ARTICLE IN BRIEF
Researchers have identified clinical and imaging phenotypes, triggers and risk factors for the re-emergence of deficits after stroke.
Post-stroke recrudescence (PSR), the transient worsening of deficits after they have initially resolved, is an under-recognized phenomenon that is associated with several identifiable triggers, clinical features, and health and demographic risk factors, according to study published in the August 7 online edition of JAMA Neurology.
PSR is believed to occur in the setting of toxic metabolic factors, such as infections and certain drugs. Still, no studies have attempted to identify the triggers, risk factors, and clinical phenotype of PSR.
In a crossover cohort and case-control study of about 150 patients, researchers found that PSR can be triggered by conditions like hyponatremia, infections, and benzodiazepine use, and is more prevalent in women and in African-Americans. Patients who smoke, have diabetes, and those with severe stroke or stroke from small-vessel disease have a higher risk for PSR.
“Post-stroke recrudescence is anecdotal in the medical literature — there are only a couple of case reports — but we see this phenomenon not uncommonly,” said Aneesh B. Singhal, MD, FAAN, vice chair of the department of neurology at Massachusetts General Hospital, in an interview with Neurology Today. “We thought it would be a good idea to characterize it and make physicians more aware of it.”
STUDY DESIGN, FINDINGS
For the study, researchers at Massachusetts General Hospital used the hospital's Research Patient Data Repository to identify patients — who were admitted between January 1,2000 and November 30, 2015 — with a primary or secondary diagnosis of cerebrovascular disease, who underwent magnetic resonance imaging (MRI) at least once, and who had the term “recrudescence” in their inpatient or outpatient clinical notes.
After reviewing 1,700 records, the researchers identified a total of 164 episodes of PSR, defined by such diagnostic criteria as worsening of residual post-stroke focal neurologic deficits; chronic stroke on brain imaging; no acute lesion on diffusion-weighted imaging (DWI); cerebral ischemia considered unlikely (for example, symptom duration of less than one hour without a new DWI lesion); and no clinical suspicion or electroencephalographic evidence of seizure around the time of the event.
The patient records revealed that PSR occurred after both ischemic and hemorrhagic stroke, after an average of 3.9 years, and the underlying strokes were of varying sizes but usually affected white-matter tracts. Triggers associated with PSR included infection, hypotension, hyponatremia, insomnia or stress, and benzodiazepine use. New deficits tended to have a sudden onset, were mild to moderate in severity, and resolved within hours or days after resolution of the trigger.
In addition, they found that the group with PSR included more women (62.1 percent), African-Americans (17.2 percent), and patients who self-identified as “other” race/ethnicity (13.1 percent). Patients were more likely to develop PSR if they had diabetes, dyslipidemia, were smokers, had infarcts from small-vessel disease, or had worse NIH Stroke Scale (NIHSS) scores at the time of the stroke.
The findings have important implications for physicians, Dr. Singhal said. “At discharge, the physician should start counseling the stroke patient that there is a risk, maybe 10 percent, that they will be readmitted with a subset of their initial stroke symptoms in certain settings, such as insomnia, benzodiazepine use, infections, and sodium imbalances,” he said.
Dr. Singhal added: “We should teach primary care physicians to avoid benzodiazepine exposure, be more careful about sodium – for example, avoid drugs that cause hyponatremia – and to treat infections much more aggressively. If a patient re-presents with a worsening of their original symptoms, they should consider the diagnosis of recrudescence rather than transient ischemic attack (TIA) or a new stroke, and their workup would include urinalysis, pneumonia, and cultures.” He noted that erroneous diagnosis of PSR as TIA, which is frequent, can lead to medication changes that may confer additional risk without added benefit.
The proper diagnosis of PSR offers benefits for doctors, patients, and caregivers, Dr. Singhal said. “There will be less anxiety for the patient and their family that there's been a stroke. Physicians can provide better counseling about the future. And you can avoid extensive, unnecessary testing. It may be reasonable to still do a brain MRI if you're unsure it's a stroke, but if the scan is negative and the patient meets criteria for PSR, it would not lead down garden path to vascular imaging, cardiac ultrasound, Holter monitoring, blood testing, etc. – all of which lead to increased expense and prolong the inpatient stay.”
“I think this is an important study and especially worthy of notice by non-stroke neurologists,” said Jonathan Edlow, MD, professor of emergency medicine at Harvard Medical School and vice chair of emergency medicine at Beth Israel Deaconess Medical Center. “They took something relatively common — with the caveat they only found about 150 patients over 15 years, but then again, they had strict criteria — and not well known outside the world of stroke neurology, and studied it very carefully. This is clearly the best article on this subject.”
But, he added, “I wouldn't go so far as to call it practice-changing. Every patient with a stroke in their first several hours of care will get a urinalysis and get their sodium checked, etc. All of these toxic, metabolic, and infectious processes will be diagnosed within hours or days. But the study findings will remind non-neurologists that this phenomenon does exist and to think about it. And in future studies of stroke mimics, recrudescence should be one of the stroke mimics, since this is just one flavor of stroke mimic.”
“This study basically provides confirmatory evidence for what people are doing already,” said George Wittenberg, MD, associate professor of neurology at the University of Maryland School of Medicine and staff physician at the VA Maryland Healthcare System. “There's some value in that. I think that if the clinical suspicion is still there for new stroke, then the physician would still be treating the patient as if they had a stroke. The risk factors they describe for PSR — diabetes, smoking — are already risk factors for stroke. They don't provide a predictive model you'd use to plug in numbers and say, this person has 1 percent chance of having a stroke.”
“They were able to confirm stroke or no stroke with MRI,” he continued, “but a lot of patients with PSR might be diagnosed clinically without MRI — or maybe with just computed tomography to rule out hemorrhage. Also, [regarding] changes in NIHSS, some patients may not have a baseline for that. They had a certain amount of information many people wouldn't have in clinical situations.” On the other hand, he added, “this may be very useful in the acute rehabilitation setting, in which the issue of new stroke or PSR comes up frequently, and in which baseline information is more available.”
“The positive [aspect of these findings] is that we may be moving toward a predictive model like the ABCD2 score that could be used quickly in the emergency department to calculate the likelihood of recrudescence,” Dr. Wittenberg said. “If you had that predetermined number, combined with clinical judgments, you could better make that decision: do a full stroke workup, or pursue other factors?”
“From the perspective of an emergency physician, who has to make a ‘go-no go’ call to pursue a ‘code stroke’ within 5-10 minutes of a patient's arrival, knowing this information does not help me very much,” Dr. Edlow said. “The findings of this paper should not dissuade front-line physicians who are evaluating a patient with the rapid or abrupt onset of focal neurological symptoms from pursuing stroke.”
“A point of debate has always been, do you want to go fast and miss some things, or go slow and be more accurate with a stroke diagnosis?” Dr. Edlow said. “One might take a message from this and say, maybe be just a little slower – we need to have a sodium level, or a urinalysis or a chest X-ray, etc. But this would come at a cost – losing the ability to treat some patients who are actually having an acute stroke. To conclude that the acute stroke evaluation must include all the triggers for recrudescence (with the important exception of a point of care glucose measurement), which you're not going to have available within the first minutes of a case, would lead to missed opportunities for treatment in some patients.”
EXPERTS: TRIGGERS AND PHENOTYPES FOR POST-STROKE RECRUDESCENCE
LINK UP FOR MORE INFORMATION:
© 2017 American Academy of Neurology
•. Topcuoglu MA, Saka E, Silverman SB, et al Recrudescence of deficits after stroke: Clinical and imaging phenotype, triggers, and risk factors http://jamanetwork.com/journals/jamaneurology/fullarticle/2646625. JAMA Neurol
2017; Epub 2017 Aug 7.
Neurology Today Quick Links