X. Wu, V.B. Kalra, H.P. Forman
Department of Diagnostic Radiology

C.C. Matouk
Department of Neurology and Neurosurgery

G. Mongelluzzo, R. Liu, A. Malhotra
Department of Diagnostic Radiology

Yale School of Medicine
New Haven, Connecticut

We would like to thank Heit et al¹ for their study “Cerebral Angiography for Evaluation of Patients with CT Angiogram-Negative Subarachnoid Hemorrhage: An 11-Year Experience” on the utility of digital subtraction angiography in patients with negative findings on CT angiography and subarachnoid hemorrhage. This is a laudable effort in addressing an issue with great heterogeneity in literature. However, we would like to raise a few questions regarding the article.

First, the statement that all patients with negative findings on CTA should be considered for DSA should be viewed with caution, especially for patients with perimesencephalic hemorrhage (pSAH). The authors reported that 2 aneurysms and 1 case of vasculitis were identified on DSA as causes of pSAH, which were initially missed on CTA. Heit et al¹ stated in the “Materials and Methods” section that if an aneurysm was identified by DSA after negative findings on CTA, the CTA was reviewed retrospectively. However, the results of that review were not available in the article. It would be helpful to know the number of cases with positive findings that could be retrospectively seen on CTA with the hindsight of the DSA results. On the other hand, in our own review of the literature, we found very few and questionable cases of pSAH in which imaging was of utility after initial negative findings on CTA.²

The authors quoted Delgado Almandoz et al³ to support the utility of follow-up DSA after negative findings on CTA because 1 aneurysm was detected on follow-up. On careful review of that article, in particular Fig 3B (the initial DSA that supposedly missed the 2-mm left superior cerebellar artery branch aneurysm), the aneurysm can be identified when correlated with subsequent DSA (Fig 3D), but the image quality was different, likely due to technical differences.² It would thus be more helpful to review the cases of missed aneurysms responsible for pSAH reported by Heit et al¹ and see whether they could be identified on the initial study in retrospect. On the basis of a literature review, Westerlaan et al⁴ found that 27% (19 of 71) of false-negative ruptured intracranial aneurysms could be detected at CTA retrospectively.

In the cases with pSAH caused by vasculitis or reversible cerebral vasoconstriction syndrome (RCVS), it would be unusual for RCVS to present with pSAH. Most of RCVS SAH tends to be convexity sulcal SAH.⁵ Vasospasm, although rare with pSAH, can also occur. The information about whether the 1 patient with SAH diagnosed as RCVS met the diagnostic criteria of SAH would be helpful.

Of further interest are the 16 patients with xanthochromia whose initial CTAs and subsequent DSAs had negative findings. Some publications have recommended the possible use of CTA to replace lumbar puncture in patients with thunderclap headache, and many institutions have used these papers to justify increased use of CTA for this indication.⁶ The absence of angiographic findings in the current study raises questions about that assumption.⁷

The authors’ recommendation that all patients with an initial CTA negative for pSAH should continue to undergo DSA needs further evidence and support. The current literature does not strongly support it, and this recommendation needs to be assessed in terms of its cost-effectiveness.⁸

References

1. Heit JJ, Pastena GT, Nogueira RG, et al. Cerebral angiography for evaluation of patients with CT angiogram-negative subarachnoid hemorrhage: an 11-year experience. AJNR Am J Neuroradiol 2016;37:297–304
2. Kalra VB, Wu X, Matouk CC, et al. Use of follow-up imaging in isolated perimesencephalic subarachnoid hemorrhage: a meta-analysis. Stroke 2015;46:401–06
3. Delgado Almandoz JE, Jagadeesan BD, Refai D, et al. Diagnostic yield of repeat catheter angiography in patients with catheter and computed tomography angiography negative subarachnoid hemorrhage. Neurosurgery 2012;70:1135–42
4. Westerlaan HE, van Dijk JM, van Dijk MJ, et al. Intracranial aneurysms in patients with subarachnoid hemorrhage: CT angiography as a primary examination tool for diagnosis—systematic review and meta-analysis. Radiology 2011;258:134–45
5. Miller TR, Shivashankar R, Mossa-Basha M, et al. Reversible cerebral vasoconstriction syndrome, Part 1: epidemiology, pathogenesis, and clinical course. AJNR Am J Neuroradiol 2015;36:1392–99
6. McCormack RF, Hutson A. Can computed tomography angiography of the brain replace lumbar puncture in the evaluation of acute-onset headache after a negative noncontrast cranial computed tomography scan? Acad Emerg Med 2010;17:444–51
7. Malhotra A, Wu X, Kalra VB, et al. Cost-effectiveness analysis of follow-up strategies for thunderclap headache patients with negative non-contrast CT. Acad Emerg Med 2016;23:243–50
8. Kalra VB, Wu X, Forman HP, et al. Cost-effectiveness of angiographic imaging in isolated perimesencephalic subarachnoid hemorrhage. Stroke 2014;45:3576–82

Reply

J.J. Heit
Department of Interventional Neuroradiology
Stanford University Medical Center
Stanford, California

J.D. Rabinov
Department of Interventional Neuroradiology
Massachusetts General Hospital
Boston, Massachusetts

We thank Dr Wu and colleagues for their comments regarding our recent article “Cerebral Angiography for Evaluation of Patients with CT Angiogram-Negative Subarachnoid Hemorrhage: An 11-Year Experience.”¹ We agree that there remains much heterogeneity in the literature regarding CTA-negative subarachnoid hemorrhage, which leads to complicated management decisions that necessarily compare the financial cost of repeat imaging with the risk of missing a ruptured cerebral aneurysm.

In response to the first question raised by our colleagues, a retrospective review of the CTA studies in patients with perimesencephalic SAH (pSAH) due to rupture of an aneurysm did not reveal the culprit aneurysms. The vessel irregularity in the patient with pSAH due to vasculitis or vasculopathy was not convincingly detectable on the original CTA. This patient presented with a headache 8 days after delivering a baby, which might suggest a diagnosis of reversible cerebral vasoconstriction syndrome, but further clinical follow-up was not available. Thus, our results do suggest that DSA is helpful in cases of CTA-negative pSAH. We would argue that the referenced data by Westerlaan et al,² in which ruptured aneurysms were missed by CTA and identified in 27% of cases on re-review, should lead to additional caution regarding the reliability of CTA. Missed aneurysms may be found retrospectively, but to our knowledge, no study compares the sensitivity of a secondary and independent review of a CTA with negative findings with a digital subtraction angiogram. Such a study would certainly be of interest to undertake.

The authors’ article questioning the cost-effectiveness of digital subtraction angiography for evaluation of pSAH was well-written and compelling.³ However, we find it challenging to calculate accurately the cost of missing a cerebral aneurysm in a patient who subsequently has a second SAH and is left with a poor clinical outcome. Rerupture of a missed aneurysm in a single young patient would be expected to incur millions of dollars in health care costs if that patient survives and is left with a large disability, as occurs in one-third of patients with rupture of a cerebral aneurysm.⁴ There is variability in the literature in determining the yield of diagnostic cerebral angiography in patients with CTA-negative SAH also makes cost-effectiveness analyses difficult.

We disagree with the publications that suggest that CTA should replace a lumbar puncture in patients with the sudden onset of a severe headache. Cerebral aneurysms very rarely cause headaches in the absence of subarachnoid hemorrhage. A patient with a severe headache and a noncontrast head CT that does not demonstrate evidence of SAH should always undergo lumbar puncture. If the lumbar puncture is positive for xanthochromia, cerebral vessel imaging should be performed to identify a treatable cause of the SAH. Performing a CTA before a lumbar puncture would lead to a large number of incidentally identified unruptured aneurysms, and a lumbar puncture would still be required to determine whether the identified aneurysms should be treated acutely. This strategy would be expected to lead to increased costs due to additional imaging follow-up of these incidentally identified aneurysms and likely overtreatment of small cerebral aneurysms.

Last, we are all informed by our personal biases, values, and experiences as physicians. As interventional neuroradiologists who care for patients with ruptured aneurysms, we are very cognizant of the risk of re-rupture of a cerebral aneurysm, which is almost always a devastating or fatal event. Although perimesencephalic hemorrhage is very unlikely to be secondary to a ruptured cerebral aneurysm, we continue to believe that the minimal risk of diagnostic cerebral angiography (<0.2% at the authors’ institutions) outweighs the risk of missing a ruptured aneurysm by not performing the criterion standard examination.

References

1. Heit JJ, Pastena GT, Nogueira RG, et al. Cerebral angiography for evaluation of patients with CT angiogram-negative subarachnoid hemorrhage: an 11-year experience. AJNR Am J Neuroradiol 2016;37:297–304
2. Westerlaan HE, van Dijk JM, van Dijk MJ, et al. Intracranial aneurysms in patients with subarachnoid hemorrhage: CT angiography as a primary examination tool for diagnosis—systematic review and meta-analysis. Radiology 2011;258:134–45
3. Kalra VB, Wu X, Forman HP, et al. Cost-effectiveness of angiographic imaging in isolated perimesencephalic subarachnoid hemorrhage. Stroke 2014;45:3576–82
4. Al-Shahi R, White PM, Davenport RJ, et al. Subarachnoid haemorrhage. BMJ 2006;333:235–40