Journal of Medical Imaging and Radiation Oncology 59 (2015) 163–169

RADIOLO GY—O R I G I N A L A RT I C L E

bs_bs_banner

Magnetic resonance angiography or digital subtraction catheter angiography for follow-up of coiled aneurysms: Do we need both? Annah Lane,1 Philip Vivian1 and Alan Coulthard1,2 1 Department of Medical Imaging, Royal Brisbane and Women’s Hospital, Brisbane, Queensland, Australia 2 Academic Discipline of Medical Imaging, University of Queensland, Brisbane, Queensland, Australia

A Lane BSc MBBS; P Vivian BSc MBBS; A Coulthard BMedSci MBBS FRCSEd FRCR FRANZCR. Correspondence Dr Annah Lane, Radiology Registrar, Royal Brisbane and Women’s Hospital, Herston, Brisbane, Qld 4029, Australia. Email: [email protected] Conflict of interest: None. Submitted 25 August 2014; accepted 23 December 2014. doi:10.1111/1754-9485.12288

Abstract Introduction: Endovascular treatment of intracranial aneurysms is now common. Digital subtraction catheter angiography (DSA) is the gold standard for imaging follow-up of coiled intracranial aneurysms; however, it is an invasive procedure with a risk of complications. Time-of-flight magnetic resonance angiography (MRA) is an alternative non-invasive imaging method. In a previous study, we found that MRA at 1.5T was comparable to DSA for detection of aneurysm recurrence in this patient group and subsequently added MRA to the routine follow-up protocol. In the current study, we further compare MRA with DSA to determine whether MRA could safely replace DSA. Methods: Patients who had endovascular coiling procedure for intracranial aneurysm from 10/2004 to 6/2010 were identified from our database. A radiologist and a radiology registrar compared MRA and DSA for all patients who received both modalities. DSA was considered as the reference technique. ‘Agreement’ or ‘disagreement’ between modalities was noted regarding absence or presence of aneurysm recurrence. Results: The study group comprised 86 treatments of 80 aneurysms in 73 patients. There were 83/86 agreements between modalities and 3 disagreements. In one case, MRA identified a recurrence that was not seen on the corresponding DSA. In two cases, DSA showed a minor recurrence that was not seen on MRA. Conclusion: Of the two MRA ‘misses’, neither would have resulted in different management. MRA is a safe and accurate modality for follow-up of coiled aneurysms and can replace DSA. Key words: angiography; diagnostic imaging; digital subtraction; endovascular procedures; intracranial aneurysm; magnetic resonance angiography.

Introduction Intracranial aneurysms are an important health concern.1 The Guglielmi detachable coil introduced in 19912,3 was a breakthrough in the treatment of intracranial aneurysms4–6 and, since then, endovascular coiling has become a valid and popular treatment alternative to surgical clipping.4,7 However, the International Subarachnoid Aneurysm Trial found that coiled aneurysms were more likely to recur than clipped aneurysms.4 Follow-up imaging is necessary to evaluate the stability of aneurysm occlusion and the need for further

© 2015 The Royal Australian and New Zealand College of Radiologists

treatment.8,9 Most recurrences appear in the first 6–12 months post-treatment.10 Recurrence can form due to coil compaction or regrowth of a residual neck.11 Risk factors contributing to an increased rate of recurrence include rupture status, large aneurysm size, low fundus to neck ratio, low coil packing density, initial presence of an intraluminal thrombus and incomplete obliteration at time of initial treatment.11,12 Digital subtraction catheter angiography (DSA) is considered the gold standard for imaging follow-up of treated aneurysms.9 However, DSA is invasive, time consuming, expensive, exposes patients to ionising radiation and

163

A Lane et al.

Table 1. HR-MRA protocol

TR (milliseconds) TE (milliseconds) Excitations Flip angle (degrees) Matrix size Slices Field of view (centimetres) Voxel size (millimetres) Slice thickness (millimetres) Acquisition time

1.5T

3T

35 1.58 1 25 256 × 256 120 15 × 15 0.6 × 0.6 × 0.3 0.3 10 minutes, 46 seconds

23 4.2 1 15 256 × 256 112 150 0.6 × 0.6 × 0.3 0.3 7 minutes, 35 seconds

potentially nephrotoxic contrast media and is associated with the risk of complications, including access site haematoma and stroke.13 Time-of-flight magnetic resonance angiography (TOF-MRA) is non-invasive, avoids many of the risks inherent in the DSA procedure14,15 and has become a popular option for follow-up of coiled aneurysms.14–18 In a previous study we utilised a high resolution (HR) MRA protocol, modified from Yamada et al.19 (see Table 1) and compared it with a standard resolution MRA protocol and DSA in patients followed-up after coiling. We showed that (HR) MRA was comparable with DSA for detection of aneurysm recurrence.15 Following this study we changed our practice for aneurysm follow-up to MRA at 3 months and DSA at 6 months post-coiling. We felt that further prospective data were necessary before we could remove DSA from our follow-up protocol. The current study examines this period of prospective dual imaging follow-up to determine whether we could safely dispense with DSA in favour of MRA in these patients.

Methods The study design was a retrospective single-centre case review. Ethics approval was obtained from the hospital Human Research Ethics Committee. The study group comprised all patients who had an endovascular coiling procedure for intracranial aneurysm between October 2004 (when the change in imaging follow-up protocol was made) until June 2010. We sought patients who had had MRA follow-up within 6 months of the procedure, with subsequent DSA within 3 months. Patients were therefore excluded if they did not have both an MRA and DSA within a suitable time frame (MRA within 6 months and DSA within 12 months), if the MRA and DSA were performed more than 3 months apart, if the DSA was performed first, if they had follow-up at another facility or if they did not complete 6 months of standard follow-up for any reason. Where several pairs of imaging were available in patients with multiple follow-up episodes, only the initial pair of follow-up images was considered. All patients had a 3D TOF-MRA study obtained at 3T using a dedicated head array coil (Siemens

164

Magnetom Trio; Siemens, Erlangen, Germany; Table 1). Post-processing was performed using console software (MR B17) and an off-line workstation (Leonardo: Syngo VD10B, Siemens, Berlin, Germany). DSA was performed by an interventional neuroradiologist using a dual-plane angiographic system (Siemens Axiom Artis FA, Siemens, Frankfurt, Germany). Standard two-dimensional projections were acquired, with additional views including 3D rotational acquisition where indicated. Console software (Syngo VB11L) and a Leonardo workstation (Syngo VB30A) were used for DSA image post-processing. A neuroradiologist and a radiology registrar reviewed each case. The MRA studies were reviewed using both the available maximum intensity projections (MIPs) and the TOF source data. If discrepancies were found between the MIP images and the source data, the latter was reviewed on a reporting workstation using multiplanar reconstruction and MIP. Planar DSA images were reviewed and where available 3D rotational DSA image sets. For each case, the presence or absence of residual or aneurysm filling was assessed, and a consensus opinion was reached on whether the two modalities of imaging used for follow-up were considered to ‘agree’ or ‘disagree’ in terms of findings (no aneurysm recurrence vs. aneurysm recurrence or residual aneurysm).

Results Two hundred sixty-seven patients (182 women, 85 men, mean age 52.5, age range 6–82) with 287 aneurysms were treated with 314 endovascular coiling procedures between 13/10/2004 and 07/06/2010. After exclusions (194 patients excluded), 86 paired imaging episodes formed the study group representing 86 coiling procedures of 80 aneurysms in 73 patients. The study group comprised 49 female and 24 male patients. Aneurysms were located as follows: internal carotid artery (n = 21), anterior communicating artery (ACOM; n = 15), anterior cerebral artery (n = 5), middle cerebral artery (n = 5), posterior communicating artery (PCOM; n = 18), anterior choroidal artery (n = 1), posterior cerebral artery (n = 4), superior cerebellar artery

© 2015 The Royal Australian and New Zealand College of Radiologists

MRA vs. DSA for imaging coiled aneurysms

Fig. 1. Case 1: (a) MIP reformat of MRA performed 4 months post-coiling of a right PICA aneurysm showing a small neck recurrence (arrow). (b) DSA performed 6 months post-coiling showing a successfully coiled right PICA aneurysm (arrow) with no visible recurrence or residual aneurysm.

(n = 3), posterior inferior cerebellar artery (PICA; n = 2), basilar artery (n = 6). Mean maximum aneurysm diameter was 7.4 mm (median 6.2 mm, range 4.5– 20 mm). The mean time difference between the follow-up paired MRA and DSA investigations was 2.5 months (median 3 months, range 0–3 months). The mean time difference between coiling and initial followup MRA (for the purposes of recurrence detection, rather than another indication e.g. vasculitis/vasospasm) was 2.15 months (median 3 months, range 0.5–6 months), and the mean time difference between coiling and initial follow-up DSA was 3.83 months (median 6 months, range 2–8 months). For 83 (96.5%) of the 86 episodes of paired imaging follow-up the MRA and DSA findings were in agreement. Of these, 33/83 (39.8%) showed no evidence of recurrence on MRA or DSA. In 45/83 (54.2%) both MRA and DSA showed small areas of filling at the aneurysm neck, not requiring further treatment. In 5/83 (6%) both modalities showed neck recurrence that required repeat endovascular treatment. In three (3.5%) of 86 episodes a discrepancy between the MRA and DSA findings was identified regarding presence or absence of aneurysm recurrence. Of these, in one of three cases MRA revealed a small recurrence that was not identified on the corresponding DSA. In two of the three cases DSA identified a recurrence that was not seen on the MRA.

Case 1 A 56-year-old female presented with subarachnoid haemorrhage (SAH). A 6-mm ruptured aneurysm of the

right PICA was coiled acutely. MRA 4 months post-coiling revealed a small recurrence (Fig. 1a), not visualised on the DSA performed at 6 months post-coiling (Fig. 1b). The recurrence was still visible on another follow-up MRA performed at 20 months and was not seen on the corresponding DSA at 23 months. The recurrence has not required treatment.

Case 2 A 70-year-old female had coiling of a 9-mm ruptured ACOM aneurysm and a 7-mm unruptured PCOM aneurysm after presenting with SAH. An MRA at 4 months post-coiling showed no aneurysm recurrence for both aneurysms (Fig. 2a); however, a small neck recurrence at the ACOM site was visible on the DSA performed at 7 months post-coiling (Fig. 2b). The recurrence was visible on the subsequent MRA, performed at 17 months postcoiling and has been followed with MRA since (Fig. 2c). No further aneurysm treatment has been considered necessary.

Case 3 A 41-year-old male had a 9-mm right PCOM aneurysm coiled after presenting with a third cranial nerve palsy. Two small additional aneurysms were left untreated (left parophthalmic and carotid siphon). An MRA at 3 months post-coiling did not show a recurrence (Fig. 3a); however, the DSA at 6 months revealed a small neck recurrence (Fig. 3b). The recurrence was visible on the

Fig. 2. Case 2: (a) MIP reformat of MRA performed 4 months post-coiling. No recurrence was seen at the site of the coiled ACOM aneurysm. (b) DSA performed 7 months post-coiling revealed a small neck recurrence (arrow) at the coiled ACOM aneurysm. A successfully coiled PCOM aneurysm is also visible. (c) The recurrence of the coiled ACOM aneurysm is visible on MIP reformat of MRA performed 17 months post-coiling (arrow).

© 2015 The Royal Australian and New Zealand College of Radiologists

165

A Lane et al.

Fig. 3. Case 3: (a) MIP reformat of MRA performed 3 months post-coiling with no recurrence of the coiled right PCOM aneurysm visible. (b) DSA performed 6 months post-coiling that shows a recurrence at the site of the coiled right PCOM aneurysm (arrow). (c) MIP reformat of MRA performed 3 years post-coiling on which a recurrence is visible (arrows).

MRA performed 3 years post-coiling (Fig. 3c). No further aneurysm treatment has been performed to date, although treatment with a flow diverting stent is being considered. This patient had multiple aneurysms under surveillance, which meant that the follow-up plan in this case would likely be different to a patient with just the one treated aneurysm.

Discussion Comparisons of MRA and DSA for detection of residual/ recurrent aneurysms have been reported in the literature. Kwee20 performed a systematic review and metaanalysis in 2007 looking at diagnostic performance of MRA compared with DSA in the follow-up of intracranial aneurysms treated with Guglielmi detachable coils. Sixteen studies were included in the analysis, 14 of which investigated TOF-MRA and seven investigated contrast enhanced MRA (CE-MRA). Pooled sensitivity and specificity for TOF MRA for the detection of residual aneurysmal flow were 83.3% (95% confidence interval (CI) 70.3–91.3%) and 90.6% (95% CI 80.4–95.8%), and for CE-MRA were 86.8% (95% CI 71.4–94.5%) and 91.9% (95% CI 79.8–97.0%), respectively. The paper concluded that both TOF-MRA and CE-MRA may achieve a moderate to high diagnostic performance. However, the authors commented that the methodological quality of the studies was moderate and pooled estimates heterogeneous and that the results, therefore, should be interpreted with caution. Since the Kwee study, further reports have shown TOF-MRA to be an excellent imaging modality for follow-up of coiled aneurysms.14,16,17,21–23 Of those studies that looked specifically at whether MRA can replace DSA as the primary imaging modality for follow-up of intracranial aneurysm coiling,14,16,17,24 only one concluded that DSA could not be replaced.24 Deutschmann et al.,25 while investigating the effect of aneurysm size and location on diagnostic accuracy of TOF-MRA, found that TOF-MRA was worse than DSA for detection of recurrence if the aneurysm was