Introduction

Acute severe headache is a common presenting complaint for 1–2 % of patients who attend the emergency department [1]. Most have primary headache disorders (such as migraine and tension-type headache) associated with benign clinical course; however, a subgroup of 1–4 % of these patients has a potentially life-threatening subarachnoid haemorrhage (SAH) [1–3].

Cerebral aneurysm without SAH has also been postulated as a cause of acute severe headache. It has been reported that 15–37 % of patients with documented SAH have experienced an attack of acute severe headache before the index episode of bleeding [4–7].

Although digital subtraction angiography (DSA) has been the “gold standard” diagnostic test in this clinical situation, its performance is often impractical as a screening tool because of its cost, availability and invasiveness. Compared with DSA, computed tomographic angiography (CTA) is a non-invasive examination technique that has proved valuable for the screening of cerebrovascular disease [8–11].

The purpose of this study was to investigate the prevalence and characteristics of intracranial vascular lesions in patients with acute severe headache with the use of CTA.

Materials and methods

The protocol of this retrospective study was approved by the institutional review board, and requirement for informed consent was waived as the patients’ data were evaluated retrospectively and anonymously.

Patient population

The databases of our Hospital Information System and Radiology Information System-Picture Archiving and Communication System (PACS) were searched for patients presenting with acute severe headache, who had also undergone cerebral CTA. Patients were included in this study if they met the following criteria: (1) age ≥ 15 years; (2) no history of previous surgery for intracranial vascular lesions; (3) sudden-onset headache different in quality and intensity from previous headaches; (4) no severe neurological deficits (e.g. seizure, hemiparesis/hemiplegia, visual abnormality); (5) unenhanced head CT did not reveal evidence of SAH or other intracranial abnormality (e.g. mass, hydrocephalus, cerebral oedema). Patients who had SAH were excluded from the study, because we believe that evaluation of patients with SAH is a different issue.

From January 2003 until July 2012, 512 patients satisfied these criteria; this group included 251 men and 261 women (mean age, 46.2 years ± 12.4; range, 16–86 years).

CT angiography

The CTA examinations were performed using 16-detector-row CT (MX8000 IDT, Philips Medical Systems, Haifa, Israel; pitch, 0.35; gantry rotation time, 50 ms; 120 kV; and 200 mAs; n = 207), a 64-detector-row CT (Brilliance 64, Philips Medical Systems; pitch, 0.52; gantry rotation time, 50 ms; 120 kV; and 300 mAs; n = 150), or a 256-detector-row CT (Brilliance iCT, Philips Medical Systems; pitch, 0.40; gantry rotation time, 75 ms; 120 kV; and 160 mAs; n = 155).

The imaging range was planned in a caudocranial direction from the foramen magnum through the vertex. For optimal intraluminal contrast enhancement, the delay time between the start of contrast material administration and the start of imaging was determined for each patient individually by using a bolus-tracking technique. A total of 100 ml iohexol (Omnipaque 300; GE-Healthcare, Princeton, NJ, USA) or iopromide (Ultravist 300; Schering, Berlin, Germany) was administered at a rate of 3 ml/s through an 18-gauge needle positioned in a peripheral vein.

The volumetric data thus obtained were transferred to a workstation with commercially available software (RAPIDIA 3D; Infinitt, Seoul, Korea) for further processing. Transverse sections were reconstructed with a section width of 0.5 mm. CTA images were processed from the obtained source images by using two different methods: (1) volume-rendered technique (VRT) algorithm; (2) VRT images after automatic segmentation of a pre-contrast image dataset (i.e. any overlapping bony structures, calcification or surgical materials).

All acquired CTA images were transferred to a PACS workstation (Pi-ViewStar, Infinitt) and independently evaluated by two board-certified neuroradiologists, who had 17 (D.Y.Y.) and 12 (H.C.K.) years of experience in CT vascular imaging and angiography. The CTA images were presented to the blinded reviewers in random order.

For each patient, reviewers were asked to classify the CTA findings into one of three categories: (1) no significant vascular lesion, (2) aneurysm and (3) vascular lesions other than aneurysm. Atherosclerotic changes (wall calcification, luminal irregularity, stenosis or occlusion) were not considered as significant vascular lesions, as they are not rapidly progressive and usually do not cause acute severe headache. For patients in whom (an) aneurysm(s) was detected on CTA, the location, size (mm, maximum dimension) and shape (lobulating vs oval/round) of the aneurysm were recorded. The size measurements were performed on the workstation using electronic callipers after appropriate magnification.

In any cases of disagreement between the two reviewers, a final decision was reached by consensus. Cases requiring adjudication included those for which the first reviewer detected an aneurysm that was not identified by the second reviewer, aneurysms detected by both reviewers but at different anatomical sites, aneurysms identified by both reviewers in which the size measurement differed by > 1 mm, or aneurysms identified by both reviewers in which the assessed shape was different.

Digital subtraction angiography

All DSA was performed transfemorally with a DSA unit (Integris Allura; Philips Medical Systems, Best, the Netherlands) with an image intensifier matrix of 1,024 × 1,024 pixels. DSA was performed with bilateral selective carotid artery injections and either unilateral or bilateral vertebral artery injections. Anteroposterior, lateral and, if necessary, oblique view(s) of each vessel were obtained by the injection of 6–9 ml iodixanol (Visipaque 320; GE-Healthcare, Princeton, NJ, USA).

The two independent reviewers (D.Y.Y. and H.C.K.) also evaluated DSA images. DSA images were analysed separately from CTA images; anonymous DSA images were given in random order to the reviewers, 8 weeks after each reviewer completed the analysis of CTA images. In cases of disagreement, a final conclusion was reached by consensus.

Data analysis

The primary analysis of this study was the estimation of the prevalence of aneurysms detected in patients with acute severe headache. The prevalence was estimated and a 95 % confidence interval was constructed for the estimate. As a secondary analysis, to investigate gender- and age-related differences in vascular lesions, the chi-squared test was used for comparison between male and female patients and between patients ≤ 45 years and those > 45 years of age.

Results

Thirty-four (6.6 %; 95 % confidence interval, 4.5–8.8 %) of the 512 patients had intracranial vascular lesions on CTA, including aneurysm (n = 31), moyamoya disease (n = 2) and arterial dissection (n = 1). No statistically significant difference was observed in the prevalence of vascular lesions between male (15/251, 6.0 %) and female (19/261, 7.3 %) patients (P = 0.6784). The prevalence in patients > 45 years of age (23/280, 8.2 %) was higher than in patients ≤ 45 years of age (11/232, 4.7 %), but the difference between the two groups did not reach statistical significance (P = 0.1637).

Thirty-three saccular aneurysms were detected in 31 patients: two patients had two aneurysms each. The characteristics of the aneurysms and other vascular lesions are summarised in Table 1. Aneurysms arose most commonly on the internal carotid artery (n = 12), followed in frequency by the anterior communicating artery (n = 7), the middle cerebral artery (n = 7), the posterior communicating artery (n = 5) and other locations (n = 2). Maximal diameters ranged from 2.0 to 13.1 mm (mean, 3.9 ± 2.6 mm).

Table 1 Characteristics of patients with positive computed tomographic angiography (CTA) findings
Full size table

Fourteen patients with aneurysms detected by CTA subsequently underwent DSA, which confirmed all 14 aneurysms. No additional aneurysms were found by DSA. In the 14 patients who underwent DSA, 9 patients were treated with coil embolisation (n = 5) or clip ligation (n = 4) (Fig. 1).

Fig. 1
figure 1

A 57-year-old woman with acute severe headache. Unenhanced head CT revealed no haemorrhage (not shown). a Left anterior oblique image of CTA (volume-rendered technique image after automatic segmentation) shows the anterior communicating artery aneurysm (arrow). b DSA image confirms the presence of aneurysm (arrow). The aneurysm was successfully treated with clipping

Full size image

Discussion

Acute severe headache (also described as thunderclap headache, which refers to a sudden and severe headache with maximum intensity at onset) is the most common symptom of SAH. In previous studies, SAH has been detected in this type of headache with highly variable frequencies, ranging from 11 % to 71 % [2–15], which probably reflects differences in patient selection [15].

There has been considerable debate regarding the nomenclature of this term. In fact, the term “thunderclap headache” was originally coined to describe what was believed to be the presenting symptom of an unruptured cerebral aneurysm [16], and it was suggested that aneurysms may present with acute severe headache as a result of sudden expansion, luminal thrombosis or intramural haemorrhage [17]. This headache has been suspected to indicate a high risk of future rupture, because 20–50 % of patients with aneurysmal SAH have had a “warning” or “sentinel” headache preceding haemorrhage [7, 18, 19]. In addition to ruptured or unruptured aneurysms, similar sudden headaches with normal neurological examinations can be the presenting feature of cerebral venous sinus thrombosis, arterial dissection, hypertensive encephalopathy, pituitary apoplexy or spontaneous intracranial hypotension [20–24]. On the other hand, it is clear that in the absence of any organic intracranial pathological condition, sudden onset headache may occur as a benign and idiopathic headache disorder in most patients [12, 25, 26].

Acute severe headache is a clinical emergency that requires a swift evaluation to exclude SAH. It is generally recommended that all patients presenting with acute severe headache should be evaluated for the possibility of SAH with unenhanced head CT, and CSF examination (if needed). The reported accuracy of CT for detecting SAH is in the 98–99 % range [27] and even 100 % for fifth-generation systems [28]. However, the most appropriate diagnostic evaluation of patients presenting with acute severe headache without evidence of SAH is still controversial. A question of particular importance is whether further neurovascular imaging is required to definitively exclude a vascular lesion (mostly unruptured aneurysms) in patients with acute severe headache, negative CT and normal neurological examination [29–31].

Until now, DSA with selective cerebral arterial injections remains the diagnostic technique of choice for identifying intracranial vascular lesions. However, DSA has significant limitations for screening use in these patients because of its invasiveness, the need for hospitalisation and downstream financial costs. Furthermore, DSA has a 1 % risk of a disabling neurological complication and a 0.1 % risk of mortality [32].

There are several non-invasive alternative imaging techniques to DSA, including CTA and magnetic resonance angiography (MRA). CTA is a non-invasive, rapid and feasible diagnostic imaging technique that has proved valuable for identifying cerebrovascular lesions. It can easily be added to the routine unenhanced CT and can be easily repeated with the relatively non-invasive intravenous contrast material administration [33], thereby allowing an effective screening for vascular lesions. CTA has some disadvantages compared with MRA in that it requires an injection of iodine-based contrast material (which may cause idiosyncratic reactions or a deterioration in renal function in vulnerable groups) and it is associated with radiation exposure (typically ∼6 mSv in an effective dose, remarkably lower than that of DSA). The primary risk associated with radiation exposure is an increased risk of cancer and the degree of risk depends on the total amount of radiation dose received. However, the benefit of receiving a prompt and accurate diagnosis of serious diseases may outweigh the risk associated with radiation exposure. Often cerebral aneurysms remain undetected until they burst and the patient suffers a SAH with potentially life-threatening consequences. With advances in CTA, especially in the multidetector-row technique, larger body volumes can be examined within shorter time periods at high enough spatial resolution to provide good delineation of arterial/venous flow [34].

Results of previous studies [34–37] have suggested that CTA has high sensitivity and specificity for the detection of intracranial aneurysms. In a recent meta-analysis [36], CTA had a pooled sensitivity of 97.2 % and specificity of 97.9 % for detecting cerebral aneurysms on a per-patient basis (detection of at least one aneurysm per patient), which is more relevant for screening purposes. Aneurysm size has been an important factor in aneurysm detection, with studies of CTA consistently indicating the relatively low sensitivity of this technique in the detection of very small aneurysms [34–36]. McKinney et al. [35], in their series, performed with 64-channel CT, found that the sensitivity of CTA for aneurysms less than 4 mm was 92.3 %, whereas it was 100 % for those measuring 4–10 mm. Additionally, CTA was found to be useful in the diagnosis of other causes of headache, including arteriovenous malformation, arterial dissection, moyamoya disease and venous sinus thrombosis [38–40].

MRA is another non-invasive imaging method that can be used to screen for intracranial aneurysms in high-risk groups [41–44]. MRA has the advantages of lack of exposure to ionising radiation, no requirement for intravenous contrast material and large volume coverage. MRA, however, is not ideal because of the relatively high costs, long acquisition time and limited access for critically ill patients.

In this study, we used CTA as a diagnostic tool in patients with acute severe headache for intracranial vascular lesions including aneurysms that—once ruptured—are associated with a high risk of death or disability. We found a prevalence of 6.6 % for intracranial vascular lesions in patients with acute severe headache, negative CT and normal neurological examination. When only aneurysm was included, the prevalence decreased to 6.1 %. To our knowledge, only one study in the English literature has evaluated the use of CTA or MRA as a non-invasive diagnostic tool for cerebral aneurysms in patients with acute severe headache. Carstairs et al. [45] reported a prevalence of aneurysms of 4.3 % in their series of 116 patients with headache and negative CT.

There have been no studies on the prevalence of cerebral aneurysms in the general population with the use of CTA that could have been used as a comparison group. One could also compare the prevalence of detection in autopsy or angiography studies. According to autopsy studies, the overall prevalence of cerebral aneurysm in the general population is variable, but two large recent contemporary series [46, 47] suggested a prevalence of 2.1 %. The 2.1 % prevalence noted in these autopsy series includes both ruptured and unruptured aneurysms. If we had used the prevalence of unruptured aneurysms noted in those studies, the prevalence would have ranged between 0.4 % and 0.8 %. The prevalence of incidental aneurysms determined from vascular imaging in patients without SAH is close to the true value in the general population. In a meta-analysis of 2052 angiograms performed for brain tumours or other indications, the prevalence of cerebral aneurysms was 2.3 % [48]. The 6.1 % prevalence of unruptured aneurysms in our study group is almost 3-times the prevalence of 2.1–2.3 % in the general population, on the basis of available evidence from autopsy and angiographic studies [46–48]. Moreover, a few very small aneurysms undetected by CTA in our study could have been readily detected in the autopsy and angiography series; thus, the differences may be an underestimation.

There are several limitations of our study. First, our study did not perform a head-to-head comparison of CTA with the “gold standard”, DSA. The overall frequency of false positives (or false negatives) in our study remains unknown, because not all patients underwent DSA to confirm the presence or absence of vascular lesions. Second, we did not investigate the association between the presence of vascular lesions and potential risk factors such as family history, hypertension, alcohol use and smoking. Because of the retrospective nature of the analysis, no confirmation of complete information of the patients was possible.

In conclusion, CTA appears to be a feasible tool for diagnosing intracranial vascular lesions in patients with acute severe headache. The prevalence of vascular lesions in our series was 6.6 %, which is higher than that of the general population. We believe that our results provide a good evidence base for further evaluation in a large prospective study in future.