The clinical conundrum of convexal subarachnoid hemorrhage

Nontraumatic subarachnoid hemorrhage (SAH) classically involves the subarachnoid space of the basal cisterns and the cerebral convexity, with ruptured aneurysms of the circle of Willis accounting for about 80% of the cases. The remainder, nonaneurysmal forms of SAH, are a heterogeneous group, in which about one-third (5% of total SAH) are limited to the cisterns that surround the brainstem, and are thought to be due to rupture of venous structures in the perimesencephalic region,^1,2 although other potential mechanisms have been postulated.³

A much less well-characterized form of SAH is the variety that occurs in the cerebral convexity as a localized focus of bleeding in one or several adjacent sulci, without associated SAH at the base of the brain or elsewhere. This form of “convexal” SAH (cSAH) is frequently encountered in clinical practice, and presents at times with acute headache suggestive of SAH, but often it is an unexpected finding on imaging in patients evaluated for a variety of symptoms, including change in mental status, transient focal neurologic deficits, or partial seizures. This variety of SAH often escapes etiologic diagnosis despite comprehensive imaging and laboratory workup. It is thus encouraging that, in this issue of Neurology®, Kumar et al.⁴ report data on a retrospective study of 29 patients with cSAH that shed some light on the subject. Two main diagnostic groups emerged in their clinical series: 1) patients older than 60 years who often presented with transient sensory symptoms and neuroimaging studies that showed features suggestive of cerebral amyloid angiopathy (CAA); 2) patients younger than 60 years who presented frequently with abrupt thunderclap headache, who were thought to have features of the reversible cerebral vasoconstriction syndrome (RCVS).⁵ Although this study has the inherent limitations of retrospective series from referral institutions, the clinical characterization of the 2 groups of patients with potentially different mechanisms of cSAH in 2 different age groups is helpful as a starting point for further analysis of this form of SAH.

One of the limitations of the series is that only 55% of patients underwent cerebral angiography. As the authors acknowledge, mycotic aneurysms, cortical vein thrombosis, micro-AVM, pial or dural fistulas cannot be excluded in patients who did not undergo cerebral angiography. While CT angiography and magnetic resonance angiography are excellent noninvasive studies, their ability to detect dural fistulas and their sensitivity in the detection of aneurysms less than 3 mm in diameter remains inferior to angiography.⁶ In addition, the relatively low frequency of cerebral angiography limits the generalizability of the observations related to the diagnosis of the RCVS in most patients less than 60 years old. Since the diagnosis of RCVS requires not only the clinical syndrome of sudden severe headache, but also angiographic demonstration of segmental cerebral arterial vasoconstriction and its reversibility,⁷ such mechanism can only be presumptive when based on clinical information alone. Since angiographic arterial narrowing was described in only 4 of 16 patients less than 60 years old, and reversal of constriction was demonstrated by angiography in only 1 patient, it may be prudent to view the cause of cSAH in patients in this age group who do not undergo angiographic imaging and follow-up as belonging to an “undetermined” group, pending further characterization of this entity by systematic and prospective evaluation of the cerebral vasculature.

The diagnosis of CAA was based on clinical and, mostly, imaging criteria, without pathologic confirmation of this entity. Although the latter is the only way of positively diagnosing CAA, the clinical and radiologic criteria applied by the authors are consistent with that diagnosis,⁸ further favored by the subsequent occurrence of intracerebral hemorrhage in 4 of 10 patients diagnosed with CAA.

At variance with other reported series,^9,10 this study documented a low frequency (1 of 16 subjects) of the posterior reversible encephalopathy syndrome (PRES) as a mechanism of cSAH, which was present in 8 of 20 subjects in the younger than 60 years age group in those 2 studies combined. Is this due to an artifact of case selection in the series of Kumar et al.? Since cases were accrued by searching for the International Classification of Diseases–9 code for SAH, it is possible that patients presenting with the primary diagnosis of PRES may or may not have cSAH coded as a separate diagnosis. This and other remaining issues related to the etiologic characterization of cSAH are likely to be elucidated in future prospective studies. In the meantime, we are grateful to Kumar et al., who have added an important set of clinical observations to our understanding of cSAH. This will hopefully spawn more studies that will further define the diagnostic categories that underlie this elusive form of SAH.

DISCLOSURE

Dr. Kase serves on scientific advisory boards for Sanofi-Aventis and Lundbeck A/S and receives research support from the NIH (NHLBI N01-HC-25195 [Co-I], NINDS 2 R01-NS17950-27 [Co-I], NINDS 2U01NS38529-04A1 [Co-I], and NINDS 1 R01 NS 44876 01A2 [Co-I]). Dr. Nguyen serves as an Associate Editor of Frontiers in Vascular and Interventional Neurology and estimates that 80% of her clinical effort is spent on cerebral angiography and cerebral aneurysm embolization.