Pouya Tahsili-Fahadan, MD
The introduction of endovascular thrombectomy (EVT) has revolutionized acute treatment of stroke due to a large vessel occlusion (LVO). However, patients who present after 6 hours from stroke onset (defined by previous EVT trials; 7.3 hours based on a recent meta-analysis) are not eligible for either IV thrombolysis or EVT. Salvageable brain tissue that can be saved by cerebral reperfusion may exist well beyond this treatment window.
The DAWN (“DWI or CTP Assessment with Clinical Mismatch in the Triage of Wake-Up and Late Presenting Strokes Undergoing Neurointervention with Trevo”) trial studied this very population of patients. In this multi-center, prospective, randomized, open-label trial, “EVT+standard medical care” (EVT group) was compared with “standard medical care alone” (control group) in treatment of patients who presented with acute ischemic stroke due to a large vessel occlusion (LVO) within 6-24 hours from their last known well (LKW) and showed evidence of salvageable brain tissue. Presence of salvageable brain tissue was based on a mismatch between severity of their neurological deficits and the volume of infarcted brain on presentation as follows:
Enrolled patients had good pre-stroke baseline defined as modified Rankin scale, mRS, 0 to 1. Infarct volume was assessed on diffusion-weighted imaging (DWI) sequence of brain MRI or perfusion CT scan using the automated RAPID software.
The DAWN trial measured “utility-weighted mRS” at 90 days as the primary outcome. In this disability score, the mRS was weighted based on the values obtained from both patient- and clinician-centered studies. As shown in the table below, there are important differences between “mRS” and “utility-weighted mRS” in assessment of morbidity:
– “Utility-weighted mRS” ranges from 0 (death) to 10 (no deficits). In contrast to “mRS,” a higher score indicates a better outcome.
– The weighted values highlight that both “mRS” of 5 and 6 are highly and equally undesirable outcomes (corresponding to “utility-weighted mRS” 0).
– While low “mRS” 0 and 1 are also considered as good outcomes in the “utility-weighted mRS” (10.0 and 9.1, respectively), the perceived level of disability on “utility-weighted mRS” steeply increases between “mRS” 2 to 4. The difference is most evident from “mRS” 3 to 4, corresponding to a large decline in the “utility-weighted mRS” from 6.5 to 3.3.
DAWN enrolled 206 patients: 107 in the EVT group and 99 in the control group. Baseline characteristics of the patients in both groups were overall comparable, except that the EVT group had a higher percentage of atrial fibrillation (40% vs. 24%) and stroke on awakening (63% vs. 47%) compared to the control group. Although most of the cases were out of the window for IV tPA, 5 patients who had persistent LVO 6-24 hours from LKW underwent EVT (versus 13 patients in the control group).
EVT methodology was fairly homogenous: Trevo stent-retriever was used in 102 (out of 105) patients, only 10% received general anesthesia, and carotid angioplasty was performed in only 3 cases. Reperfusion was achieved in 84% of patients by the median of 13.6 hours from LKW. Rate of recanalization at 24 hours after stroke onset was 77% and 39% for the EVT and control groups, respectively.
At 90 days, the “utility-weighted mRS” was 5.5 for the EVT group compared to 3.4 in the control group (absolute difference 2.1). Also, 49% of patients in the EVT group achieved functional independence (defined as mRS 0-2) vs. only 13% in the control group (see Figure 1 in the article). In other words, number-to-treat (NTT) after EVT for less disability and functional independence at 90 days was 2 and 2.8, respectively. Of note, rate of functional independence in the EVT group (49%) was similar to the pooled data from previous EVT trials (46%) performed within 6 hours from stroke onset. However, rate of functional independence in the control group of this study was significantly lower than previous EVT trials (13% vs. 26%). This difference between control groups can be partly attributed to high rates of IV thrombolysis among patients of the previous EVT trials and enrollment of more severe strokes in DAWN. Results of ongoing trials such as IWITNESS may extend the indication for IV thrombolysis beyond the current 4.5 hours guidelines.
Also, 48% of the patients in the EVT group showed early therapeutic response defined as a decrease in the NIHSS score of ≥10 from baseline or an NIHSS score of 0-1 at discharge. Median increase in the infarct volume at 24 hours was only 1 ml for the EVT group versus 13 ml among controls.
In addition, EVT+standard medical care was found superior to standard medical care only across pre-specified subgroups: treatment groups (A, B, and C), gender, age (<80 and ≥ 80), baseline NIHSS (10-17 and >17), occlusion site (ICA and MCA), type of stroke onset (wake-up, witnessed, and un-witnessed), interval between LKW and randomization (6-12 and 12-24 hours), and interval between symptoms onset and randomization (0-6 and >6 hours).
Serious adverse events, including mortality and stroke-related death at 90 days, as well as symptomatic intracerebral hemorrhage, were similar between the two groups. Rate of procedure-related complications was very low.
DAWN is an extremely important study with significant clinical implications for patients and practitioners. First, DAWN introduced a new aproach to outcome analysis by introduction of utility-weighted mRS, an outcome measure that further incorporates patients’ desires and clinicians’ view of favorable outcomes. I expect this approach to be more commonly used in the future trials. Second, a new way to assess salvageable brain tissue (penumbra) was utilized. In lieu of an imaging-based penumbra used in the previous studies, a “clinical-imaging mismatch” was determined. Most importantly, DAWN advanced treatment of stroke from a time-based to a tissue-based window.
Extrapolation of these findings to daily practice needs to be approached cautiously. The extension of the treatment window to 24 hours shall not cause any delays in treatment. “Time is [still] brain” and all cases of acute stroke have to be approached emergently. However, protocols to transfer patients to an EVT-capable center need to be revised thoroughly as more patients may now benefit from EVT. A small proportion of patients with LVO within 6-24 hours from stroke onset meet the clinical-imaging inclusion criteria used in this study, and it is yet to be determined whether patients with smaller clinical-imaging mismtach (milder symptoms, larger infarct volumes, and smaller salvagable tissue) can also benefit from the extended window for EVT. Infarct size was determined by RAPID software based on DWI or CTP images, and neither the software nor the opportunity for hyper-acute MRI or CTP are widely available in all stroke centers. Determination of infarct size based on other quantitative (volumetric) or qualititative (such as ASPECTS) assessment of non-contrast head CT needs to be studied.