A conversation with Philippa Lavallée, MD, Department of Neurology and Stroke Centre, Bichat University Hospital, about the importance of atypical symptoms in patients with TIA.

Interviewed by José G. Merino, MD, FAHA, Associate Professor of Neurology, University of Maryland School of Medicine.

They will be discussing the paper, “Clinical Significance of Isolated Atypical Transient Symptoms in a Cohort With Transient Ischemic Attack,” published in the June 2017 issue of Stroke.

Dr. Merino: Could you please briefly summarize the key findings and put them into context of what was known before you did the study?

Dr. Lavallée: Conventional wisdom considers that some transient symptoms such as diplopia, vertigo, dysarthria and even a sensory deficit limited to one limb or the face are not compatible with the diagnosis of TIA when they occur in isolation. Daily experience in the stroke unit and TIA clinic shows that it is not true. In our study, we enrolled 1,850 patients seen in our TIA clinic who had transient symptoms and found that 10% of the patients with stroke or TIA had one of these isolated atypical symptoms and that 10% of the patients with atypical symptoms had an acute infarct on brain MRI and 18% had an underlying disease that placed them at high risk of stroke recurrence.

Dr. Merino: As you mention, you enrolled patients referred to your TIA clinic, and these patients were screened by a senior vascular neurologist who confirmed the suspicion that they have a TIA. Do you think that your findings can be generalized to a general population or to patients who present to the emergency room or their doctor’s office with atypical transient symptoms?

Dr. Lavallée: Yes. In fact, the message of the study is particularly important for physicians who are not stroke specialists, such as GPs or emergency physicians, because they are generally the first to evaluate patients with these atypical symptoms. A diagnosis of TIA should be considered in patients with acute transient neurological deficits if the symptoms are compatible with a cerebral arterial territory, even when the deficits are atypical transient symptoms. Of course, they must also consider other more common alternative diagnoses like paroxysmal benign vertigo in patients presenting with isolated vertigo.

Dr. Merino: This study seems to have important implications for practice. How may your findings change the approach to patients with atypical transient symptoms seen in different settings?

Dr. Lavallée: Considering the high risk of stroke after TIA, it is important to consider TIA diagnosis in all patients with acute transient neurological deficit if the deficit could be explained by an arterial cerebral occlusion. When obvious alternative diagnoses are excluded by a detailed medical history, examination and ancillary tests, patients with atypical transient symptoms should be investigated by a stroke specialist.

Dr. Merino: Is there a difference in the incidence of acute DWI lesions between patients with atypical transient symptoms and those with a more “classical” presentation?

Dr. Lavallée: As I noted previously, in our series, 10% and 11.5% of patients with isolated atypical and typical symptoms (such as a motor deficit or aphasia), respectively, had an ischemic lesion on DWI. The difference was not statistically significant. The high incidence of DWI lesions in patients with isolated atypical symptoms was mainly driven by patients with dysarthria and partial sensory symptoms.

Dr. Merino: Do you think that your findings should lead to changes in clinical guidelines?

Dr. Lavallée: Definitely. Our study shows that it is important to enlarge the list of symptoms suggestive of TIA and include some atypical symptoms.

Dr. Merino: Did you collect all the data prospectively? What were the major challenges you faced when doing this study?

Dr. Lavallée: The data were collected prospectively by stroke specialists using a CRF. Our team is highly motivated, and the time taken to fill the CRF was short (less than 10 minutes). When we decided to open the TIA clinic, our first goal was proving the efficiency of emergent treatment of TIA patients by stroke specialists to decrease the risk of recurrent stroke. From a research perspective, we collected all the details about patient medical history, clinical symptoms, work-up result, vascular risk factors, treatment. Some were chosen on purpose as we had a prespecified hypothesis, but other ideas came after we closed the data collection. We were able to perform other analysis, as the CRF was very detailed. All stroke neurologists were asked to fill the CRF (on paper) at the time they saw the patient. I personally reviewed all the CRF, and I would strongly advise to test first the CRF before launching a study with a lot of data analyzed and to check with all your collaborators that each item is well understood. A research nurse anonymizes and enters the data in an eCRF. Regarding French regulation of database management, we obtained the authorization of the CNIL (an independent administrative organization). At that time, it was not mandatory to obtain informed consent from the patient.

Dr. Merino: You have a very efficient system to refer, triage and evaluate patients with a possible TIA, and I wonder if your model can be implemented in other settings. How is the clinic set up and staffed? How can you get the testing done in a timely fashion? 

Dr. Lavallée: The TIA clinic is an outpatient clinic attached to a stroke center. Between 7 a.m. and 5 p.m., patients are admitted in a day hospital that is part of the neurovascular unit. This day hospital was preexisting to the TIA clinic; as a result, we did not recruit any supplementary nurse or neurologist. After 5 p.m., patients are evaluated in a room located in the stroke unit by the stroke neurologist on call. The inclusion of the TIA clinic as part of the stroke unit is very important. It allows us to hospitalize TIA patients if necessary (e.g. those with high-grade arterial stenosis), and to use the diagnostic resources of the stroke unit (duplex ultrasound and transcranial Doppler, immediate access to MRI/CT, rapid access to the vascular department). It was also important to provide a direct access to first line physician (GP, emergency department) to reduce the delay of admission. We have created a dedicated phone line open at all times for physician referrals.  Patients are admitted after a phone interview of the physician to be sure that symptoms are compatible with TIA symptoms. This phone triage was effective, as only 20% of the patients were finally misdiagnosed. From our point of view, all stroke units and centers should have a TIA clinic.

Dr. Merino: Do you do outreach to physicians in the Paris region to teach them about the importance of rapid referral to your TIA clinic?

Dr. Lavallée: Before opening the TIA clinic, we sent to all the neurologists, GPs, cardiologists and ophthalmologists in the Paris region a leaflet providing important information about TIA (symptoms, risk of stroke, need for prompt work-up) and information on how to reach our TIA clinic. That represents about 10,000 leaflets to send. Since the opening of the clinic, the mailing process has been done 3 times. We implement the leaflet with the result of the data published in the Lancet Neurology in 2007 showing a 80% risk reduction of stroke. To date, 10,000 leaflets have been sent.

Dr. Merino: What advice can you give neurologists who want to set up a TIA clinic in their hospital or community?

Dr. Lavallée: The key points are to provide access for TIA patients at all times, staff the clinic with stroke specialists who may provide prompt consultations and have immediate access to relevant work-up studies including MRI and arterial imaging (MRA or CT angiography or Duplex US/transcranial Doppler), and have the ability to hospitalize patients if necessary (in our experience, it represents about 20% of the patients).

In our experience, the need for immediate access to cardiac echography is exceptional. Cardiac echography and Holter ECG are performed in the month following the event.