Mild Hypothermia Reduces Tissue Plasminogen Activator–Related Hemorrhage and Blood Brain Barrier Disruption After Experimental Stroke: Editorial Commentary on Tang et al .,2013

Thrombolysis with recombinant tissue plasminogen activator (rt-PA)—still the only proven effective stroke treatment—causes significant hemorrhagic transformation in 3 to 6% of patients. Critical risk factors include age, pre-existing white matter disease, and hypertension (NINDS tPA Study Group, 1997). Each of these risk factors causes known changes that might increase susceptibility to hemorrhagic transformation, such as reduced levels of plasminogen activator inhibitor one (PAI-1) and microvascular thickening and medial fibrosis. Although the exact mechanism of thrombolysis-related hemorrhagic transformation remains unclear, we do hope to find a neuroprotectant that might diminish the risk, and hypothermia has long been touted for such a role (Lyden et al., 1993; Meden et al., 1994; Kallmunzer et al., 2012). Previous work suggests that hypothermia could influence the efficacy of thrombolytics (Yenari et al., 1995). Research in the laboratory of Drs. Yenari, Tang, and colleagues provides important evidence that hypothermia may indeed reduce hemorrhagic risk during thrombolysis (Tang et al., 2013).

In this issue of Therapeutic Hypothermia and Temperature Management, the authors report a study on mice using the traditional middle cerebral artery occlusion (MCAo) model with a nylon filament (Tang et al., 2013). The choice of this model will seem odd to clinicians, but the inert thread can be removed at a known time, allowing precise control of reperfusion. Rt-PA or hypothermia or both were then given to the mice in varying combinations. To mimic the common clinical scenario, rt-PA and hypothermia were given together, much as we try to do in the ongoing ICTuS 2/3 trial. In anticipation of hoped-for future trials of field-based hypothermia, the investigators also studied the effect of hypothermia prior to reperfusion and rt-PA. When used simultaneously, hypothermia was highly protective, reducing infarct size and clinical deficit scores. When hypothermia preceded reperfusion and rt-PA administration, infarct size was also reduced. The critical and notable finding in this article, however, is that hypothermia powerfully limited the blood brain barrier (BBB) disruption and the hemorrhagic conversion seen with rt-PA use. While long hoped for by everyone in this field, these data are among the first, and certainly the most rigorous, to demonstrate a beneficial effect of hypothermia on hemorrhagic transformation associated with rt-PA.

The use of the filament MCAo model in mice allowed the authors to perform a number of elegant measurements, but mice respond to rt-PA considerably differently than humans or even rats. For example, using Western blots on whole brain extracts, they showed a reduction of endogenous rt-PA with hypothermia but no change in PAI-1 levels. If this effect were confirmed in humans, it suggests that hypothermia might impede the efficacy of thrombolysis during hypothermia. We are testing this hypothesis by examining recanalization rates in the ICTuS 2/3 trial. Another significant issue in this work is the use of young animals free of any co-morbidities such as hypertension or vasculopathy. Thus, while very encouraging, the results reported by Tang et al. (2013) do require further confirmation in animal models—and of course in humans—that include such confounding disorders.

The fundamental significance of these data is—as many have hypothesized—hypothermia appears to exert a salutary effect on the BBB that reduces the risk of thrombolysis-related hemorrhagic transformation. In clinical use, we hope for this exact result. In addition, we certainly hope that hypothermia will benefit efficacy outcomes as well as reduce safety concerns.