days between the two groups (28% versus 46%; P=0.062). In Group 1, successful revascularization was associated with a lower mortality (24% versus 64%; P=0.033; RR: 0.37 [95%CI: 0.16 to 0.87]) and an overall improvement in clinical outcomes at 90 days (P=0.015), including a trend toward a higher rate of good clinical outcomes (mRS ≤2: 17% versus 0%; P=0.24) and a statistically significant higher rate of “acceptable outcomes” (mRS ≤3) at 90 days (44% versus 7%; P=0.04; Figure 1).
Group 1 had 3 ICHs with significant amounts of blood resulting in 4 or more point worsening in the NIHSS score. These include 1 PH-2 in a patient with platelet count of 16 000/μL in the setting of gestational trophoblastic cancer, chemotherapy, and sepsis, who was treated under protocol violation (Figure 2A), 1 PH-1 in a patient with INR of 1.8 who also received 30 mg of IA rt-PA (Figure 2B), and 1 subarachnoid hemorrhage (SAH) in the setting of a distal ICA perforation with active contrast extravasation in a patient with INR of 2.3 who was treated at 7.2 hours after symptoms onset (Figure 2C). In addition, there was a case of a small SAH along the cerebral convexity that was also classified as SICH by definition of the DSMB in a patient with platelet count of 64 000/μL in the setting of metastatic gynecological malignancy and presumed DIC, who also received 5 mg of IA rt-PA (Figure 2D). The incidence of SICH did not significantly differ between groups (11.4% versus 8.5%; P=0.53—after excluding minor SAH case: 8.6% versus 8.5%; P=1.0). Similarly, there was no significant difference in the incidence of either symptomatic PH-2 (2.9% versus 1.9%; P=0.52) or clinically significant procedural complications (11.4% versus 5.6%, P=0.25) among the 2 groups. Major groin complications requiring surgery or blood transfusion were observed in 1/35 (2.9%) and 2/270 (0.7%) of the patients in Group 1 and 2, respectively (P=0.31).
This study provides favorable data about the safety and efficacy of endovascular treatment of acute stroke in patients with “significantly abnormal hemostasis.” Even though the hemorrhagic event numbers were relatively low generating wide confidence intervals, there was no suggestion of higher periprocedural complication rates in patients with significantly abnormal INR, PTT, or platelet count. Indeed, the only PH-2 that was seen in this patient group occurred in a subject with a very low platelet count (16 000/μL) who was treated under protocol violation (lower limit of platelet count for eligibility was 30 000/μL). In addition, the remaining SICHs occurred in patients who either received IA rt-PA or suffered intraprocedural vascular perforation. This suggests that a more conservative approach that precludes the use of thrombolytics might result in even lower rates of SICH.
Interestingly, patients with “significantly abnormal hemostasis” had a lower chance of having good clinical outcomes (mRS ≤2) at 90 days despite having similar recanalization rates to patients with “normal hemostasis.” This is presumably related to an apparent lower prestroke health status in this patient group as evident by their higher incidence of atrial fibrillation and diabetes mellitus as compared to the patients with abnormal hemostasis as well as by the high frequency of associated malignancies (17%), ongoing sepsis (5.7%), recent cardiac surgery or myocardial infarction (11.4%), and renal disease with or without hemodialysis (11.4%) that was observed in these patients. There was no difference in the rates of mortality or “acceptable outcome” (mRS ≤3) between the 2 groups. More importantly, recanalization status had a major impact on the final outcome of patients with abnormal hemostasis, supporting the need for aggressive reperfusion therapy in these patients who otherwise are not candidates for intravenous thrombolysis.
Endovascular therapy in patients with elevated INR has been only previously reported in 2 small case series. Linfante et al reported on 2 octogenarian patients who presented with distal MCA branch occlusions (M2 and M3 segments, respectively) in the setting of warfarin therapy and elevated INR (1.7 and 1.9, respectively). Both patients underwent IA thrombolysis with low-dose rt-PA (5 mg and 3.5 mg, respectively), resulting in marked clinical improvement within 48 hours (NIHSS score drop from 17 to 0 and 9 to 2, respectively). There were no hemorrhagic complications.10 Janjua et al reported on 3 consecutive patients (age: 58 to 79 years; NIHSS score: 12 to 17) on active warfarin therapy (baseline INR: 1.99 to 2.25) who were treated with low-dose IA reteplase after anticoagulation was reversed with fresh frozen plasma. There were no SICH. Early neurological improvement occurred in 2 of the patients.11 Our findings further support the safety of endovascular treatment in this patient population. Only 2 SICH occurred in the 20 patients with INR >1.7 who underwent thrombectomy despite the fact that 8/20 patients also received IA thrombolytics. This is the same rate of SICH that was found in the treatment arm of the PROACT II trial.3
To the best of our knowledge, no data have been previously reported on the endovascular treatment of stroke patients with elevated PTT. However, heparin is frequently administered in the setting of endovascular therapy of acute stroke. The optimal level of anticoagulation for these patients has not yet been defined, but the PROACT trial suggested that high-dose heparin increases the risk of SICH.2 Our pooled analysis demonstrated that none of the 11 patients with pretreatment PTT >45 seconds had SICH. This finding further supports the idea of offering endovascular treatment for these patients in a more universal basis.
Significant thrombocytopenia (<100 000/μL) is considered a contraindication for the use of rt-PA, and thrombolytic trials have in general excluded these patients. Therefore, data were previously available only on patients with platelet counts ≥100 000/μL. Platelet counts have been evaluated as a potential risk factor for SICH after thrombolysis in at least 4 previous studies of which only 1 identified lower platelet counts (<150 000/μL) as an independent risk factor.12 In this pooled analysis, we identified 6 patients with platelet count less <100 000/μL who underwent thrombectomy. This limited data did not raise any major safety concerns because only 1 significant ICH (PH-2) occurred in this group and this particular patient was treated outside protocol (platelet count of 16 000/μL). However, none of these patients had a good outcome (mRS of 4 in 2 patients; death in 4 patients) despite successful recanalization in 4 of the 6 patients. This is most likely related to the poor baseline health status of thrombocytopenic patients as seen in our cases, where 3/6 patients had advanced malignances and/or sepsis and another 2/6 patients had recently undergone major cardiac surgeries.
Our analysis is somewhat limited by the overall small number of patients in Group 1 as well as by the diversity of the etiologies of abnormal hemostasis in these patients. In addition, this work carries all the limitations of a posthoc analysis. Moreover, the lack of a formal protocol to manage abnormal hemostasis as well as to standardize the use of intraprocedural heparin and thrombolytics led to a diversity of approaches that could have impacted on our results. Finally, the data regarding the periprocedural management of the abnormal hemostasis patients were collected through retrospective review of the clinical research forms and medical records.
It should also be noted that 1 of the major limitations of the MERCI and Multi MERCI trials in terms of outcome analysis is that these trials did not exclude patients based on their baseline (prestroke) health status. Therefore, many patients with baseline mRS >2 were enrolled. This obviously affected the overall chances of these patients achieving good outcomes (mRS ≤2) and likely unfavorably impacted their mortality rates as well. As discussed above, one would expect these patients with poor baseline health status to be overrepresented in the “abnormal hemostasis” group.
When considering endovascular treatment for patients with abnormal hemostasis one must carefully assess for risk factors for hemorrhagic complications after acute revascularization. These include the use and dose of lytic agents, the presence of edema or mass effect on head CT, high stroke severity, older age, hyperglycemia, concurrent heparin use, and the timing of therapy.13 In addition, given our findings with only 9% of the patients with abnormal hemostasis achieving good outcomes overall and only 17% achieving good outcomes despite successful revascularization, the patient’s family should be educated that even if revascularization is successful, the existing data suggest that the ultimate outcome could still very well be quite poor. Finally, the process underlying the abnormal hemostasis likely plays a major role in the ultimate outcomes of these patients.
Patients with abnormal hemostasis who undergo thrombectomy do not appear to be at a significantly higher risk for SICH or other serious complications. In this patient group, successful revascularization appears to be associated with an overall improvement in clinical outcomes and a lower mortality. However, their outcomes are in general inferior to the outcomes of patients without abnormal hemostasis.
Raul G. Nogueira is a member of the Scientific Advisory Boards for Concentric Medical Inc, ev3 Neurovascular Inc, and Coaxia Inc. Wade S. Smith was the Principal Investigator for the MERCI and Multi MERCI trials and has stock ownership in Concentric Medical Inc.
This work has been presented at the 2008 International Stroke Conference (New Orleans, February 2008).
- Received May 7, 2008.
- Revision received July 1, 2008.
- Accepted July 9, 2008.
del Zoppo GJ, Higashida RT, Furlan AJ, Pessin MS, Rowley HA, Gent M. PROACT: a phase II randomized trial of recombinant pro-urokinase by direct arterial delivery in acute middle cerebral artery stroke. PROACT Investigators. Prolyse in acute cerebral thromboembolism. Stroke. 1998; 29: 4–11.
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.
IMS Trial Investigators. Combined intravenous and intra-arterial recanalization for acute ischemic stroke: the Interventional Management of Stroke Study. Stroke. 2004; 35: 904–911.
IMS Trial Investigators. The Interventional Management of Stroke (IMS) II Study. Stroke. 2007; 38: 2127–2135.
Smith WS, Sung G, Starkman S, Saver JL, Kidwell CS, Gobin YP, Lutsep HL, Nesbit GM, Grobelny T, Rymer MM, Silverman IE, Higashida RT, Budzik RF, Marks MP. Safety and efficacy of mechanical embolectomy in acute ischemic stroke: results of the MERCI trial. Stroke. 2005; 36: 1432–1438.
Smith WS, Sung G, Saver J, Budzik R, Duckwiler G, Liebeskind DS, Lutsep HL, Rymer MM, Higashida RT, Starkman S, Gobin YP; Multi MERCI Investigators, Frei D, Grobelny T, Hellinger F, Huddle D, Kidwell C, Koroshetz W, Marks M, Nesbit G, Silverman IE. Mechanical thrombectomy for acute ischemic stroke: final results of the Multi MERCI trial. Stroke. 2008; 39: 1205–12.
Williams DO, Borer J, Braunwald E, Chesebro JH, Cohen LS, Dalen J, Dodge HT, Francis CK, Knatterud G, Ludbrook P, et al. Intravenous recombinant tissue-type plasminogen activator in patients with acute myocardial infarction: a report from the NHLBI thrombolysis in myocardial infarction trial. Circulation. 1986; 73: 338–346.
Berger C, Fiorelli M, Steiner T, Schabitz WR, Bozzao L, Bluhmki E, Hacke W, von Kummer R. Hemorrhagic transformation of ischemic brain tissue: asymptomatic or symptomatic? Stroke. 2001; 32: 1330–1335.
Khatri P, Wechsler LR, Broderick JP. Intracranial hemorrhage associated with revascularization therapies. Stroke. 2007; 38: 431–440.