Background Ruptured and thrombosed cerebral aneurysms pose a risk for rebleeding that may go undetected because they are angiographically occult. The presence of multiple aneurysms may further complicate efforts in accurately identifying the true source because their treatment may circumvent further necessary investigations.
Methods This case series and literature review illustrates the background and clinical features of ruptured thrombosed cerebral aneurysms.
Results Thrombosed ruptured cerebral aneurysms have several common anatomic and physiologic mechanisms for their occurrence. They may evade detection if not considered part of a diligent thorough approach that includes careful analysis of the pattern of initial bleeding on non-contrast head CT scan.
Conclusions Despite negative angiographic studies, and even when multiple cerebral aneurysms are present, diffuse subrachnoid hemorrhage may still warrant further investigation.
Statistics from Altmetric.com
The bleeding pattern on initial presentation often dictates the management plan for patients with non-traumatic subarachnoid hemorrhage and negative catheter cerebral angiography. If the pattern of hemorrhage is diffuse (not perimesencephalic or cortical), serious potential causes include thrombosed aneurysms and vascular lesions of the cervical spine. The catheter cerebral angiogram is usually repeated in a week. Prior reports have shown that in such cases with diffuse subarachnoid hemorrhage, significant pathology that requires surgical treatment is discovered on the second study 16% (7/44) of the time.1 Until such a diagnosis is made, the unpredictable potential for rebleeding is obvious and underlies the urgency in appropriate decision making.
However, when the ruptured aneurysm fails to opacify because of temporary aneurysm dome thrombosis and another aneurysm is seen, this vigilant search for the actual source of bleeding may be derailed. The treating physician may falsely attribute the hemorrhage to the visible aneurysm, anatomically exclude it, and not pursue additional surveillance investigations while the risk of rebleeding remains unchanged.
This case series and literature review highlights the importance of maintaining a high degree of discipline and vigilance when confronted with a detectable vascular pathology that is not consistent with the hemorrhage pattern. Thrombosed angiographically occult ruptured cerebral aneurysms serve as illustrative examples of such a situation. Even when a neighboring aneurysm is treated, if the clinical suspicion remains, early repeat catheter angiography improves the odds of detecting a potentially thrombosed aneurysm that more reliably explains the initial hemorrhage pattern. We present two cases that demonstrate the importance of not only accurately interpreting the hemorrhage pattern on a non-contrast head CT scan but also maintaining a strong clinical suspicion despite a catheter angiogram revealing unrelated pathology.
Case No 1
A 60-year-old woman was admitted with a headache and found to have diffuse subarachnoid hemorrhage in the right Sylvian fissure, and right basal and prepontine cisterns (figure 1A). CT angiography (CTA) was performed on a Siemens dual source SOMATOM scanner with a contrast volume of 50 mL at 4 mL/s with a slice thickness of 1 mm. For this patient, CTA of the brain was reviewed by the neuroradiologist and ourselves and was found not to reveal any cerebral aneurysms or vascular abnormalities. Catheter angiography was performed using a Siemens biplane Artis zee, with a hand injection of 8 mL of Isovue 250 over 2 s in the selected artery. Catheter angiography demonstrated a 4 mm inferiorly directed right paraclinoid aneurysm. No vascular pathology was noted in the territory of the middle cerebral artery bifurcation on the right. Because we attributed the hemorrhage to this paraclinoid aneurysm, and planar imaging yielded a satisfactory profile of the aneurysm, no additional rotational angiography was performed. We proceeded with coil embolization of this putative source of the subarachnoid hemorrhage (figure 1B). The remainder of her hospital course was uneventful and she was discharged home after 14 days with a normal neurologic examination.
Two weeks later (4 weeks after her initial subarachnoid hemorrhage), she reported worsening headaches and was admitted to a local hospital where a CT scan demonstrated acute hyperdensity in the right Sylvian fissure (figure 1C). Repeat catheter angiography demonstrated a previously occult small 2 mm superiorly and posteriorly directed right middle cerebral artery bifurcation aneurysm with a 1 mm neck. The initial hemorrhage pattern was re-evaluated, and it was decided that this was the original aneurysm that had ruptured but was previously thrombosed and therefore angiographically occult (figure 1D). During embolization, the neck of the aneurysm was noted to be quite narrow, barely accommodating the orifice of the 1.7 F microcatheter. This newly discovered aneurysm was successfully embolized and she was fortunately discharged home after several days of monitoring with a normal neurologic examination (figure 1E).
Case No 2
A 47-year-old woman was admitted for a sudden onset severe headache while lifting weights at the gym. A CT scan demonstrated diffuse subarachnoid hemorrhage involving the ambient and quadrigeminal cisterns on the left, the left Sylvian fissure, the interhemispheric fissure, and intraventricular hemorrhage (figure 2A). CTA demonstrated a 2.5 mm right (contralateral) anterior choroidal artery aneurysm. Catheter angiography was performed and in addition to the right carotid lesion, a 2 mm left posterior communicating artery infundibulum was found. Vessels near the quadrigeminal plate cistern were also evaluated and found to be normal (figure 2B). Given the pattern of the hemorrhage, the left posterior communicating artery region was explored with open surgery and found to be an infundibulum without signs of hemorrhage but nevertheless wrapped. The patient was then brought back for embolization of the right anterior choroidal artery aneurysm as a safeguard. At this point, the subarachnoid hemorrhage was attributed to either the left posterior communicating artery blister aneurysm, which was wrapped, or the right supraclinoid artery aneurysm, neither of which adequately explained her initial hemorrhage pattern. Her clinical course was stable and she was discharged home.
After 2 months she was brought back for surveillance elective catheter angiography and prepared for possible flow diversion embolization of a possible blister left posterior communicating artery aneurysm when an entirely new aneurysm was found, measuring 2 mm in maximal dimensions with a 1 mm neck at the left P2/P3 junction of the posterior cerebral artery (figure 2C). The initial hemorrhage pattern was then re-evaluated and it was decided that this new aneurysm was likely the original ruptured aneurysm but was thrombosed and therefore angiographically occult. During embolization, the neck of the aneurysm was noted to be quite narrow, barely accommodating the orifice of the 1.7 F microcatheter. This newly discovered aneurysm was embolized, and fortunately she was discharged home after several days of monitoring with a normal neurologic examination (figure 2D).
The pattern of subarachnoid hemorrhage, usually one of three types, determines the yield of finding critical lesions on subsequent studies after an initially normal cerebral angiogram.2 In both cases, the initial head CT scans were neither perimesencephalic or cortical, and should have raised a stronger clinical suspicion for something other than what was initially attributed to the hemorrhage.
For the first case, the subarachnoid hemorrhage was erroneously attributed to the paraclinoid aneurysm, which precluded further investigation. In retrospect, the location of the aneurysm did not adequately correlate with the overall pattern of hemorrhage, particularly with so much blood located in the right Sylvian fissure. Further investigation, despite the embolization of the paraclinoid aneurysm was, in retrospect, warranted. Fortunately, despite rehemorrhage, the severity of the injury was relatively mild.
For the second case, the subarachnoid hemorrhage was erroneously attributed to an infundibulum or junctional dilatation. The angiographic finding in this case, as demonstrated in figure 2B, did fit the radiographic criteria, as suggested by Stehbens, in comprising a dilatation triangular in shape, with a base measuring no greater than 3 mm, and a posterior communicating artery in continuity with the apex of the triangle.3 In 1964, Fox et al4 reported a series of five patients with subarachnoid hemorrhage where only junctional dilatations were demonstrated angiographically. Three of these individuals were subjected to craniotomy, and in each instance only the dilatation was found. In no case was there evidence of it having bled.4 Epstein et al5 evaluated seven junctional dilatations of the posterior communicating artery. In all specimens, adventitia, media, internal elastic membrane, and intima were present and of normal appearance. They concluded histologically that such dilatations were a normal variation, and not a site having predilection to bleeding or future true aneurysmal dilatation.5
In both cases, failing to recognize the possibility of spontaneous thrombosis of the ruptured aneurysm on initial catheter angiographic images delayed surveillance imaging. Spontaneous thrombosis of ruptured aneurysms has been reported as far back as Dandy in 1944, where autopsy studies suggested a 15% occurrence rate. Fodstad and Liliequist found spontaneous disappearance of ruptured aneurysms on repeat catheter angiography in 3% of patients with subarachnoid hemorrhage.6 Partial spontaneous thrombosis is more frequently seen with giant aneurysms and has been reported to occur in more than half of cases.7–9
Several proposed anatomic and physiologic mechanisms for spontaneous thrombosis have been proposed. Animal modeling has determined a chamber volume to orifice ratio of greater than 28 : 1 to be associated with spontaneous aneurysm thrombosis.10 Aneurysms with a small orifice experience lower flow velocity, different flow direction, and smaller shear forces than those with a large orifice.11 One study looked at seven ruptured aneurysms that were initially angiographically occult but subsequently recanalized, and found a trend of smaller aneurysms. There was no mention of neck size. All but one of these cases demonstrated, as in our case, an aneurysm diameter of less than 3 mm. They hypothesized that vasospasm, thrombosis, arterial dissection, and operator interpretative error contributed to the initial finding. They found the anterior communicating artery complex to be the most common site for initially occult thrombosed aneurysms.12
Several physiologic mechanisms are thought to contribute to spontaneous thrombosis of a ruptured cerebral aneurysm. These include: increased intracranial pressure, nearby vasospasm, systemic hypotension, and possibly systemic administration of antifibrinolytics.13 The final common mechanism from these processes consists of lower flow within the dome of the aneurysm which increases the propensity for coagulation. The exact role of each of these is unclear. One study of 83 angiographically negative subarachnoid hemorrhage patients demonstrated an equivocal role for aminocaproic acid with about an equal 5–6% risk of ischemic stroke regardless of whether it was administered.14
As expected, spontaneous thrombosis does not appear to reduce the risk of future aneurysm re-rupture. Other reports, like ours, have shown subsequent recanalization after temporary thrombosis of a single aneurysm.15 ,16 Whittle's series of 12 patients with partial or complete thrombosis of giant intracranial aneurysms demonstrated that nearly half developed recurrent subarachnoid hemorrhage.9 The mechanism likely entails recanalization of the aneurysmal clot at some point and re-exposure of the aneurysmal wall to pulsatile arterial blood flow along with protease activity.
In our series, there were no signs of vasospasm, systemic hypotension, administration of aminocaproic acid, or a large aneurysm dome diameter. In fact, the only contributing anatomic or physiologic factors may have been increased intracranial pressure and an aneurysm neck diameter of 1 mm or less, making for a relatively large chamber volume to orifice ratio.
Successful detection of ruptured cerebral aneurysms therefore lies not exclusively on the results of the catheter angiographic images but on thoughtful interpretation of all the available data, including the clinical history and pattern of subarachnoid hemorrhage on CT. Despite being commonly referred to as the ‘gold standard’ in cerebrovascular imaging, and the availability of newer applications, such as three-dimensional rotational reconstructions, there still remains a conspicuous false negative rate.17
Multiple cerebral aneurysms occur in about 30% of cases.18 When multiple aneurysms are present, angiographic signs thought to be fairly indicative of recent rupture include focal spasm, focal mass effect, and aneurysm daughter sacs. Larger size has been traditionally reported as an important determinant in the site of rupture but more recent series seem to suggest morphology is more significant.19 However, it is obvious that when the aneurysm is completely thrombosed, angiographic criteria are meaningless, and determining the cause of the subarachnoid hemorrhage in the initial stages may depend more on the non-contrast CT scan.
Exploratory surgery could be considered when there is a high clinical suspicion that a ruptured and thrombosed aneurysm exists, despite a negative catheter angiographic study. Although published nearly 20 years ago, one series of five patients over 5 years were found to have aneurysms discovered only at surgery with previous negative angiograms and were successfully clipped.13 The limitations of the catheter angiogram studies performed 20 years ago may account for this finding. But these cases still shared the characteristic diffuse bleed pattern highly suggestive of a ruptured aneurysm.
Ruptured and thrombosed aneurysms pose a risk for rebleeding that may go undetected because they are angiographically occult. The presence of multiple cerebral aneurysms may further misdirect efforts to accurately identify the ruptured aneurysm because their treatment may circumvent further necessary early surveillance investigations. Careful analysis of the pattern of initial bleeding on non-contrast head CT scan is valuable in successfully determining the site of hemorrhage early on. Despite negative angiographic studies, and even when multiple cerebral aneurysms are present, diffuse subrachnoid hemorrhage may warrant further investigation.
Competing interests None.
Contributors MC is the sole author of this manuscript.
Patient consent Obtained.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement The author agrees to share any data on request.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.