Pattern and frequency of recurrent transient ischemic attacks

Pattern and Frequency of Recurrent Transient Ischemic Attacks Elizabeth T. Walz, MD, Tia Brink, RN, and Andrew Slivka, MD

Objectives. Natural history studies and stroke prevention trials in patients with transient ischemic attacks (TIAs) have focused primarily on stroke or death outcomes and not recurrent TIA, yet treatment decisions are often based on recurrences. This study was undertaken to evaluate the frequency and pattern of recurrent TIAs in patients presenting with their first TIA. Methods. We monitored TIA recurrence for an average of 16 months in 47 consecutive patients hospitalized in a tertiary care center within 3 weeks of their first TIA. Treatment in 22 patients was aspirin or ticlopidine alone, and 12 underwent endarterectomy. Remaining patients received warfarin, dipyridamole, or pentoxifylline alone or with aspirin, including two who underwent endarterectomy. Treatment was changed in only four patients during the follow-up period. Results. Thirty-four patients (72%) had more than one TIA; 23 of these (68%) had two to five recurrences and 11 (32%) had more than five TIAs. Nineteen of the 34 patients with multiple TIAs had recurrences within 1 month of the initial TIA. Only four patients (12%) had recurrent TIAs throughout the follow-up period. Stroke occurred in five patients (11%); one patient had a single prior TIA, and the rest had multiple TIAs. No risk factors or treatments predicted which patients had single or multiple TIAs. Conclusions. Most patients have few recurrent TIAs, and recurrences usually occur within the first month of the initial TIA. The patterns of TIA recurrence identified should aid in treating patient and in devising appropriate outcome measures in treatment trials in which TLA is an endpoint. Key Words: Transient ischemic attack--Recurrence.

Natural history studies and stroke prevention trials in patients with transient ischemic attacks (TIAs) have focused primarily on stroke or death outcomes. 1-12Because recurrent TIAs may prompt treatment alterations, information about the pattern of recurrent TIAs is necessary to evaluate treatment efficacy in individual patients as well as treatment trials using recurrent TIA as an outcome. This study was designed to determine the frequency and pattern of recurrent TIAs in patients presenting with the first TIA. Potential predictors of repeated TIAs including treatment were also examined. Methods At The Ohio State University, data on all hospitalized patients with TIA or ischemic stroke are entered into the From the Department of Neurology, Ohio State University, Columbus, OH. Received April 23, 1996;accepted September 4, 1996. Address reprint requests to Elizabeth Walz, MD, 457 Means Hall, 1654 Upham Dr, Columbus, OH 43210. Copyright © 1997by National Stroke Association 1052-3057/97/0603-0005503.00/0

cerebrovascular database. Patients are classified according to the National Institute of Neurological Diseases and Stroke Classification of Cerebrovascular Diseases IIU 3 A TIA is defined as focal neurological symptoms, thought to be caused by ischemia, of less than 24 hours' duration, and ischernic stroke as symptoms persisting beyond 24 hours. Patients with isolated vertigo, diptopia, syncope, or unexplained unconsciousness are excluded, as are patients with focal symptoms related to other potential causes (eg, brain tumor, migraine, or seizure). Only patients with the first TIA within 3 weeks of presentation were included because of concerns that longer intervals between the first TIA and evaluation might impair the historical accuracy of timing and number of TIAs. Patients were included in the study if they have had multiple TIAs before hospitalization. Patients were recruited over a 2-year period from June 1990 to June 1992. Follow-up was completed for all patients by December 1992. Patients were interviewed by a neurologist trained in cerebrovascular disease and a cerebrovascular nurse clinician. Data obtained included history of hypertension, diabetes, hypercholesterolemia, cardiac disease

Journal of Stroke and Cerebrovascular Diseases, Vol. 6, No. 3, 1997: pp 121-124

121

WALZ, BRINK, A N D SLIVKA

122 Table 1. Patient characteristics

1 TIA

No 2-5 TIAs >5 TIAs follow-up Total

Number I3(27%) 23(47%) 11(22%) Median age (yr) 66 67 70 Sex (% male) 54% 78% 45% Hypertension 4 8 4 Diabetes 2 2 I Cholesterol 0 3 1 Tobacco use 5 9 2 Cardiac 1 8 2 Prior stroke 4 4 1

2(4%) 49 56 100% 1 I 0 1 0 0

65 65% 17 6 4 17 11 7

including possible sources of emboli, and tobacco use. The vascular distribution and duration of TIAs as well as the treatment before the qualifying event and subsequent therapies were also recorded. Patients were then followed up by phone interview, conducted by the nurse clinician, at 1, 3, 6, 9, 12, 16, 20, and 24 months after entry into the study and were seen at least once during the follow-up period by the neurologist. The date, description, and duration of recurrent TIAs and any changes in treatment were recorded, as was the occurrence of stroke or death. The time between the initial TIA and subsequent TIA(s) and the number of recurrent events were determined. Patients with one TIA were compared with those with more than one TIA by Fisher's Exact Test to determine whether demographic features or treatment predicted further TIAs.

Results Forty-nine patients were identified as having the first TIA within 3 weeks of evaluation. Patient characteristics are presented in Table 1. Two patients were lost to follow-up, both with two TIAs during their 2-day hospitalization. Neither was followed up after the hospitalization because of lost contact. For the remaining 47 patients, average follow-up was 16 months. There are no significant differences between patients with a single TIA and those with multiple events (P > .05, Fisher's Exact Test). Of the 47 patients followed up, 34 (72%) had more than one TIA during the follow-up period; 25 of these patients had their first recurrence before hospitalization. Twentythree patients (68%) had a total of two to five recurrent TIAs, and the rest (32%) had more than five TIAs, ranging from five to 140 events. Patients with more than one TIA had three patterns of recurrence (Fig. 1). Nineteen of the 34 patients (56%) had recurrent episode(s) within I month of the initial event and had no further events afterwards. Eleven patients (32%) had recurrent TIAs in the first year

occurring within 2 months (four patients), 4 months (two patients), 7 months (three patients), and 9 to 12 months (two patients) of the initial event; 10 patients had a single recurrence, and the other had two TIAs at 9 months and then no further events during the follow-up period. The other four patients (12%) had multiple TIAs throughout the follow-up period at irregular intervals. One of these patients had only three further TIAs, one at 16 months and two at 24 months. Another of these patients was followed up for 2 months then died of renal and liver failure. In 27 patients (55%), the duration of the initial TIA and subsequent TLAs was less than 30 minutes. Ten patients (20%) had symptoms that lasted more than 2 hours. In all but five patients the duration of subsequent TIAs was the same as that of the presenting event. In only two patients were recurrences in another vascular distribution (one in the opposite carotid, the other vertebrobasilar). The etiology of the TIA did not account for the pattern of recurrence because in all three groups the majority had the event(s) secondary to atherosclerotic disease. Patients received a variety of treatments at the discretion of the primary medical physician or the cerebrovascular neurologist. Twenty-one patients were treated with aspirin alone, and t2 patients underwent carotid endarterectomy, with nine receiving aspirin postoperatively. Five patients were treated with warfarin alone, and the remaining patients were treated with various combinations of dipyridamole, pentoxifylline, aspirin, and warfarin. After recurrences, warfarin was added to aspirin in two patients, treatment changed from aspirin to warfarin in one, and treatment was changed to ticlopidine in another. The remaining patients continued with the therapy that was initiated at the time of entry into the study. Initial treatment with aspirin or warfarin as a single agent or in combination with other agents did not predict recurrent

18 16 14

~'.~

~

2-5 events

~

> 5 events

i/

O~

12

E

o

n ~8

t0 8 6 4

7, i, 7, 7, ii 7, 9;, 7, i/ zl

7,

within I

mo

2 mo

4 mo

7 mo

9-10 mo

Time to TIA Recurrence

Figure 1. Patternof Recurrent TIAs.

Irregutor

RECURRENT TRANSIENT 1SCHEMIC ATTACKS

events compared with patients with single events (P > .05, Fisher's Exact Test). Nine patients had a history of previous cerebral infarction. After entry, six patients (12%) developed subsequent infarctions. One of 13 (8%) patients with a single TIA had an infarction after carotid endarterectomy, and five of 34 patients (15%) with recurrent TIAs had a stroke, occurring 6 to 24 months after the initial TIA (P > .05, Fisher's Exact Test). Three patients (7%) died during the study period, one of multiorgan failure, one of cardiac failure, and one of unknown cause. Discussion This study describes the frequency and pattern of TIA recurrences, which had been studied previously in a limited manner. More than two thirds of patients had more than one TIA; the majority had two to five TIAs and then no further recurrences. We could not identify any demographic or treatment variables that predicted TIA recurrence. Three patterns of recurrence were identified. Most patients had recurrent TIAs within 1 month of the initial attack and then no further events. Thirty-two percent had a recurrence 1 to 12 months after the initial TIA, with almost all having a single recurrence, and only 12% had recurrences at irregular intervals throughout the follow-up period. We confirmed that most TIAs last less than 2 hours, and the frequency of TIAs does not appear to predict stroke risk. 1"4,8q°Our 1-year stroke rate of 12% is in the range reported by others) ,3,9,1° This study was performed at a center with primarily tertiary care, and only hospitalized patients were included; therefore, this may not reflect the pattern of TIAs in the community. The majority of patients had the first recurrence before hospitalization, possibly prompting the referral and admission. This finding suggests that the number of recurrences would be greater than in the community in general. In addition, we studied a small number of patients; these trends need to be confirmed in a larger series. Data on TIA recurrence is limited and confounded in most cases by treatment. In series of patients with TIA, the incidence of multiple TIAs ranges from 19% to 77%, although in most of these studies, as in ours, patients were treated with antiplatelet agents, anticoagulants, or carotid endarterectomy. 1,4~s,~4q8Most recurrences are in the same vascular distribution as the initial event, which we confirmed in our study. In a study of 144 patients with TIA followed up for 3 years, the cumulative number of attacks increased for the first 18 months and more attacks occurred in patients with vertebrobasilar than carotid distribution events. 7 The attack rate leveled off after 18 months for patients with carotid TIAs but continued to increase for those with vertebrobasilar TIAs. Several studies have noted that patients with multiple TIAs before entry were more likely to have recurrent TIAs than those with a

123 single presentation) ,2,9Our data show that although most patients have recurrent TIAs, the TIAs are few in number and tend to occur soon after the presenting event. Two TIA series have included patients who were not treated. Marshall reported 14 of 60 patients (23%) with a single TIA and 46 of 60 patients (77%) with recurrent TIAs after an average follow-up of 2 years, s Of the 46 patients with recurrent TIAs, the length of time the patients experienced TIAs was less than 1 month in 30%, 1 to 3 months in 9%, and longer than I year in 59%, although the pattern of TIAs during this time was not described. Acheson and Hutchinson reported 81 patients with TIA followed up for approximately 3 years, of whom only six were treated. 8 Fourteen of these patients (17%) had one TIA a day or more for 2 weeks, and 32 patients (40%) had regular TIAs over 1 to 4 weeks. The other 35 patients (43%) had recurrent TIAs at irregular intervals of days, weeks, or months. Our results were similar, despite treatment, in that 56% of patients with multiple TIAs had recurrences within 1 month and no further events. The remaining 44% had recurrent TIAs at irregular intervals, but most (30%) had only one recurrence 1 to 12 months after the initial event. In our study, patients received a variety of treatments, including aspirin, warfarin, dipryidamole, pentoxifylline, and carotid endarterectomy alone or in combination. Therapy was changed after recurrent events in only three patients. Treatment with aspirin or warfarin did not affect the frequency or pattern of recurrence. Other medications and endarterectomy were not used in enough patients to determine whether they would have affected TIA frequency. Patients with recurrent TIAs often have alterations in treatment regimens. Few treatment trials have examined TtAs as an outcome, although aspirin, ticlopidine, pentoxifylline, and carotid endarterectomy have all been reported to alter TIA frequency. 19"2sCarotid endarterectomy, pentoxifylline, and tidopidine decrease the number of patients who experience recurrent TIAs.21-23Aspirin decreases the instance of patients with more than two TIAs within a week) 9 Aspirin has also been reported to decrease the instance of patients with more TIAs in the 6 months after the beginning of treatment than in the 3 months before treatment compared with placebo-treated patients. 24 Knowing the frequency and patterns of recurrent TIAs would help in evaluation of treatment efficacy for TIAs. Our study shows that most patients with multiple TIAs have recurrences within the first month after the initial event and no subsequent attacks, irrespective of treatment. Using a single TIA recurrence as a treatment failure in trials without also recording the total number of TIAs therefore appears to be of limited value. In conclusion, this study describes the frequency and pattern of TIA recurrence, which has not previously been

124 well described. This study shows that patients tend to have relatively few TIAs, and they cluster soon after the initial ischemic event. Patients infrequently have TIAs over a prolonged period. Furthermore, this study confirmed results of prior studies concerning the duration and distribution of TIAs as well as the risk of stroke. This information can be helpful in patient treatment decisions about when to add another antithrombotic agent or when to discontinue newly added agents. These results also suggest that if TIA is to be used as an endpoint in treatment trials, a single recurrence may be insufficient to evaluate treatment efficacy.

Acknowledgment The authors thank John Kissel, MD, for his comments and suggestions in preparation of the article.

References 1. Baker RN, Ramseyer JC, Schwartz WS. Prognosis in patients with transient cerebral ischemic attacks. Neurology 1968;18:1157-65. 2. Simonsen N, Christiansen HD, Heltberg A, Marquardsen J, Pedersen HE, Sorensen PS. Long-term prognosis after transient ischemic attacks. Acta Neurol Scand 1981;63: 156-68. 3. Sorensen PS, Marquardsen J, Pedersen H, Heltberg A, Munck O. Long-term prognosis and quality of life after reversible cerebral ischemic attacks. Acta Neuro} Scand 1989;79:204-13. 4. Friedman GD, Wilson WS, Mosier JM, Colandrea MA, Nichaman MZ. Transient ischemic attacks in a community. JAMA 1969;210:1428-34. 5. Marshall J. The natural history of transient ischaemic cerebro-vascular attacks. Q J Med 1964;131:309-23. 6. Ostfeld AM, Shekelle RB, Klawans HL. Transient ischemic attacks and risk of stroke in an elderly poor population. Stroke 1973;4:980-6. 7. Ziegler DK, Hassanein RS. Prognosis in patients with transient ischemic attacks. Stroke 1973;4:666-73. 8. Acheson J, Hutchinsin EC. Observations on the natural history of transient cerebral ischaemia. Lancet 1964;2: 871-4. 9. Dennis M, Bamford J, Sandercock P, Warlow C. Prognosis of transient ischemic attacks in the Oxfordshire Community Stroke Project. Stroke 1990;21:848-53.

WALZ, BRINK, AND SLIVKA

10. Cartlidge NE, Whisnant JP, Elveback LR. Carotid and vertebral-basilar transient ischemic attacks. Mayo Clin Proc 1977;52:117-21. 11. Hankey GJ, Slattery JM, Warlow CP. The prognosis of hospital-referred transient ischaemic attacks. J Neurol Neurosurg Psychiatry 1991;54:793-802. 12. Toole JF, Yuson CP, Janeway R, et al. Transient ischemic attacks: a prospective study of 225 patients. Neurology 1978;20:746-53. 13. Special report from the National institute of Neurologic Disorders and Stroke. Stroke 1009;21:637-76. 14. Levy DE. How transient are transient ischemic attacks. Neurology 1988;38:674-7. 15. Werdelin L, Juhler M. The course of transient ischemic attacks. Neurology 1988;38:677-80. 16. Cunneally PM, Dyken ML, Futty DE, et al. Cooperative study of hospital frequency and character of transient ischemic attacks. VIII. Risk factors. JAMA 1978;240: 724-46. 17. Fields WS, Lemak NA. Joint study of extracranial arterial occlusion. IX. Transient ischemic attacks in the carotid territory. JAMA 1976;235:2608-10. 18. Fields WS, Maslenikov V, Meyer JS, Hass WK, Remington RD, MacDonald M. Joint study of extracranial arterial occlusion. V. Progress report of prognosis following surgery or nonsurgical treatment for transient cerebral ischemic attacks and cervical carotid artery lesions. JAMA 1970;211:1993-2003. 19. The SALT Collaborative Group. Swedish aspirin lowdose trial of 75 mg aspirin as secondary prophylaxis after cerebrovascular ischaemic events. Lancet 1991;338: 1345-49. 20. Ticlopidine Aspirin Stroke Study Group. A randomized trial comparing ticlopidine hydrochloride with aspirin for the prevention of stroke in high-risk patients. N Engl J Med 1989;321:501-7. 21. Veterans Affairs Cooperative Study Group. Efficacy of carotid endarterectomy for asymptomatic carotid stenosis. N Engl J Med 1993;328:221-7. 22. The Ticlopidine Aspirin Stroke Study Group. Efficacy of ticlopidine and aspirin for prevention of reversible cerebrovascular ischemic events. Stroke 1993;24:1452-57. 23. Herskovits H, Famulari A, Tamaroff L, et al. Preventive treatment of cerebral transient ischemia: comparative randomized trial of pentoxifylline versus conventional antiaggregants. Eur Neurol 1985;24:73-81. 24. Field WS, Lemak NA, Frankowski RF, Hardy RJ. Controlled trial of aspirin in cerebral ischemia. Stroke 1977;8: 301-14. 25. Barnett HJM. Drug and surgical issues in stroke prevention. Cerebrovasc Dis 1994;4(suppt 1):16-25.