TOTAL: Use of Routine Thrombectomy in Primary PCI Takes Another Blow

SAN DIEGO, CA—Performing manual thrombectomy routinely prior to primary PCI does not improve 180-day outcomes among patients with STEMI over PCI alone, according to results from the TOTAL trial. In addition, the risk of stroke was slightly higher following thrombectomy.

The findings, which were presented March 16, 2015, at the American College of Cardiology/i2 Scientific Session in San Diego, CA, and simultaneously published online in the New England Journal of Medicine, are consistent with those from the TASTE trial.

“Both TOTAL and TASTE emphasize the need to conduct large randomized trials to test common interventions even when small trials appear promising,” said Sanjit S. Jolly, MD, of McMaster University (Hamilton, Canada).

From August 2010 to August 2014, Dr. Jolly and colleagues randomized 10,732 patients with STEMI who were referred for primary PCI within 12 hours of symptom onset to routine manual thrombectomy with Export catheters (Medtronic) before PCI or PCI alone at 87 centers in 20 countries. The primary analysis included only the 10,063 patients (mean age 61 years; 77% men) who underwent primary PCI after randomization.

Baseline characteristics were generally similar in the 2 groups, although those in the thrombectomy group were less likely to be smokers and had longer times to hospital presentation after symptom onset (128 vs 120 minutes; P = .02). Overall, 78.4% of patients had a high thrombus burden (TIMI thrombus grade 4 or 5).

The rate of crossover was 4.6% from thrombectomy to PCI alone and 1.4% from PCI alone to thrombectomy. In addition, bailout thrombectomy was used in 7.1% of the control group. Patients assigned to routine thrombectomy were less likely to receive upfront or bailout glycoprotein IIb/IIIa inhibitors (37.4% vs 41.4%) and more likely to undergo direct stenting (38.3% vs 21.3%; P < .001 for both). Procedure time was longer in the thrombectomy group (median 39 vs 35 minutes; P < .001).

No Impact on Clinical Outcomes

Following PCI, thrombectomy was associated with lower rates of incomplete ST-segment resolution of less than 70% (27.0% vs 30.2%) and distal embolization (1.6% vs 3.0%; P < .001 for both).

Those effects did not, however, translate into improvements in clinical outcomes. The rate of cardiovascular death, recurrent MI, cardiogenic shock, or NYHA class IV heart failure through the first 180 days (primary outcome) was 6.9% with thrombectomy and 7.0% with PCI alone (HR 0.99; 95% CI 0.85-1.15). The finding was consistent across subgroups, including in patients with large thrombus burden.

Similarly, there were no differences between groups in the rates of cardiovascular death, a composite of the primary outcome plus stent thrombosis or TVR, components of the primary and secondary outcomes, or major bleeding (table 1).

Stroke was more frequent in the thrombectomy arm both at 30 days (0.7% vs 0.3%; HR 2.06; 95% CI 1.13-3.75) and 180 days (1.0% vs 0.5%; HR 2.08; 95% CI 1.29-3.35), although an assessment of net clinical benefit incorporating the primary outcome and stroke at 180 days revealed no difference between groups (7.5% vs 7.2%; HR 1.04; 95% CI 0.90-1.20).

Other analyses—full intention-to-treat, as-treated, and per-protocol—provided results consistent with the primary.

Larger Trials Take Down Smaller Studies

Evidence from prior studies regarding the utility of manual thrombectomy performed before primary PCI has been mixed. The 1,071-patient TAPAS trial showed that myocardial blush grade (primary outcome) was improved and 1-year cardiac mortality was reduced, resulting in a guideline recommendation for routine thrombectomy, which was followed by a substantial increase in its use.

Consistent with the TOTAL results, however, subsequent trials—including INFUSE-AMI and TASTE—failed to demonstrate clinical benefits of thrombectomy. TOTAL differed from those other 2 trials in that it was the largest, it “used detailed event definitions, and the outcomes were ajudicated in a blinded fashion by an independent central committee rather than relying on registry data,” Dr. Jolly and colleagues note.

The TOTAL results, which failed to show an association between beneficial changes in ST-segment resolution and distal embolization and clinical outcomes, “[caution] against changing practice on the basis of trials showing an improvement in surrogate outcomes,” they write.

They acknowledge some limitations of the trial, however, including the fact that interventionalists performing the procedures were not blinded to treatment assignment, the inclusion of some patients with low thrombus burden, and the inability to rule out a benefit of selective—rather than routine—use of thrombectomy.

Dr. Jolly said it is important to note that bailout thrombectomy was needed in some patients in the PCI-alone group. “Clinical judgment is still needed, especially if patients are deteriorating and [have] a large thrombus burden and your current strategies are not working,” he said.

Findings Suggestive, but Not Definitive

In an editorial accompanying the NEJM paper, Filippo Crea, MD, of the Catholic University of the Sacred Heart (Rome, Italy), says the main strengths of TOTAL were the study design and large sample size.

But he points out that the stroke finding is difficult to interpret and likely due to chance—a possibility also raised by the study authors—for several reasons. First, the study was not powered to detect a difference in stroke and the number of events was small. And second, Dr. Crea writes, “the increase in stroke risk with thrombectomy was not confined to the periprocedural period, as would be expected if the procedure were causal ….”

Dr. Jolly said the increase in stroke was unexpected and requires confirmation in future studies, adding that the trial investigators are currently conducting a detailed analysis and a case-by-case review with neurologists to help understand the causes.

Although the findings of TASTE and TOTAL together suggest putting an end to the routine use of manual thrombectomy, neither trial excludes the possibility of a benefit in high-risk patients, he says, noting that event rates in both were lower than expected. He adds that a recent meta-analysis incorporating TASTE supports a potential impact in high-risk patients by showing lower rates of MACE, late death, reinfarction, and stent thrombosis with thrombectomy.

At a press briefing, panelist David E. Kandzari, MD, of the Piedmont Heart Institute (Atlanta, GA), said the TOTAL results will likely reduce but not eliminate use of manual thrombectomy because it has been shown to be useful in facilitating procedures and reducing some periprocedural complications, despite the lack of an effect on clinical outcomes.

However, that is assuming that no safety concerns arise, he said, adding that even though the number of strokes was low, “any difference in stroke would be very clinically meaningful.”  

---

Sources:
1. Jolly SS, Cairns JA, Yusuf S, et al. Randomized trial of primary PCI with or without routine manual thrombectomy. N Engl J Med. 2015;Epub ahead of print.
2. Crea F. Coronary microvascular obstruction—a puzzle with many pieces [editorial]. N Engl J Med. 2015;Epub ahead of print.

Disclosures:

• The trial was funded by Medtronic, the Canadian Institutes of Health Research, and the Canadian Network and Center for Trials Internationally (CANNeCTIN).
• Dr. Jolly reports receiving grant support from Medtronic, the Canadian Institutes of Health Research, and CANNeCTIN during the conduct of the study, and personal fees from AstraZeneca, and St. Jude.
• Dr. Crea reports receiving grant support from Biotronik and Boehringer Ingelheim.
• Dr. Kandzari reports receiving consulting fees/honoraria from Boston Scientific, Medtronic, Micell Technologies, and Thoratec.

Related Stories:

• TASTE: No Appetite for Manual Thrombus Aspiration
• Meta-analysis: Manual, Not Mechanical, Thrombectomy Holds Advantages Over PCI
• INFUSE-AMI: Bolus Abciximab Works, Thrombus Aspiration Doesn’t for Large STEMI