Ruptured Dissecting Aneurysm as a Cause of Subarachnoid Hemorrhage of Unverified Etiology
Background and Purpose The clinical features of “aneurysmal” subarachnoid hemorrhage (SAH) of angiographically unverified etiology were reviewed to clarify the incidence and natural history of dissecting aneurysms as the hemorrhagic source of SAH.
Methods We reviewed 30 patients with SAH of unverified etiology in whom initial CT scan showed a diffuse or anteriorly distributed subarachnoid blood clot. Ten of the patients had stenotic or occlusive lesions (SOCL) on initial angiography, and these were the main focus of this study.
Results Among the 10 patients with SOCL on initial angiography, the lesions were located on the anterior circulation in 6 and on the posterior circulation in 4. Ruptured dissecting aneurysms were confirmed by exploratory surgery or autopsy in 6 patients. Subsequent rupture occurred in 6 of the 10 patients (60%), and all 6 of these patients died as a result.
Conclusions The incidence (6/30) of dissecting aneurysms as the cause of SAH of unverified etiology was unexpectedly high, especially when initial angiography disclosed SOCL (6/10). The moribund patients with SOCL showed a high rate of rebleeding, and the untreated recurrent hemorrhages were fatal. Further MRI study is indicated for these patients to demonstrate the intramural hematoma. Compared with the devastating mortality caused by the subsequent ruptures, the extent of surgical morbidity was minor. Surgical intervention could therefore be justified when the following neuroradiological findings are present: (1) SOCL evident on angiography, (2) distribution of SAH on CT compatible with the location of the SOCL, and (3) intramural hematoma on MRI in the same region as the SOCL.
Approximately 13% to 22% of patients suffering from spontaneous subarachnoid hemorrhage (SAH) are diagnosed as having SAH of unverified etiology when the source of bleeding cannot be determined despite extensive neuroradiological investigation.1 Several causes of bleeding can be considered in patients with normal angiograms who show diffuse or anteriorly located hemorrhages in the basal cisterns: tiny or thrombosed aneurysm, dural arteriovenous malformation (AVM), spinal AVM, or trauma.1 Some reports indicate that vertebral or carotid dissecting aneurysms could cause such bleeding; dissecting aneurysms often demonstrate subtle angiographic features in the acute stage without any presentation of typical double lumen or pearl and string signs.1 Although identification of a double lumen is the main diagnostic feature of dissecting aneurysms, some ruptured dissecting aneurysms appear only as stenotic or occlusive lesions (SOCL). Here we discuss the significance, incidence, and natural course of dissecting aneurysms among SAH cases of unverified etiology, particularly those in which initial angiography disclosed nonatherosclerotic SOCL.
Subjects and Methods
During an 11-year period from April 1984 through August 1995 at Showa General Hospital, the authors reviewed and evaluated 767 patients with nontraumatic SAH, 648 of whom underwent thorough angiographic investigation to determine the source of the SAH within 48 hours after onset. All the patients showed the typical clinical symptoms and signs of SAH, which was confirmed by CT scan and/or lumbar puncture. Among these patients, 594 had saccular aneurysms and 14 had vertebrobasilar dissecting aneurysms. In the remaining 40 patients with negative initial angiographic findings, other neuroradiological modalities disclosed 1 dural AVM, 2 spinal AVMs, and 1 giant thrombosed aneurysm as the source of hemorrhage. Of the remaining 36 patients, 6 with a perimesencephalic pattern of hemorrhage had an invariably good prognosis. The remaining 30 patients with diffuse or anteriorly located subarachnoid blood clots on early CT scan were divided by the presence or absence of SOCL on initial angiography. It was hypothesized that patients with SOCL would be more likely to have dissecting aneurysms. In total, 10 patients had SOCL (5 men and 5 women ranging in age from 41 to 78 years, with a mean age of 53.0 years) (Table 1⇓). Patients with positive signs of subtle abnormality at the arterial bifurcation and those with SOCL were considered as candidates for surgery. For the latter group, further MRI study focused on the region of angiographically evident SOCL is indicated to visualize the intramural hematoma along the arterial flow void. Exploratory surgery disclosed four semifusiform aneurysms in the former group and three dissecting aneurysms in the latter group. Sixteen patients without SOCL and subtle angiographic signs at arterial bifurcation underwent serial angiography, which disclosed saccular aneurysms in 9 patients. The initial status of patients was evaluated according to World Federation of Neurological Surgery grade, and the Glasgow Outcome Scale was used for assessment of the patients’ condition at 6 months after the initial hemorrhage.
Incidence and Outcome of Subsequent Rupture
Among 10 patients with SOCL, the rate of rebleeding of the patients with a severe clinical grade (3 to 5) was significantly higher than that of the patients with a fair clinical grade (1 to 2) (two-tailed P=.00476). The rebleeding rate in 10 patients with angiographically confirmed SOCL was significantly higher than that of the other 20 patients without SOCL (two-tailed P=.030). By Fisher’s exact test, this result was statistically significant with 95% confidence (Table 2⇓).
Among 10 patients, 6 had diffuse subarachnoid clots and 3 had focal subarachnoid clots limited to the unilateral sylvian cistern. The remaining patient was diagnosed as having SAH by lumbar tap only. In the 9 patients with positive CT scans, the clot was distributed predominantly in the cisterns adjacent to the SOCL.
All the patients with SOCL met the following conditions: (1) initial angiography was performed within 48 hours after onset, (2) angiographic SOCL indicated single subtle luminal narrowing or vascular occlusion at a point other than a vascular bifurcation, and (3) adequate four-vessel studies disclosed absence of vasospasm and no atherosclerotic changes in the other arterial systems. The SOCL locations are listed in Table 1⇑ (Fig 1⇓).
Among the 10 patients with SOCL, MRI findings were unavailable for the first 4 patients (pre-MRI era.). The other 6 patients were scheduled for investigation, but 4 with severe clinical grade did not undergo MRI because of unstable vital signs and subsequent ruptures. The 2 patients who underwent MRI study showed intramural hematomas (Fig 2A⇓ and 2C⇓).
On the basis of high rebleeding rate, exploratory surgery was considered to be indicated when the following neuroradiological findings were obtained: (1) SOCL evident on angiography, (2) distribution of SAH compatible with the location of the SOCL on CT, and (3) intramural hematoma in the same region as the SOCL on MRI. The diagnosis of ruptured dissecting aneurysm was confirmed in all the surgical cases (Fig 2B⇑ and 2D⇑). In our series, surgical morbidity and mortality were minor (0%). Wrapping or clip-reinforced wrapping procedures were performed in our series. No subsequent ruptures were observed during the follow-up period (41 to 95 months) (Table 1⇑).
Among 3 autopsied patients, dissecting aneurysms were verified on the basilar artery, vertebral artery, and internal carotid artery, respectively. Fragmentation of the degenerated internal elastic lamina was most commonly recognized on microscopic study. Multiple intramural hemorrhages without luminal connections were also observed in all patients (Fig 3A⇓ and 3B⇓).
Despite improvements in neuroradiological techniques, SAH of unverified etiology is not rare.1 Several studies have demonstrated that 4.7% to 8.6% of patients with such SAH eventually succumb to a subsequent hemorrhage.2 3 Some of these patients were later found to have aneurysms that were not detected during initial diagnostic evaluation.3 In contrast, little attention has been focused on dissecting aneurysms as a cause of SAH of unverified etiology.4 In our 11-year experience, the hemorrhagic sources of spontaneous SAH were revealed to be aneurysms in 594 patients and dissecting aneurysms in 20 patients. Among the latter 20, 6 had SOCL only on the initial angiograms. A substantial proportion (30%) of these dissecting aneurysms showed subtle abnormality on initial angiography. We verified a high ratio (6/30) of dissecting aneurysms as the cause of SAH cases in which the etiology was unverified by angiography. We also demonstrated that patients with SOCL of severe clinical grade showed a high incidence of rebleeding and that recurrent hemorrhages were fatal. The same tendency has been described in previous studies of ruptured dissecting aneurysms with negative initial angiographic findings. There have been 12 reported cases of dissecting aneurysm as a cause of SAH with unverified etiology either in the carotid or posterior circulation. Six of the patients with SAH and dissection in the carotid distribution have been reported,4 5 6 7 8 9 and the clinical outcome was devastating in all cases. In vertebrobasilar distribution, there are reports of six cases in which initial angiography failed to reveal the cause of the SAH and the dissecting aneurysm was revealed by subsequent angiography.10 11 12 13 14 15
Angiography does not always permit a definitive diagnosis because luminal narrowing or vascular occlusion at a site other than a vascular bifurcation is also seen in atherosclerotic or other vascular diseases. It has been suggested that the pathognomonic sign for a dissecting aneurysm is a double lumen, although this is infrequently demonstrated.5 Angiographic signs may be understood in terms of the status of thrombosis in the pseudolumen; thus, a double lumen could result from minimum thrombosis, a pearl and string sign could result from partial thrombosis, and subtle SOCL could result from early progression of thrombosis. Although dissecting aneurysms often demonstrate only subtle abnormalities in the acute stage, early thrombosis might not necessarily indicate healing of the dissected pseudolumen. This was demonstrated in our series and in other reported cases4 5 6 7 8 9 10 11 12 13 14 15 (Table 1⇑). Cerebral vasospasm due to SAH occasionally has the appearance of a string sign. Cerebral vasospasm commonly occurs in multiple vessels and is hardly seen at the acute stage. Differentiating SOCL from incidental atherosclerotic stenosis may be important. Another definitive diagnostic feature of arterial dissection is intramural hematoma. MRI is very valuable when combined with angiography because MRI can directly demonstrate the intramural hematoma.16 In case 6 of our patients, a high-intensity lesion with marked contrast enhancement on T1-weighted imaging along the C1 portion of the right internal carotid artery was observed at the subacute stage (Fig 2A⇑). In case 7, an isointense lesion along the M1 segment of the right middle cerebral artery was observed at the acute stage on proton-density imaging. In this case, T1-weighted imaging was not diagnostic because of its inappropriate slice position (Fig 2C⇑). In both cases, we clearly demonstrated intramural hematoma, shown as a crescentic (case 6) or curvilinear (case 7) isointensity to high-intensity structure in the lumen, in the same region where SOCL were revealed angiographically. In our series, there was no false-positive MRI demonstration of intramural hematoma in cases without SOCL. With the use of MRI, we could successfully diagnose dissecting aneurysms as the hemorrhagic sources among SAHs of unverified etiology and discriminate them from atherosclerotic stenosis.
Our experience and review of the literature suggested that these lesions posed a significant risk of rebleeding if untreated. The outcome in all patients presenting with a severe clinical grade was fatal (Table 1⇑).4 5 6 7 8 9 10 11 12 13 14 15 When such lesions are encountered as the likely cause of SAH, diagnosis with the aid of MRI and obliteration of the dissected lesions are essential. Early surgery may be justified to prevent rerupture: our series and a review of the literature revealed a greater risk of subsequent rupture within the first week (Table 1⇑).4 5 6 7 8 9 10 11 12 13 14 15 If a diagnosis of dissecting aneurysm is highly suspected, trapping of the aneurysm or proximal ligation of the artery with or without extracranial-intracranial bypass surgery would be the optimal surgical treatment for prevention of rebleeding, although the best surgical procedure is still being debated.5
In conclusion, there was an unexpectedly high ratio (6/30) of dissecting aneurysms as the cause of SAH of unverified etiology. A substantial proportion (30%) of dissecting aneurysms showing only subtle angiographic SOCL in the acute stage was noted. There was a high mortality rate (6/10) due to subsequent rupture of such SOCL. We also demonstrated that patients with a severe clinical grade of SOCL showed a high rate of rebleeding and that untreated recurrent hemorrhages were fatal. Further MRI study is indicated for these patients to demonstrate the intramural hematoma. Compared with the devastating mortality caused by the subsequent ruptures, the extent of surgical morbidity was minor. Surgical intervention could therefore be justified when the following neuroradiological findings are present: (1) SOCL evident on angiography, (2) distribution of SAH on CT compatible with the location of the SOCL, and (3) intramural hematoma on MRI in the same region as the SOCL.
Reprint requests to Hirofumi Nakatomi, MD, Department of Neurosurgery, Fuji Brain Institute, 270-12 Sugita, Fujinomiya City, Shizuoka 418, Japan.
- Received February 6, 1997.
- Revision received March 21, 1997.
- Accepted March 21, 1997.
- Copyright © 1997 by American Heart Association
Rinkel GJE, van Gijn J, Wijdicks EFM. Subarachnoid hemorrhage without detectable aneurysm: a review of the causes. Stroke. 1993;24:1403-1409.
Hochberg FH, Bean C, Fisher CM, Roberson GH. Stroke in a 15-year-old girl secondary to terminal carotid dissection. Neuroradiology. 1975;25:725-729.
Kitanaka C, Tanaka J, Kuwahara M, Teraoka A. Magnetic resonance imaging study of intracranial vertebrobasilar artery dissections. Stroke. 1994;25:571-575.