Incidence of Recurrent Subarachnoid Hemorrhage After Clipping for Ruptured Intracranial Aneurysms
Background and Purpose— Because intracranial aneurysms develop during life, patients with subarachnoid hemorrhage (SAH) and successfully occluded aneurysms are at risk for a recurrence. We studied the incidence of and risk factors for recurrent SAH in patients who regained independence after SAH and in whom all aneurysms were occluded by means of clipping.
Methods— From a cohort of patients with SAH admitted between 1985 and 2001, we included those patients who were discharged home or to a rehabilitation facility. We interviewed these patients about new episodes of SAH. We retrieved all medical records and radiographs in case of reported recurrences. If patients had died, we retrieved the cause of death. We analyzed the incidence of and risk factors for recurrent SAH by Kaplan-Meier curves and Cox regression analysis.
Results— Of 752 patients with 6016 follow-up years (mean follow up 8.0 years), 18 had a recurrence. In the first 10 years after the initial SAH, the cumulative incidence of recurrent SAH was 3.2% (95% confidence interval [CI], 1.5% to 4.9%) and the incidence rate 286 of 100 000 patient-years (95% CI, 160 to 472 per 100 000). Risk factors were smoking (hazard ratio [HR], 6.5; 95% CI, 1.7 to 24.0), age (HR, 0.5 per 10 years; 95% CI, 0.3 to 0.8) and multiple aneurysms at the time of the initial SAH (HR, 5.5; 95% CI, 2.2 to 14.1).
Conclusions— After SAH, the incidence of a recurrence within the first 10 years is 22 (12 to 38) times higher than expected in populations with comparable age and sex. Whether this increased risk justifies screening for recurrent aneurysms in patients with a history of SAH requires further study.
The prevalence of intracranial aneurysms under the age of 20 is very low and increases thereafter.1 This suggests that aneurysms are not congenital but develop during life. The development of aneurysms during life and the presence of multiple aneurysms in up to 30% of patients with subarachnoid hemorrhage (SAH) indicate that patients who have had an aneurysm are at risk for developing new aneurysms.2
In patients with SAH and clipping of all detected aneurysms, new aneurysms can develop at a new location (de novo) or at the clip site (regrowth).3–7 In addition, some aneurysms might have been missed at the angiogram at the time of the SAH. The International Study of Unruptured Intracranial Aneurysms (ISUIA) study showed that the rupture risk of small unruptured aneurysms in patients with a previous SAH from another aneurysm is higher than in patients with similar aneurysms but without a history of SAH.8 The ongoing development of aneurysms and higher rupture risks after a previous episode of SAH suggest that patients with a history of SAH are at considerable risk of a new SAH.
We performed a long-term follow-up study in a large cohort of patients who regained independence after SAH and in whom all detected aneurysms had been occluded by means of clipping. We determined the incidence of and the risk factors for a recurrent episode of SAH.
From a database of patients admitted to the University Medical Center Utrecht with SAH, we selected all patients admitted between 1985 and 2001 who met the following inclusion criteria: (1) subarachnoid hemorrhage confirmed by computed tomography (CT) or lumbar puncture; (2) presence of a saccular aneurysm confirmed by conventional angiography or CT-angiography; (3) clipping of the ruptured aneurysm and all additional aneurysms; (4) age at time of SAH ≥20 years; and (5) discharge to home or a rehabilitation facility. Patients who were discharged to a nursing home and patients with one or more aneurysms left untreated were not included.
After approval of the ethical committee of our hospital, we contacted the general practitioner of all eligible patients to know if the patient was still alive. If patients had died, we asked for the date and the cause of death. If a patient had died in a hospital or other facility, we reviewed the medical records. Subsequently, we sent a letter to all patients who were still alive. In this letter, patients 70 years of age or younger were invited to the outpatient clinic. For patients older than 70 years, we announced a telephone call. If a patient had no phone number or an exdirectory one, we sent another letter asking the patient to contact us. At the outpatient clinic or during the telephone interview, we asked the patients about new episodes of SAH. For all patients, we retrieved data on sex, age, family history of intracranial aneurysms, smoking habits, alcohol use, history of hypertension, and number of aneurysms at the time or the initial SAH.
Recurrent Subarachnoid Hemorrhage
For patients who reported new episodes of SAH, we retrieved all medical data from the hospital where patients had been treated and reviewed all CT scans of the brain and all angiograms of the intracranial vessels of the initial episode if available and of the recurrence. New episodes of SAH were defined as SAH proven by CT, lumbar puncture, or autopsy after treatment of all aneurysms that had been found at the time of the initial SAH. On basis of surgery reports, CT scans and angiograms, autopsy reports, or—if patients died before CT-angiography could be performed—by a pattern of hemorrhage on CT highly suggestive for the site of rupture, we categorized patients into: (1) aneurysm located at a site remote from the original clipped aneurysm, (2) aneurysm located at the same site as the original clipped aneurysm, or as (3) unclassified if patients died before CT-angiography could be performed and the site of rupture could not be derived from the hemorrhage pattern. The aneurysms at other sites than the clip were subdivided in (1) de novo (aneurysm not visible on the initial CT-angiogram at the time of the SAH); (2) additional (aneurysm visible in retrospect but not identified on the initial CT-angiogram); or (3) possible de novo (when the CT-angiogram at time of SAH was not available for review, but the aneurysm was not described in the radiology report). The aneurysms located at the clip were classified as (1) regrowth (postoperative angiogram showed complete clipping of the aneurysm); (2) remnant (postoperative angiogram showed incomplete clipping of the aneurysm); or (3) possible regrowth (when no postoperative angiogram was performed).
Patients with a history suggestive of aneurysmal rupture but without confirmation of the diagnosis because they had died before reaching hospital were classified as sudden death probably as a result of recurrent SAH. A history was classified as suggestive for aneurysmal rupture if we had an eyewitness account of a sudden severe headache before onset of coma. Patients with sudden loss of consciousness not proceeded by acute headache or patients who were found dead were not classified as suggestive for aneurysmal rupture.
The risk for recurrent SAH was assessed by survival analysis. We calculated the incidence rate per 100 000 patient-years and the cumulative incidence at 5, 10, and 15 years with corresponding 95% confidence intervals (CIs) for recurrent SAH and sudden death probable as a result of recurrent SAH combined and for recurrent SAH separately. If patients had died during the follow-up period, were lost to follow up, or were admitted to a nursing home, they were censored at that point of time.
We calculated incidence rate ratios with corresponding 95% CIs by dividing the incidence rate in our population by a population-based incidence rate of SAH. This population incidence rate was derived from a population in Australia and New Zealand with comparable age (45 to 54 years).9 Adjustments were made for differences in sex distribution.
We used Cox regression to calculate hazard ratios (HRs) and corresponding 95% CIs of baseline characteristics associated with the risk of recurrent SAH. Again, we performed the analyses for recurrent SAH and sudden death probable as a result of recurrent SAH combined and for recurrent SAH separately. The following factors were included in the analyses: age at time of SAH (categorical per 10 years), sex, smoking (dichotomous current vs former plus never smokers and categorical), family history of intracranial aneurysms (defined as one or more first-degree relative[s] with a verified aneurysm or a history suggestive of SAH, history of hypertension [dichotomous], and number of aneurysms at the time of SAH [dichotomous: one versus multiple]). We performed univariate Cox regression analysis for all risk factors and multivariate Cox regression analysis with forward selection of variables with probability values <0.20 in the univariate analysis.
Between 1985 and 2001, 930 patients survived after SAH and were successfully treated for the ruptured aneurysm. One hundred fifty-four did not meet the inclusion criteria (Figure 1). Of the 776 remaining patients, 24 (3%) could not be contacted. The total follow up of the 752 patients included in the study was 6016 years with a mean period of follow up of 8.0 years (range, 0.2 to 20.1 years). During the follow-up period, 107 patients had died. In 10 patients, the cause of death was a recurrent SAH. Seven patients died suddenly before reaching the hospital, of whom 2 had a history suggestive of SAH. Other causes of death were cancer (21); cardiovascular disease (19); infectious diseases (9); other causes, including dementia, trauma, renal failure (14); or unknown (27). In addition, during follow up, 23 patients were censored because they moved abroad (6) or were admitted to a nursing home (17). The baseline characteristics of the 752 patients are listed in Table 1.
Recurrent Subarachnoid Hemorrhage
A new episode of SAH had occurred in 18 (2.4%; 95% CI, 1.5% to 3.8%) of the 752 patients during the follow-up period. The incidence rate in our patients for the first 10 years after SAH was 286 per 100 000 (95% CI, 160 to 472 per 100 000). In the population-based study, the incidence rate derived from a population in Australia and New Zealand with a comparable age and sex distribution was 13 per 100 000. The risk of a subarachnoid hemorrhage in our population was therefore 22 (95% CI, 12 to 38) times higher compared with the population based incidence (Table 2).9
The cumulative incidence of new episodes in the first 10 years after the index SAH was 3.2% (95% CI, 1.5% to 4.9%) for all certain recurrences (Figure 2 and Table 2). When the 2 patients with sudden death probable as a result of SAH were included the cumulative incidence was 3.5% (95% CI, 1.8% to 5.3%).
In the 18 patients with a recurrent SAH after successful treatment, 19 recurrent aneurysms were found: 13 patients had an aneurysm located at a new location, 3 patients had an aneurysm located at the clip site from the previous operation, one patient had both an aneurysm located at a new location and an aneurysm located at the clip site, and one patient had an unclassified aneurysm. Of the 14 patients with aneurysms at a new location, the CTA or angiogram was available for review in 8 patients: 4 aneurysms were certain de novo aneurysms and 4 were classified as additional. The remaining 6 aneurysms were classified as possible de novo aneurysms. Of the 4 patients with an aneurysm at the clip site all were classified as possible regrowths.
The mean interval between the initial SAH and the recurrence was 6.5 years (range, 0.2 to 17 years). We found no recurrent SAH from a (possible) de novo or regrowth aneurysm within the first 33 months after operation.
Risk Factors for Recurrent Subarachnoid Hemorrhage
In the univariate Cox regression analysis for recurrent SAH, significant risk factors were age (HR, 0.6 per 10 years; 95% CI, 0.4 to 0.9), a verified history of familial SAH (HR, 3.8; 95% CI, 1.1 to 13.2), current smoking (HR, 4.8; 95% CI, 1.3 to 17.4), and the presence of multiple aneurysms at time of the SAH (HR, 5.7; 95% CI, 2.3 to 14.5). Gender (HR, 0.4 for men; 95% CI, 0.1 to 1.3) was not a statistically significant risk factor but had a probability value >0.2 and was therefore also included in the multivariate analysis. In the multivariate forward Cox regression analyses, current smoking (HR, 6.5; 95% CI, 1.7 to 24.0), age (HR, 0.5 per 10 years; 95% CI, 0.3 to 0.8), and the presence of multiple aneurysms at time of the SAH (HR, 5.5; 95% CI, 2.2 to 14.1) were statistically significant risk factors. In the Cox regression analysis for recurrent SAH, including sudden death probable resulting from SAH, the same risk factors were detected in both the univariate as the multivariate analysis with comparable hazard ratios in the multivariate analysis.
We found that in patients who had recovered to an independent state after an episode of aneurysmal SAH and in whom all detected aneurysms were treated by means of clipping, the risk of a recurrence the first 10 years after treatment is 22 times higher than the risk of SAH in a healthy cohort with comparable age and sex. Independent risk factors for subsequent episodes of SAH were current smoking, young age, and multiple aneurysms at the time of the initial SAH.
We collected >6000 patient-years of follow up with a mean follow-up time of 8 years. Two previous studies with much smaller numbers of patients and less stringent inclusion criteria than in our study have reported on the incidence of recurrent SAH.4,10 These studies found a slightly lower risk of a recurrent SAH in patients treated after SAH. The first study followed 220 clipped patients who survived >3 years over a mean period of 9.9 years (approximately 2100 patient-years of follow up).10 Six patients had a recurrent SAH and the cumulative incidence of recurrences calculated in that study was 2.2% at 10 years after the original treatment. The other study followed 102 surgically treated patients over a mean follow-up period of 4.4 years (443 patient-years of follow up).4 Only one recurrent SAH was reported and therefore the overall risk of recurrent SAH was 0.26% per year. We found that the risk of a recurrent SAH is small in the first few years after treatment and increases thereafter.11 The risk for recurrent SAH of 0.26% per year during a follow-up period of 4.4 years can therefore not be extrapolated to longer follow-up times. Moreover, more than one third of the patients included in this study did not have an SAH as initial presentation but had an incidentally detected aneurysm. The ISUIA study showed that the rupture risk of incidentally detected aneurysms in patients without a history of SAH is lower than in patients with additional aneurysms and a history of SAH. Furthermore, loss to follow up and selection bias occurred more often in this study than in our study. Only patients who underwent late follow-up angiography were included (<10% of the treated population), and therefore patients who died of a recurrent SAH before follow up could be performed were missed. Risk factors for recurrent SAH were not evaluated in the 2 studies.
We compared the incidence of recurrent SAH in our study with a large population-based study on the incidence of SAH in Australia and New Zealand.9 We chose this study as comparison because it is a recent study with over 1.7 million patient-years of follow up and data are reported for men and women separately in different age groups. The SAH incidence in this study was comparable with an incidence study performed between 1978 and 1980 in The Netherlands and with the incidence of SAH found in a metaanalysis of 15 studies from populations other than Finland and Japan.12,13 In Finland and Japan, the SAH incidences are higher than in other countries (around 20 per 100 000 patient-years).13–15 In our study, the incidence of recurrent SAH was 286 per 100 000. Thus, even if we compare our incidence of recurrences with relatively high incidences of SAH as in Finland or Japan, the difference between the incidence in the population and the incidence of recurrent SAH is impressive and is included within the limits of the confidence interval of our estimate (incidence rate ratio, 22; 95% CI, 12 to 38).
There are some limitations of our study that need to be acknowledged. Although the number of patients lost to follow up in our study was very small (3%), we did not have follow up for all patients. Furthermore, in 27 patients, the exact cause of death was unknown, and an additional 5 patients died suddenly of unknown cause. If some or all of these patients would have had a subarachnoid hemorrhage, the actual recurrence rate is higher than the one observed in our study. Conversely, if none of the patients who were lost to follow up had a recurrence, we have slightly overestimated the rate of recurrent SAH by not including the follow-up years of these patients.
In 4 patients, the recurrence was caused by an aneurysm at the clip site. Unfortunately, a postoperative angiogram was not performed in these patients. In some patients, the aneurysm might have been incompletely clipped. Even after surgery by experienced neurosurgeons, neck remnants are found in nearly 10% of the cases.16 We therefore cannot exclude that some patients had small remnants of the initial ruptured aneurysm.
We excluded patients who were discharged to a nursing home because consent for follow up in these patients is in general hard to obtain because most of these patients have severe cognitive deficits. Moreover, from a clinical point of view, the risk of recurrence is less relevant for such patients. Our results therefore apply only to patients discharged home or to a rehabilitation center, which will in general be a younger population than the patients discharged to a nursing home. Whether our results can be generalized to coiled patients is unclear. One can assume that the formation rate and rupture risk of de novo aneurysms in coiled patients is comparable with that in clipped patients. The number of regrowth aneurysms or remnants caused by impaction of coils, however, may be higher than by clipping and therefore these patients are possibly even at a higher risk of a recurrence.
The high risk of a recurrent SAH indicates that having an SAH is not a single event in a lifetime. New aneurysms have previously been found in up to 16% of the patients with a history of SAH.3–7 Screening for aneurysms in these patients might be effective, but screening also has disadvantages. Screening carries the risks of angiography and preventive treatment, especially in older patients. In addition, not all aneurysms will be treated, and the knowledge of having an untreated aneurysm has a negative effect on the quality of life of these patients.17 Moreover, screening also has psychosocial consequences.18 Thus, whether a regular screening program for detection of new aneurysms is beneficial for patients who survived after treatment for SAH requires careful weighing of the pros and cons, for example, in a decision analysis.
This study was supported by an established clinical investigator grant from the Netherlands Heart Foundation to G.J.E.R. (grant D98.014) and a grant from the Netherlands Organization for Scientific Research/ZonMw (grant 945-02-007).
- Received June 9, 2005.
- Revision received July 4, 2005.
- Accepted August 23, 2005.
Rinkel GJ, Djibuti M, Algra A, van Gijn J. Prevalence and risk of rupture of intracranial aneurysms: a systematic review. Stroke. 1998; 29: 251–256.
Juvela S. Risk factors for multiple intracranial aneurysms. Stroke. 2000; 31: 392–397.
Wermer MJH, Schaaf van der IC, Velthuis BK, Algra A, Buskens E, Rinkel GJE, for the ASTRA Study Group. Follow-up screening after subarachnoid hemorrhage: frequency of detected aneurysms and risk factors for aneurysm formation and growth. Brain. In press.
Tsutsumi K, Ueki K, Morita A, Usui M, Kirino T. Risk of aneurysm recurrence in patients with clipped cerebral aneurysms. Results of long-term follow-up angiography. Stroke. 2001; 32: 1191–1194.
Juvela S, Poussa K, Porras M. Factors affecting formation and growth of intracranial aneurysms: a long-term follow-up study. Stroke. 2001; 32: 485–491.
Wiebers DO, Whisnant JP, Huston J 3rd, Meissner I, Brown RD Jr, Piepgras DG, Forbes GS, Thielen K, Nichols D, O’Fallon WM, Peacock J, Jaeger L, Kassell NF, Kongable-Beckman GL, Torner JC. Unruptured intracranial aneurysms: natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet. 2003; 362: 103–110.
The ACROSS Group. Epidemiology of aneurysmal subarachnoid hemorrhage in Australia and New Zealand: incidence and case fatality from the Australasian Cooperative Research on Subarachnoid Hemorrhage Study (ACROSS). Stroke. 2000; 31: 1843–1850.
Tsutsumi K, Ueki K, Usui M, Kwak S, Kirino T. Risk of recurrent subarachnoid hemorrhage after complete obliteration of cerebral aneurysms. Stroke. 1998; 29: 2511–2513.
Herman B, Leyten AC, van Luijk JH, Frenken CW, Op de Coul AA, Schulte BP. Epidemiology of stroke in Tilburg, The Netherlands. The population-based stroke incidence register: 2. Incidence, initial clinical picture and medical care, and three-week case fatality. Stroke. 1982; 13: 629–634.
Linn FH, Rinkel GJ, Algra A, van Gijn J. Incidence of subarachnoid hemorrhage: role of region, year, and rate of computed tomography: a meta-analysis. Stroke. 1996; 27: 625–629.
Inagawa T. Trends in incidence and case fatality rates of aneurysmal subarachnoid hemorrhage in Izumo City, Japan, between 1980–1989 and 1990–1998. Stroke. 2001; 32: 1499–1507.
Ohkuma H, Fujita S, Suzuki S. Incidence of aneurysmal subarachnoid hemorrhage in Shimokita, Japan, from 1989 to 1998. Stroke. 2002; 33: 195–199.
van der Schaaf IC, Brilstra EH, Rinkel GJ, Bossuyt PM, van Gijn J. Quality of life, anxiety, and depression in patients with an untreated intracranial aneurysm or arteriovenous malformation. Stroke. 2002; 33: 440–443.
Wermer MJ, van der Schaaf IC, Van Nunen P, Bossuyt PM, Anderson CS, Rinkel GJ. Psychosocial impact of screening for intracranial aneurysms in relatives with familial subarachnoid hemorrhage. Stroke. 2005; 36: 836–840.