International Stroke Conference
February 6–8, 2019

Robert W. Regenhardt, MD, PhD

Moderators: Jean Claude Baron, Andrew Demchuk

The first talk, by Nerses Sanossian, was titled “Neuro-protection in the Pre-hospital Setting.” Sanossian began by introducing the ischemic cascade in which there are rapid changes over minutes to hours (O2 depletion, energy failure, terminal depolarization, ion homeostasis failure), secondary changes over minutes to days (excitotoxicity, SD-like depolarizations, disturbance of ion homeostasis), and delayed changes over days to weeks (inflammation, apoptosis). While no neuro-protective agent has proven successful in phase 3 clinical trials, he asserts that these agents still hold promise. Reconsideration of mechanistic targets is important as there has been a shift in treatment paradigm with thrombectomy now the priority for LVO strokes. As these patients are collected from locations of symptom onset, transported to thrombectomy capable hospitals, and admitted for acute care, when is the best time to offer a neuro-protective agent? Future trials must balance the timing against the cost. An agent could be administered pre-hospital, post-arrival/pre-imaging, post-tPA, pre/during transfer if needed, pre-thrombectomy, or post-thrombectomy. There are unique considerations for each timepoint. In the pre-hospital setting, consent can be difficult, there is no imaging available, agents should be safe in ICH, have no interaction with tPA, easily stored in ambulances, administration should be easy (avoid pumps and compounding), agents should be effective despite fewer patients recanalizing, and have robust experimental data at early time points. In-hospital initiation allows standard consent, imaging is available (could tailor agent to stroke subtype), pharmacies can store and dispense agents, agents can be combined with recanalization (improved delivery to target tissue and opportunity for targeting reperfusion injury and hemorrhagic transformation), and patients can be more carefully selected after imaging for those most likely to benefit.

Indeed, a long criticism of many animal models was that they utilized ischemia followed by reperfusion; in the thrombectomy era, now 80% of LVO patients may achieve reperfusion making the prior model more translatable. Several pre-hospital trials have paved the way for future work, including FAST-MAG testing magnesium (completed 2015), RIGHT 2 testing glyceryl trinitrate (completed 2019), FRONTIER testing NA1 (recruiting until 2020), and PHAST-TSC testing trans-sodium crocetinate (approved to start 2019). FAST-MAG had an average symptom onset to agent administration time of 45 min. The agent to tPA time was 92 min and the agent to thrombectomy time was 230 min, illustrating how much faster these pre-hospital agents can be administered. One of the latest agents, trans-sodium crocetinate (PHAST-TSC), is intended to work by enhancing diffusion of oxygen to hypoxic tissues by altering the structure of water molecules in plasma. Preclinical data suggests there may be a benefit for both ischemic and hemorrhagic stroke, making this an exciting pre-hospital target.

Michael Hill gave the second talk, titled “Expanding the Time Window for Thrombectomy: Neuro-protection in the Era of Endovascular Treatment.” Hill began by reminding us that despite thrombectomy’s enormous efficacy for LVO strokes, 37% of those treated are still disabled and 10% still die. The problem is not solved. One approach may be expanding the time window by administering an agent, perhaps pre-hospital, that can prevent infarct expansion and allow more patients to be candidates for thrombectomy. While time is a convenient “surrogate for physiology,” there is significant variation between individual patients and it is only one variable in the equation for infarct progression. Indeed, imaging characteristics may be more important than time as shown in the DAWN and DEFUSE 3 trials for thrombectomy and the WAKEUP and EXTEND trials for tPA.

Hill asked, “Can we help patients on the way to thrombectomy?” He believes we can as about 1/3 have infarct progression while being transferred for thrombectomy. He favors the approach of in-field delivery of an agent to “suspend” strokes. “Freezing ischemic penumbra evolution” has been shown in a rodent model using NA1. Hill also argued that in the era or thrombectomy our human patients are now better modeled by preclinical rodent monofilament ischemia and reperfusion experiments since humans are now achieving reperfusion. Much of the preclinical work with this model is now more translatable. Hill stated that the ESCAPE- NA1 trial is testing the efficacy of NA1 specifically in patients undergoing thrombectomy (recruiting until 2020). It is recruiting at 50 sites and has over 800 already enrolled. He told the audience to look for the results next year.

“Preventing Reperfusion Injury: Neuro-protection after Acute Stroke Treatment” by Lauren Sansing was the third talk. Sansing outlined reasons for the renewed interest in neuro-protection. Like the previous speaker, she made the point that while thrombectomy is highly effective, there is still room for improvement as only 10% of those treated achieve full recovery with mRS 0. Previous trials of neuro-protective agents may have failed as there were low rates of recanalization. In the thrombectomy era, drug is more likely to be delivered to target tissue after recanalization. In the past, there was less regulation on preclinical design. More recently, both the NIH and most journals have increased the level of rigor expected. There has also been increased interactions between basic scientists and clinicians through efforts of NINDS, Stroke Net, and academic groups.

Sansing believes that inflammation plays a key central role in many underlying mechanisms of ischemic injury. She described that the inflammatory response after stroke can be described as a balance of injurious effects and reparative effects. The injurious effects involve IL-1β, TNF, IL-6, endothelial cell activation, platelet aggregation, recruitment of pro-inflammatory neutrophils, monocytes, T cells, and inflammatory microglia. These processes lead to infarct growth, neuronal loss, and astrogliosis. In contrast, the reparative effects involve TGFβ, IL-10, IL-4, BDNF, alternatively activated microglia and macrophages, and regulatory T and B cells. These processes lead to clearance of debris, resolution of inflammation, angiogenesis, and neurogenesis.

Sansing further discussed thrombo-inflammation and the no-reflow phenomenon, in which there is infarct growth into the penumbra, failure of collateral flow, and tissue infarction despite recanalization of the parent vessel. She described that inflammation may be the underling mechanism, including NFΚB, MMP-9, PAI-1, PAR4, leukocyte margination, and neutrophil-platelet aggregates. Inflammation is also involved in several biologic comorbidities, such as aging, diabetes, and hyperglycemia.

Sansing then outlined several promising immunomodulatory targets in the translational pipeline. IL-6 receptor antagonism is one example; a meta-analysis of 24 studies showed elevated IL-6 is associated with poor outcomes. Tocilizumab (anti-IL-6R) is FDA approved for RA and GCA and has shown promising results in preliminary studies. Another is PAR1 functional selectivity with 3K3A-APC. Multiple preclinical studies in aged mice of both sexes with comorbidities such as hypertension have demonstrated effects. The RHAPSODY trial, a phase 2A dose escalation of 3K3A-APC in conjunction with tPA and/or thrombectomy, showed possibly lower rates of hemorrhagic transformation. Sphingosine 1 phosphate signaling with the use of fingolimod, which is FDA approved for MS, is another target. It modulates several cell types but is best known for inhibiting egress of memory T cells from lymph nodes. A meta-analysis of many preclinical models and early preclinical trials are promising. The FAMTAIS trial is ongoing. Lastly, IL-1 receptor antagonism has been explored in over 25 papers and 76 experiments. The phase 2 trial, SCIL-STROKE, enrolled 80 patients to examine Anakinra. While there was no effect on 90-day outcomes, a mediation analysis with baseline IL-6 showed promise and raised new questions.

Angel Chamorro gave the final talk, titled “Vasculo-protection for Stroke.” He opened by stating “Failure is not the opposite of success. It is part of success.” He summarized major lessons learned in the field, including the importance of studying the whole neurovascular unit and not just neurons. He stated the “ischemic cascade is a mess,” underscoring its complicated, inter-related, and multidirectional mechanisms. Chamorro argued that perhaps the best target is reducing oxidative stress. Recanalization is important and full reperfusion is better than partial reperfusion.

Chamorro believes that the vessels themselves are the next exciting target for vasculo-protection. Endothelial cells produce free radicals, which can lead to tight junction opening with blood brain barrier (BBB) degradation and pericyte contraction with no reflow. This may underlie mechanisms of futile recanalization. Indeed, it appears that neuro-protective agents can act directly on the vessels as agents to suppress ROS are still effective in reducing infarct size even when unable to cross the BBB.

Chamorro then discussed uric acid, the end product of nucleotide metabolism, in depth. By reducing nitrosylation of proteins, it is one of the strongest antioxidants, according to Chamorro. In preclinical models, uric acid has met all the STAIR criteria. Furthermore, it has been studied in both sexes, in hypertensive rats, in hyperglycemic mice, and there is tPA synergism. In human stroke patients, Chamorro has shown that uric acid levels are associated with good functional outcomes at hospital discharge. The URICO-ICTUS phase 2b/3 trial showed a trend for improvement given soon after tPA. Uric acid appears to have the most pronounced effects in patients with hyperglycemia, perhaps by blunting toxic effects of glucose. Chamorro believes this may be particularly important as we now know intensive glucose control is not enough to influence outcomes (SHINE). He showed data that glucose is a treatment modifier of thrombectomy utilizing HERMES data. He concluded by stating that uric acid is ready for definitive validation in a large phase 3 trial, but funding is lacking.

The ensuing discussion involved several interesting questions, but one of the most widely debated was, How should we decide which agent to test next in humans? The NINDS recently launched a model to test up to 6 neuroprotectants in parallel. Perhaps combinations of agents will be more effective. Agents should only be considered if all STAIR criteria are met. Others questioned if we should re-test agents that previously failed in the pre-thrombectomy era given that the human condition of LVO and thrombectomy now includes reperfusion. Drug companies may lack incentive to re-test agents that are past their patents. Furthermore, it may be that patient selection is key. I believe that for any agent, we should learn lessons from the thrombectomy trials; we should select patients that are most likely to benefit from a given agent as to not dilute their effects. Time and target are critical. Considerations of stroke subtypes and white matter vs gray matter injury mechanisms will also likely be important. I agree with the speakers and appreciate their optimism. Neuro-protective strategies have a role in the future, and funding their study is essential.