Download this article's Factoid (PDF & PPT for Gold Subscribers) 

In patients undergoing endovascular therapy for acute ischemic stroke, the presence of collateral blood flow increases the likelihood of reperfusion and good functional outcome, according to a subanalysis of the DEFUSE-2 study published online February 25, 2014, ahead of print in Stroke. However, the study also suggests that good flow may be less important when reperfusion is achieved in patients with salvageable brain tissue.

In the original trial, published in The Lancet: Neurology in October 2012, 60 patients scheduled for endovascular treatment of acute ischemic stroke within 12 hours of symptom onset were assessed for an MRI profile, termed ‘target mismatch,’ which suggests the presence of viable tissue. Early MRI reperfusion was associated with good functional outcome at 90 days solely in the target mismatch group.

For the current study, investigators led by Michael P. Marks, MD, of Stanford University Medical Center (Stanford, CA), looked at the relationship between collateral status and angiographic reperfusion, subsequent infarct growth, and clinical outcome.

Collateral Flow Linked to Lesion Size

Using a 5-point scale based on baseline angiography, collateral flow was classified as either poor (0-2) or good (3-4). Collateral flow correlated with baseline National Institutes of Health Stroke Scale (NIHSS) score (P = 0.002) and median lesion volume (P = 0.009).

Patients with good collaterals were more likely to have higher reperfusion levels—gauged by Thrombolysis in Cerebral Infarction (TICI) score ranging from 0 to 3—than those with poor collaterals (P = 0.010; table 1).

Table 1. Reperfusion Rates According to Collateral Score

On day 5, patients in the target mismatch group with poor collateral scores but good reperfusion had less infarct growth on imaging than those with poor reperfusion (P = 0.009). Patients with good collateral scores and good reperfusion also had less infarct growth compared with those with poor reperfusion, although the difference did not reach statistical significance (P = 0.25). Moreover, there was a strong trend toward greater infarct growth in patients with poor collaterals/poor reperfusion vs those with good collaterals/poor reperfusion (P = 0.06). However, among those with good reperfusion, there was no difference in the amount of infarct growth between those with poor vs good collaterals. (P = 0.73).

A similar pattern was seen for mismatch patients in terms of 90-day functional outcomes. Patients with poor collaterals/good reperfusion fared better than those who did not reperfuse well (OR 12.0; 95% CI 1.6-98; P = 0.017). However, patients with good collaterals/good reperfusion only trended toward better outcomes compared with poor reperfusion (OR 4.7; 95% CI 0.8-26; P = 0.11). 

Evaluation of mismatch patients showed that those with good reperfusion had higher rates of hemorrhagic conversion if they had poor rather than good collateral flow, but the difference did not reach statistical significance (P = 0.407).

On regression analysis, there was no interaction between occlusion location and any of the variables including collateral and reperfusion status. 

Good Collaterals Incentive to Be More Aggressive 

In a telephone interview with TCTMD, Ashfaq Shuaib, MD, of the University of Alberta (Edmonton, Canada), noted that the current findings differ somewhat from the recent IMS III substudy insofar as they suggest that collateral flow is not as critical in mismatch patients with salvageable tissue. “I strongly believe that the presence of good collaterals is the best marker for outcome in thrombolytic or endovascular treatment,” he said. “However, [the current authors] are saying it may not be that straightforward.”

Nonetheless, Dr. Shuaib said that for him the clinical message is “if a CT angiogram shows good collaterals, you have extra incentive to be more aggressive because these patients are more likely to benefit from … therapy. [On the other hand,] if patients have poor collaterals, you would try to be a little more cautious.” In part, that is because patients with poor collaterals are more likely to experience hemorrhage, he added.

The key to collaterals’ beneficial role is providing compensatory blood to preserve ischemic but viable tissue, Dr. Shuaib explained. The presence of good collaterals means the therapeutic window may be extended up to 10 to 12 hours, he added.

However, collaterals vary considerably from person to person and likely over time, Dr. Shuaib noted. Flow is affected by factors such as genetics, age, and drugs. One promising area of research, he added, is therapy aimed at enhancing collateral flow.

Neglect of Collaterals Detrimental? 

David S. Liebeskind, MD, of the UCLA Stroke Center (Los Angeles, CA), said the dependence of the impact of a cerebral lesion on collateral status has been shown repeatedly in previous research, including the IMS III substudy for which he was the principal investigator. Nonetheless, he said, the current finding is an important reminder for a field overly focused on reperfusion.

Currently “endovascular therapy is at an impasse,” Dr. Liebeskind observed in a telephone interview with TCTMD. “We’ve become very good at recanalizing with the new devices, but to date in clinical trials, recanalization and reperfusion do not correlate with clinical outcomes. If you reperfuse a patient who has poor collaterals, you can still get a bad outcome.”

Dr. Liebeskind attributed clinicians’ neglect of the role of collaterals to “a certain all-or-nothing” mind-set. “Once we decide to go forward with endovascular therapy, it is extremely rare that we stop short of trying for the highest TICI [reperfusion] score we can possibly get. I think that paradigm needs to be reconsidered. If you assume that there is equipoise in the field, then you should not be going full tilt toward the most extensive revascularization you can achieve.”

Importantly, he said, if future trials do not account for the effect of collaterals, “the result may be failure to develop evidence that endovascular therapy has a role in the management of acute stroke patients,” and the availability of newer, more effective technologies may simply exacerbate the problem.

Study Details 

Target mismatch profile was defined as a ratio between hypoperfused tissue and ischemic core of ≥ 1.8, with an absolute difference ≥ 15 mL. Patients with the profile also had ischemic core volume ≤ 70 mL and tissue volume with more severe hypoperfusion ≤ 100 mL. 

Patients with occlusion of the internal carotid artery were more likely to have poor than good collaterals (48% vs 7%), while those with middle cerebral artery occlusions were more likely to have good vs poor collaterals (93% vs 52%; both P < 0.001).

Related Stories:

• Collateral Flow Associated with Angiographic, Clinical Outcomes in Acute Stroke Patients
• Imaging Shows Mixed Results in Patient Selection for Endovascular Stroke Therapy