Novel Definition of Stroke “Good Responders” Predicts 90-Day Outcome after Thrombolysis

Abstract

Introduction

Identifying predictors of good response in thrombolytic-treated stroke is important to clinical care, resource allocation, and research design. We developed a simple, novel measure of “Good Responders” to assess if 2 short-term variables could predict 90-day outcomes after thrombolysis in stroke.

Methods

Intravenous thrombolysis-treated stroke cases from June 2004 to June 2018 were analyzed from a stroke registry. Intraarterial treatment cases were excluded. Good responders (GR++) were defined as those with length of stay less than or equal to 3 days and discharge to home. Poor responders (GR- -) had length of stay more than 3 days and discharge other than home. Mixed responders (GR+/-) composed the remainder. Baseline characteristics and predictors of 90-day outcome were assessed.

Results

Of 261 patients, there were 101(38.7%) GR++, 67(25.7%) GR- -, and 93(35.6%) GR+/-. For GR++ versus GR- - versus GR+/-, there were differences in mean age (62.7, 71.2, 69.2; P = .0016), and baseline modified Rankin score (mRS) 0-2 (%: 94.9, 74.6, 84.8; P = .008). Younger age, male sex, lower values for systolic BP, glucose, and baseline mRS were associated with good responders. Older age, atrial fibrillation, symptomatic intracerebral hemorrhage, and baseline mRS greater than 2 were associated with poor responders. At 90 days, mortality was reduced in GR++ versus GR- - versus GR+/- (%alive: 92.6, 72, 86; P = .04), and mRS(0-2) (%: 36.8, 0, 11.8; P < .001).

Conclusions

Good responders to thrombolysis are younger and have better baseline functional status. Our novel definition of “Good Responders”, using 2 early variables of home disposition and short length of stay, may help predict 90-day post-thrombolytic outcome. Future work should focus on validating this definition.

Introduction

Stroke is a leading cause of mortality and disability worldwide,^1,2 and a major source of economic burden.³ An estimated $65 billion is lost to stroke in the United States, with two thirds accounted by direct costs (including acute and long term care), and the remainder third accounted by indirect costs that include lost productivity.⁴

The use of intravenous recombinant tissue plasminogen activator (rt-PA) since its approval in 1996, has become pivotal in reducing disability in stroke. About two thirds of eligible patients receive IV thrombolysis in the United States, with the rate nearly doubling over time.⁵ The time it takes to improve after rt-PA varies among individuals, but certain variables have been correlated with good outcomes. Variables correlating with early improvement are not well elucidated, but studies have found an association between early neurologic improvement within 24 hours and favorable outcome at 3 months.6, 7, 8 Even so, a proportion of patients still fail to demonstrate early recovery after thrombolysis.⁸ In these studies, the National Institute of Health Stroke Scale (NIHSS) was the sole measure of clinical improvement, plotted against differing time frames. While unquestionably useful, utilizing the NIHSS poses some limitations,⁹ cautious interpretation, and requires training and certification.

Identifying who will recover early from thrombolysis, and if early recovery will correlate with long term outcome, is still unknown. In order to eliminate NIHSS measurements and expand outcome measures to include patients that do not necessarily improve early within 24 hours, but improve nonetheless during hospitalization, we hypothesized that variables available early in the hospitalization of rt-PA patients may aide in longer term outcome prediction. We assessed if 2 clinically chosen variables (hospital length of stay [LOS] of 3 days or less, and discharge destination of home) could pragmatically predict 90-day outcome among rt-PA treated stroke patients.