Selective cerebral cooling for acute ischemic stroke

The most potent treatment for acute ischemic stroke is recanalization using intravenous thrombolytic therapy, intra-arterial thrombectomy (IAT), or both. Since a minority of patients respond to recanalization with full cure, a need exists for adjunctive cytoprotective treatments to complement recanalization and augment good outcomes. After decades of work and hundreds of trials, however, no effective therapy has proven effective as an adjunct to recanalization. Among many reasons for the failure of prior cytoprotective trials is the fact that evidence for recanalization was not required or documented. For example, the highly promising cytoprotective drug NXY-059, while effective in multiple pre-clinical stroke models—including non-human primates (NHPs)—failed in a pivotal Phase 3 clinical trial known as SAINT2. ¹ Yet, whether patients enrolled in SAINT2 recanalized remains unknown because no post-thrombolytic vascular imaging occurred.

Therapeutic hypothermia remains the single most effect adjunctive treatment for acute ischemic stroke in pre-clinical animal models.^2,3 Two recent multi-center trials of therapeutic hypothermia for ischemic stroke were halted due to poor recruitment, but trends suggested that hypothermia may have promoted pneumonia.^4,5 In patients with massive stroke, hypothermia appeared to reduce edema^6,7 but failed to benefit similar patients when combined with hemicraniectomy. ⁸ Given the incredible potential for hypothermia to improve long-term outcomes after ischemia, and the mixed record in clinical trials, there remains a great need to learn how best to combine recanalization with therapeutic hypothermia.

The advent of widespread thrombectomy for stroke follows publication of several recent clinical trials.^9–13 Importantly, multi-modality imaging allows the clinician to select patients for thrombolysis and IAT, even beyond the traditional time windows.^14–16 In other words, using imaging selection to identify patients with salvageable brain, we have opened the door to treating a much larger number of patients, even those who present late after stroke onset. The availability of IAT allows investigators the tremendous opportunity to study cytoprotective drugs in ideal patients: those with successful recanalization. Only two trials so far have studied the effect of cytoprotection when combined with IAT. The first trial to allow IAT as part of the clinical trial protocol was the RHAPSODY trial.^17,18 Recently, a larger trial allowed IAT to be used in study patients, and in a crucial innovation, required evidence of good collateral flow as an inclusion criterion. ¹⁹ In subgroup analysis, there was a trend to suggest that patients with complete or partial recanalization benefited from the study drug, confirming the idea that recanalization powerfully permits a putative cytoprotectant to influence patient outcome.

In this issue of the Journal, Wu and colleagues present the results of a study in NHPs examining the role of recanalization in permitting the efficacy of therapeutic hypothermia. ²⁰ The model uses injected autologous blood clots to occlude the middle cerebral arteries of NHPs followed by intra-arterially administered tissue plasminogen activator (tPA) to attempt recanalization. After 2.5 h, based on follow-up angiography, the vessels could be categorized as showing full, partial, or no recanalization. Then, each subject was randomized to receive a single intra-arterial injection of cold saline (100 ml lactated Ringer’s solution at 0 to 4°C) over 20 min into the middle cerebral artery. Among animals with full or partial recanalization, the intra-arterial cooling added considerable benefit, measured as reduced stroke size and improved behavioral scores. Impressive benefits were seen at 24 h and 30 days. The same group completed a pilot feasibility study in stroke patients, with promising safety data. ²¹

The principal value of this trial is the demonstration that recanalization powerfully influences the protective effect of therapeutic hypothermia. Although the absence of vascular imaging in prior human trials was always suspected of powerfully impeding success, the paper by Ji and colleagues impressively confirms that hypothesis. Other significant advantages of the study are the large sample size and reasonable attention to rigor. Using NHPs rather than rodents allows the use of human-style catheters; the NHP brain resembles human in size and complexity; and the volume-to-surface area ratio—a key driver of successful hypothermia—is closer to human in an NHP. Intra-arterial cold saline infusion would translate easily into clinical practice, as thrombectomy use spreads widely around the world.

The authors infused tPA into the occluded cerebral arteries, while the gold standard in humans for IAT includes the use of thrombectomy. The authors wisely avoided thrombectomy in this trial, knowing that intra-arterial tPA would fail to open 100% of the arteries. Thus, they could simulate the human condition in which some arteries open and some do not; this allowed the critical evaluation of the interaction between cooling and recanalization status. It could be noted that animals were not randomly assigned among the recanalization categories, but such would not be feasible, and the design used by the authors makes tremendous sense. Another powerful innovation in this study includes the use of serial magnetic resonance scanning—as would be done in many human stroke patients—to allow tracking of the stroke volume. Selective intra-arterial cooling powerfully limited the growth and expansion of ischemia into infarction.

Further studies in NHPs will be important and informative. One of the critical dilemmas we face when applying therapeutic hypothermia to humans is that we do not know the optimal timing: when to start cooling; how long to cool; the optimal target temperature depth; or how long to re-warm. Recently, it has been suggested that the duration and depth of therapeutic hypothermia should be personalized to each patient, perhaps based on the time from stroke onset to recanalization, or perhaps based on biomarkers. ²² Another critical question concerns the best method for applying hypothermia. Selective intra-arterial cooling, as used in the present study, holds tremendous promise, as the ischemic brain is cooled quickly and with minimal impact on the rest of the body.^21,23–26 This seems like the ideal way to induce hypothermia. On the other hand, if cooling should continue for any length of time, then systemic cooling will be needed, and it should also be very quick. New, more powerful methods are clearly needed to conduct human cooling as prior studies have failed to achieve target temperature in a reasonable time window. Further, in the myocardial infarction literature, there is strong data suggesting the body should be at target prior to coronary recanalization. ²⁷ Evidence of this phenomenon is lacking in stroke, but if it proves true, then selective intra-arterial cooling will be the only effective approach.

Regardless of future findings, the study of selective intra-arterial cooling by Ji et al. serves as a landmark result showing the crucial and powerful impact of recanalization on the evaluation of putative cytoprotective therapies. Future studies in humans and NHPs must carefully and rigorously address this and assure that subjects are stratified according to recanalization grade. As the recent landmark ENACT-1 showed, perhaps imaging after IAT should be mandated in clinical trials of putative cytoprotective drugs, and successful recanalization required as an inclusion criterion. ¹⁹ Although logistically challenging, such innovations ought to allow us to finally identify effective adjunctive cytoprotective therapies to accompany recanalization.