Registry Supports MERCI Trials Data, but Methods Criticized

LOS ANGELES - Results of the MERCI patient registry appear to uphold findings from the device’s two pivotal trials of patients with acute ischemic stroke, which demonstrated that successful recanalization is significantly associated with good outcome.

But audience members who spoke at the International Stroke Conference following the presentation of the registry results emphasized – to the applause of others in the audience – that the MERCI device lacks randomized data proving its safety and efficacy, and that reported outcomes for patients have not been stratified according to intubation status.

The registry is a large, nonrandomized case series documenting the postapproval use of the devices, whereas the two pivotal trials (MERCI and Multi MERCI) had strict inclusion and exclusion criteria and protocols but did not compare the device vs. medical therapy. Device-treated patients in the trials were instead compared with a placebo group from a randomized medical trial, called PROACT II (Prolyse in Acute Cerebral Thromboembolism).

At the conference, Dr. Marilyn Rymer presented the 90-day outcomes of 1,000 patients with acute ischemic stroke in the registry who were treated with the MERCI clot retriever embolectomy device. The MERCI (Mechanical Embolus Removal in Cerebral Ischemia) and Multi MERCI studies examined embolectomy with the device in similar patient groups, with a total of 305 patients.

"The registry is designed to answer the question, ‘What does the real-world, unrestrained treatment of ischemic stroke with this device look like?’" said Dr. Rymer, a medical director of the Brain and Stroke Institute at St. Luke’s Hospital in Kansas City, Mo. However, she noted that because the registry consists of nonrandomized cases, the efficacy implied in it "can’t be compared to a medical therapy."

The participating sites included every consecutive patient who was treated with the device. Treatment remained a clinical decision guided by each site’s general practice. The inclusion criteria were a diagnosis of acute ischemic stroke and at least one pass with the tool.

The primary end point was revascularization with a TICI (Thrombolysis in Cerebral Infarction) scale grade of 2a or higher; there was also a secondary functional outcome, which was the modified Rankin Scale (mRS) score at 90 days. The final analysis included 872 patients as a result of excluding 4 with insufficient procedural data and 90 who were disabled before their stroke with an mRS of 2 or more, as well as losing 34 to follow-up.

The patients’ median age was 68 years, compared with the median age of 72 in the trials. There was wide intersite variability with regard to age: At one site, the median age of patients treated was 58, and at another the median age of patients treated was 72.

"We saw the same kind of variation in terms of baseline National Institutes of Health Stroke Scale [NIHSS] score," Dr. Rymer said at the conference, which was sponsored by the American Heart Association. "The median in the registry was 17, while it was 21 in the trials." Median NIHSS scores also varied across the registry sites (range, 14-21), she said.

Overall, 305 patients received intravenous thrombolytic therapy, "But there was an incredible variation among sites" in the use of thrombolytics in conjunction with embolectomy (range, 0%-71%). Intra-arterial thrombolytic therapy also varied widely: Some 47% of patients overall received it, but the intersite rate varied from 7% to 100%.

Overall, 63% of patients were intubated, with the rate varying from 12% to 100%. "Several sites used intubation routinely in 100% of their patients, and some intubated only for airway protection. This becomes important as we begin to understand that intubation is associated with a worse outcome," Dr. Rymer said.

Most sites treated fewer than 19% of patients with angioplasty or stenting in addition to clot retrieval, but one site employed these additional treatments in 64% of patients.

The time from symptom onset to groin insertion was 6.3 hours, compared with 4.5 hours in the MERCI trials. Most patients (71%) were treated 3-8 hours after symptom onset.

Overall in the registry, 80% of patients were successfully recanalized, which was significantly more than in the two MERCI trials combined (65%). Similar numbers of patients had good 90-day outcomes (32% in both the registry and combined trials). Mortality at 90 days also was not significantly different between the registry and the trials (33% and 38%).

Symptomatic brain hemorrhage in the registry was 7% overall, not significantly different from the 8.8% seen in the MERCI trials, Dr. Rymer said. "It is notable that in the patients who were well recanalized (those with a TICI grade of 2b to 3), the symptomatic hemorrhage rate was lower (3.7% in TICI 2b and 5.4% in TICI 3)."

When the investigators examined the rate of good 90-day outcome (defined as an mRS of 0-2), they found the best outcomes in patients with the lowest baseline NIHSS scores. "As the stroke became more severe, the likelihood of good outcome went down," she said. In cases with NIHSS scores lower than 16, "the outcomes were excellent," she said, with up to 70% of those with a TICI grade of 2b or 3 experiencing a good outcome. TICI 2a provided only modest benefit, but it was consistent across the whole range of NIHSS scores, she added.

Age and recanalization status also affected mortality. "Age was a predictor of worse outcome, but recanalization did provide benefit across all ages except for the very young, who had low mortality rates in any case," she said.

A multivariate analysis identified several factors that significantly affected mortality both negatively and positively, including advancing age (odds ratio, 1.05), worse baseline NIHSS score (OR, 1.08), revascularization to a TICI grade of 2a, 2b, or 3 (OR, 0.33), heart failure (OR, 2.85), blood glucose above 140 mg/dL (OR 2.0), and intubation during the revascularization procedure (OR, 2.20)

The same multivariate model also identified factors that negatively impacted good 90-day outcomes, including worse baseline NIHSS score (OR, 0.88), advancing age (OR, 0.96), intubation during the procedure (OR, 0.43), longer duration of procedure (OR, 0.66), and a blood glucose level of 140 mg/dL or greater (OR, 0.59).

During the discussion period, several audience members questioned the relationship between intubation and poor outcomes. Dr. Joseph Broderick, chair of the department of neurology at the University of Cincinnati, said that the intubation data were interesting but could throw a statistical kink into the risk analysis. "How do you take into account those centers with a 100% intubation rate?" he asked. "Why don’t you analyze the data separately and see if the risk of intubation still holds up? There is always a risk of selection bias unless you compare the [sites] that always intubated against those that did not. Otherwise, you might be including people who were intubated because they looked dead, had heart failure, or weren’t breathing."

Three other speakers echoed the same concern, with one suggesting that intubation could lead to aspiration pneumonia, which in turn could contribute to a bad outcome.

"We don’t know what all the facts are" in relation to intubation, Dr. Rymer said. "We can only speculate."

However, the discussants also emphasized that there are still no randomized data confirming the device’s efficacy when compared with medical therapies. "There is, however, a clear association of improved outcomes with better revascularization," Dr. Rymer said in an interview.

Both MERCI and Multi MERCI were criticized for comparing their device-treated populations with only a placebo arm of PROACT II, according to an editorial by Dr. Kyra J. Becker of the University of Washington and Dr. Thomas G. Brott of the Mayo Clinic in Jacksonville, Fla. (Stroke 2005;36:400-3).

They said that the data presented to the Food and Drug Administration for the device’s approval did not meet the standards that could be achieved in a randomized trial.

"These data certainly provide no reassurance about the safety of mechanical embolectomy, yet alone the efficacy," the authors said. "The risks of treatment with the Concentric clot retriever might not be limited to arterial perforation, dissection, or distal embolization; there could be unidentified risks associated with the required immediate poststroke angiography as well as with endovascular instrumentation and injury to the arterial endothelium. Unfortunately, these questions cannot be addressed with the data presented ... by the MERCI investigators. Randomized trials have a distinct advantage in this setting; they can control for variables we know are likely to relate to procedure success and outcome, and they can control for important variables we have yet to identify."

Concentric Medical, the manufacturer of the MERCI device, sponsored the MERCI registry. Dr. Rymer reported being on the speakers bureau of the company. The five other coauthors reported varying relationships with Concentric, including being consultants or medical advisors, and having an ownership interest in Concentric.

LOS ANGELES - Results of the MERCI patient registry appear to uphold findings from the device’s two pivotal trials of patients with acute ischemic stroke, which demonstrated that successful recanalization is significantly associated with good outcome.

But audience members who spoke at the International Stroke Conference following the presentation of the registry results emphasized – to the applause of others in the audience – that the MERCI device lacks randomized data proving its safety and efficacy, and that reported outcomes for patients have not been stratified according to intubation status.

The registry is a large, nonrandomized case series documenting the postapproval use of the devices, whereas the two pivotal trials (MERCI and Multi MERCI) had strict inclusion and exclusion criteria and protocols but did not compare the device vs. medical therapy. Device-treated patients in the trials were instead compared with a placebo group from a randomized medical trial, called PROACT II (Prolyse in Acute Cerebral Thromboembolism).

At the conference, Dr. Marilyn Rymer presented the 90-day outcomes of 1,000 patients with acute ischemic stroke in the registry who were treated with the MERCI clot retriever embolectomy device. The MERCI (Mechanical Embolus Removal in Cerebral Ischemia) and Multi MERCI studies examined embolectomy with the device in similar patient groups, with a total of 305 patients.

"The registry is designed to answer the question, ‘What does the real-world, unrestrained treatment of ischemic stroke with this device look like?’" said Dr. Rymer, a medical director of the Brain and Stroke Institute at St. Luke’s Hospital in Kansas City, Mo. However, she noted that because the registry consists of nonrandomized cases, the efficacy implied in it "can’t be compared to a medical therapy."

The participating sites included every consecutive patient who was treated with the device. Treatment remained a clinical decision guided by each site’s general practice. The inclusion criteria were a diagnosis of acute ischemic stroke and at least one pass with the tool.

The primary end point was revascularization with a TICI (Thrombolysis in Cerebral Infarction) scale grade of 2a or higher; there was also a secondary functional outcome, which was the modified Rankin Scale (mRS) score at 90 days. The final analysis included 872 patients as a result of excluding 4 with insufficient procedural data and 90 who were disabled before their stroke with an mRS of 2 or more, as well as losing 34 to follow-up.

The patients’ median age was 68 years, compared with the median age of 72 in the trials. There was wide intersite variability with regard to age: At one site, the median age of patients treated was 58, and at another the median age of patients treated was 72.

"We saw the same kind of variation in terms of baseline National Institutes of Health Stroke Scale [NIHSS] score," Dr. Rymer said at the conference, which was sponsored by the American Heart Association. "The median in the registry was 17, while it was 21 in the trials." Median NIHSS scores also varied across the registry sites (range, 14-21), she said.

Overall, 305 patients received intravenous thrombolytic therapy, "But there was an incredible variation among sites" in the use of thrombolytics in conjunction with embolectomy (range, 0%-71%). Intra-arterial thrombolytic therapy also varied widely: Some 47% of patients overall received it, but the intersite rate varied from 7% to 100%.

Overall, 63% of patients were intubated, with the rate varying from 12% to 100%. "Several sites used intubation routinely in 100% of their patients, and some intubated only for airway protection. This becomes important as we begin to understand that intubation is associated with a worse outcome," Dr. Rymer said.

Most sites treated fewer than 19% of patients with angioplasty or stenting in addition to clot retrieval, but one site employed these additional treatments in 64% of patients.

The time from symptom onset to groin insertion was 6.3 hours, compared with 4.5 hours in the MERCI trials. Most patients (71%) were treated 3-8 hours after symptom onset.

Overall in the registry, 80% of patients were successfully recanalized, which was significantly more than in the two MERCI trials combined (65%). Similar numbers of patients had good 90-day outcomes (32% in both the registry and combined trials). Mortality at 90 days also was not significantly different between the registry and the trials (33% and 38%).

Symptomatic brain hemorrhage in the registry was 7% overall, not significantly different from the 8.8% seen in the MERCI trials, Dr. Rymer said. "It is notable that in the patients who were well recanalized (those with a TICI grade of 2b to 3), the symptomatic hemorrhage rate was lower (3.7% in TICI 2b and 5.4% in TICI 3)."

When the investigators examined the rate of good 90-day outcome (defined as an mRS of 0-2), they found the best outcomes in patients with the lowest baseline NIHSS scores. "As the stroke became more severe, the likelihood of good outcome went down," she said. In cases with NIHSS scores lower than 16, "the outcomes were excellent," she said, with up to 70% of those with a TICI grade of 2b or 3 experiencing a good outcome. TICI 2a provided only modest benefit, but it was consistent across the whole range of NIHSS scores, she added.

Age and recanalization status also affected mortality. "Age was a predictor of worse outcome, but recanalization did provide benefit across all ages except for the very young, who had low mortality rates in any case," she said.

A multivariate analysis identified several factors that significantly affected mortality both negatively and positively, including advancing age (odds ratio, 1.05), worse baseline NIHSS score (OR, 1.08), revascularization to a TICI grade of 2a, 2b, or 3 (OR, 0.33), heart failure (OR, 2.85), blood glucose above 140 mg/dL (OR 2.0), and intubation during the revascularization procedure (OR, 2.20)

The same multivariate model also identified factors that negatively impacted good 90-day outcomes, including worse baseline NIHSS score (OR, 0.88), advancing age (OR, 0.96), intubation during the procedure (OR, 0.43), longer duration of procedure (OR, 0.66), and a blood glucose level of 140 mg/dL or greater (OR, 0.59).

During the discussion period, several audience members questioned the relationship between intubation and poor outcomes. Dr. Joseph Broderick, chair of the department of neurology at the University of Cincinnati, said that the intubation data were interesting but could throw a statistical kink into the risk analysis. "How do you take into account those centers with a 100% intubation rate?" he asked. "Why don’t you analyze the data separately and see if the risk of intubation still holds up? There is always a risk of selection bias unless you compare the [sites] that always intubated against those that did not. Otherwise, you might be including people who were intubated because they looked dead, had heart failure, or weren’t breathing."

Three other speakers echoed the same concern, with one suggesting that intubation could lead to aspiration pneumonia, which in turn could contribute to a bad outcome.

"We don’t know what all the facts are" in relation to intubation, Dr. Rymer said. "We can only speculate."

However, the discussants also emphasized that there are still no randomized data confirming the device’s efficacy when compared with medical therapies. "There is, however, a clear association of improved outcomes with better revascularization," Dr. Rymer said in an interview.

Both MERCI and Multi MERCI were criticized for comparing their device-treated populations with only a placebo arm of PROACT II, according to an editorial by Dr. Kyra J. Becker of the University of Washington and Dr. Thomas G. Brott of the Mayo Clinic in Jacksonville, Fla. (Stroke 2005;36:400-3).

They said that the data presented to the Food and Drug Administration for the device’s approval did not meet the standards that could be achieved in a randomized trial.

"These data certainly provide no reassurance about the safety of mechanical embolectomy, yet alone the efficacy," the authors said. "The risks of treatment with the Concentric clot retriever might not be limited to arterial perforation, dissection, or distal embolization; there could be unidentified risks associated with the required immediate poststroke angiography as well as with endovascular instrumentation and injury to the arterial endothelium. Unfortunately, these questions cannot be addressed with the data presented ... by the MERCI investigators. Randomized trials have a distinct advantage in this setting; they can control for variables we know are likely to relate to procedure success and outcome, and they can control for important variables we have yet to identify."

Concentric Medical, the manufacturer of the MERCI device, sponsored the MERCI registry. Dr. Rymer reported being on the speakers bureau of the company. The five other coauthors reported varying relationships with Concentric, including being consultants or medical advisors, and having an ownership interest in Concentric.

LOS ANGELES - Results of the MERCI patient registry appear to uphold findings from the device’s two pivotal trials of patients with acute ischemic stroke, which demonstrated that successful recanalization is significantly associated with good outcome.

But audience members who spoke at the International Stroke Conference following the presentation of the registry results emphasized – to the applause of others in the audience – that the MERCI device lacks randomized data proving its safety and efficacy, and that reported outcomes for patients have not been stratified according to intubation status.

The registry is a large, nonrandomized case series documenting the postapproval use of the devices, whereas the two pivotal trials (MERCI and Multi MERCI) had strict inclusion and exclusion criteria and protocols but did not compare the device vs. medical therapy. Device-treated patients in the trials were instead compared with a placebo group from a randomized medical trial, called PROACT II (Prolyse in Acute Cerebral Thromboembolism).

At the conference, Dr. Marilyn Rymer presented the 90-day outcomes of 1,000 patients with acute ischemic stroke in the registry who were treated with the MERCI clot retriever embolectomy device. The MERCI (Mechanical Embolus Removal in Cerebral Ischemia) and Multi MERCI studies examined embolectomy with the device in similar patient groups, with a total of 305 patients.

"The registry is designed to answer the question, ‘What does the real-world, unrestrained treatment of ischemic stroke with this device look like?’" said Dr. Rymer, a medical director of the Brain and Stroke Institute at St. Luke’s Hospital in Kansas City, Mo. However, she noted that because the registry consists of nonrandomized cases, the efficacy implied in it "can’t be compared to a medical therapy."

The participating sites included every consecutive patient who was treated with the device. Treatment remained a clinical decision guided by each site’s general practice. The inclusion criteria were a diagnosis of acute ischemic stroke and at least one pass with the tool.

The primary end point was revascularization with a TICI (Thrombolysis in Cerebral Infarction) scale grade of 2a or higher; there was also a secondary functional outcome, which was the modified Rankin Scale (mRS) score at 90 days. The final analysis included 872 patients as a result of excluding 4 with insufficient procedural data and 90 who were disabled before their stroke with an mRS of 2 or more, as well as losing 34 to follow-up.

The patients’ median age was 68 years, compared with the median age of 72 in the trials. There was wide intersite variability with regard to age: At one site, the median age of patients treated was 58, and at another the median age of patients treated was 72.

"We saw the same kind of variation in terms of baseline National Institutes of Health Stroke Scale [NIHSS] score," Dr. Rymer said at the conference, which was sponsored by the American Heart Association. "The median in the registry was 17, while it was 21 in the trials." Median NIHSS scores also varied across the registry sites (range, 14-21), she said.

Overall, 305 patients received intravenous thrombolytic therapy, "But there was an incredible variation among sites" in the use of thrombolytics in conjunction with embolectomy (range, 0%-71%). Intra-arterial thrombolytic therapy also varied widely: Some 47% of patients overall received it, but the intersite rate varied from 7% to 100%.

Overall, 63% of patients were intubated, with the rate varying from 12% to 100%. "Several sites used intubation routinely in 100% of their patients, and some intubated only for airway protection. This becomes important as we begin to understand that intubation is associated with a worse outcome," Dr. Rymer said.

Most sites treated fewer than 19% of patients with angioplasty or stenting in addition to clot retrieval, but one site employed these additional treatments in 64% of patients.

The time from symptom onset to groin insertion was 6.3 hours, compared with 4.5 hours in the MERCI trials. Most patients (71%) were treated 3-8 hours after symptom onset.

Overall in the registry, 80% of patients were successfully recanalized, which was significantly more than in the two MERCI trials combined (65%). Similar numbers of patients had good 90-day outcomes (32% in both the registry and combined trials). Mortality at 90 days also was not significantly different between the registry and the trials (33% and 38%).

Symptomatic brain hemorrhage in the registry was 7% overall, not significantly different from the 8.8% seen in the MERCI trials, Dr. Rymer said. "It is notable that in the patients who were well recanalized (those with a TICI grade of 2b to 3), the symptomatic hemorrhage rate was lower (3.7% in TICI 2b and 5.4% in TICI 3)."

When the investigators examined the rate of good 90-day outcome (defined as an mRS of 0-2), they found the best outcomes in patients with the lowest baseline NIHSS scores. "As the stroke became more severe, the likelihood of good outcome went down," she said. In cases with NIHSS scores lower than 16, "the outcomes were excellent," she said, with up to 70% of those with a TICI grade of 2b or 3 experiencing a good outcome. TICI 2a provided only modest benefit, but it was consistent across the whole range of NIHSS scores, she added.

Age and recanalization status also affected mortality. "Age was a predictor of worse outcome, but recanalization did provide benefit across all ages except for the very young, who had low mortality rates in any case," she said.

A multivariate analysis identified several factors that significantly affected mortality both negatively and positively, including advancing age (odds ratio, 1.05), worse baseline NIHSS score (OR, 1.08), revascularization to a TICI grade of 2a, 2b, or 3 (OR, 0.33), heart failure (OR, 2.85), blood glucose above 140 mg/dL (OR 2.0), and intubation during the revascularization procedure (OR, 2.20)

The same multivariate model also identified factors that negatively impacted good 90-day outcomes, including worse baseline NIHSS score (OR, 0.88), advancing age (OR, 0.96), intubation during the procedure (OR, 0.43), longer duration of procedure (OR, 0.66), and a blood glucose level of 140 mg/dL or greater (OR, 0.59).

During the discussion period, several audience members questioned the relationship between intubation and poor outcomes. Dr. Joseph Broderick, chair of the department of neurology at the University of Cincinnati, said that the intubation data were interesting but could throw a statistical kink into the risk analysis. "How do you take into account those centers with a 100% intubation rate?" he asked. "Why don’t you analyze the data separately and see if the risk of intubation still holds up? There is always a risk of selection bias unless you compare the [sites] that always intubated against those that did not. Otherwise, you might be including people who were intubated because they looked dead, had heart failure, or weren’t breathing."

Three other speakers echoed the same concern, with one suggesting that intubation could lead to aspiration pneumonia, which in turn could contribute to a bad outcome.

"We don’t know what all the facts are" in relation to intubation, Dr. Rymer said. "We can only speculate."

However, the discussants also emphasized that there are still no randomized data confirming the device’s efficacy when compared with medical therapies. "There is, however, a clear association of improved outcomes with better revascularization," Dr. Rymer said in an interview.

Both MERCI and Multi MERCI were criticized for comparing their device-treated populations with only a placebo arm of PROACT II, according to an editorial by Dr. Kyra J. Becker of the University of Washington and Dr. Thomas G. Brott of the Mayo Clinic in Jacksonville, Fla. (Stroke 2005;36:400-3).

They said that the data presented to the Food and Drug Administration for the device’s approval did not meet the standards that could be achieved in a randomized trial.

"These data certainly provide no reassurance about the safety of mechanical embolectomy, yet alone the efficacy," the authors said. "The risks of treatment with the Concentric clot retriever might not be limited to arterial perforation, dissection, or distal embolization; there could be unidentified risks associated with the required immediate poststroke angiography as well as with endovascular instrumentation and injury to the arterial endothelium. Unfortunately, these questions cannot be addressed with the data presented ... by the MERCI investigators. Randomized trials have a distinct advantage in this setting; they can control for variables we know are likely to relate to procedure success and outcome, and they can control for important variables we have yet to identify."

Concentric Medical, the manufacturer of the MERCI device, sponsored the MERCI registry. Dr. Rymer reported being on the speakers bureau of the company. The five other coauthors reported varying relationships with Concentric, including being consultants or medical advisors, and having an ownership interest in Concentric.