Acta Neurochir (2015) 157:351–359 DOI 10.1007/s00701-014-2287-2
CLINICAL ARTICLE - VASCULAR
Comparison of indocyanine green fluorescent angiography to digital subtraction angiography in brain arteriovenous malformation surgery Christopher J. Bilbao & Tarun Bhalla & Shamsher Dalal & Hiren Patel & Amir R. Dehdashti
Received: 25 August 2014 / Accepted: 18 November 2014 / Published online: 10 December 2014 # Springer-Verlag Wien 2014
Abstract Background The potential utility of intraoperative microscope-integrated indocyanine green (ICG) fluorescence angiography in the surgery of brain arteriovenous malformations (AVMs) and evaluation of the completeness of resection is debatable. Postoperative catheter angiography is considered the gold standard. We evaluated the value of ICG and intraoperative catheter angiography in this setting. Methods Between January 2009 and July 2013, 37 patients with brain AVMs underwent surgical resection of their vascular lesions. ICG videoangiography and an intraoperative catheter angiography were performed in 32 cases, and a routine postoperative angiogram was performed within 48 h to 2 weeks after surgery. The usefulness of ICG findings and the ability to confirm total resection and to identify residual
Presentations: Oral Presentation at the AANS-CV section: February 2014 Clinical Trial IRB no. 2013–0455 C. J. Bilbao (*) : H. Patel Department of Neurosurgery, Philadelphia College of Osteopathic Medicine, 4190 City Avenue, Philadelphia, PA 19131, USA e-mail: [email protected] H. Patel e-mail: [email protected] T. Bhalla : S. Dalal Department of Neurosurgery, Geisinger Medical Center, 100 N. Academy Ave, Danville, PA 17822, USA T. Bhalla e-mail: [email protected]
nidus or persistent shunt were assessed and compared to intraoperative and postoperative digital subtraction angiography, respectively. Results There were 7 grade 1, 11 grade 2, 11 grade 3 and 3 grade 4 Spetzler-Martin classification AVMs. ICG angiography helped to distinguish AVM vessels in 26 patients. In 31 patients, it demonstrated that there was no residual shunting. In one patient, a residual AVM was identified and further resected. Intraoperative catheter angiography detected two additional small residuals that were missed by ICG angiography, both deep in the surgical cavity. Further resection of the AVM was performed, and total resection was confirmed by a repeat intraoperative angiogram. Postoperative angiography in a patient with a grade 4 lesion revealed one additional small deep residual AVM nidus with persistent late shunting missed on both ICG and intraoperative angiography. Overall ICG angiography missed three out of four residual AVMs after initial resection, while the intraoperative angiogram missed one. Conclusion Although ICG angiography is a helpful adjunct in the surgery of some brain AVMs, it’s yield in detecting residual AVM nidus or shunt is low, especially for deepseated lesions and higher grade AVMs. ICG angiography should not be used as a sole and/or reliable technique. Highresolution postoperative angiography must be performed in brain AVM surgery and remains the best test to confidently confirm complete AVM resection. Keywords Brain arteriovenous malformation . Fluorescence angiography . Indocyanine green videoangiography . Catheter angiography . Intraoperative
S. Dalal e-mail: [email protected] A. R. Dehdashti Department of Neurosurgery, North Shore University Hospital, 300 Community Dr., Manhasset, NY 11030, USA e-mail: [email protected]
Introduction The usefulness of intraoperative imaging during cerebrovascular surgery has been concretely elucidated since its
352
inception over 2 decades ago. It allows for localization of vascular lesions as well as confirmation of obliteration and cure. The application of digital subtraction angiography (DSA) allows for a detailed, dynamic view of the arterial and venous systems. However, there are several clinical and logistic limitations to DSA. Recently, the application of fluorescent videoangiography for visualizing neurovascular structures has become a safe and effective alternative to DSA. The role of this technique in brain AVM surgery and its usefulness in determining the completeness of arteriovenous malformation (AVM) resection as compared to DSA has been studied previously [3, 8, 10–18]. Its validity and weaknesses however need further investigation as there are some debates regarding the reliability of ICG angiography in brain AVM surgery. The purpose of this study was to retrospectively review our experience with indocyanine green (ICG) fluorescent angiography and determine whether (1) it is helpful during AVM surgery and (2) moreover equivalent to DSA for the detection of residual AVM nidus or AVM shunting.
Materials and methods ICG dye Indocyanine green (ICG) is a tricarbocyanine dye that has a molecular weight of 751.4 Da. It is a negatively charged ion that is stable at room temperature and is soluble in water but not readily soluble in saline. It is evident in the arterial phase within 10–20 s after a venous injection. An important clinical property of ICG is its fast binding to plasma proteins, particularly lipoproteins, which makes repeat administrations possible [1].
Acta Neurochir (2015) 157:351–359
Oberkochen, Germany). During the surgical procedure when fluorescence was desired, ICG (at a concentration of 25 mg in 5 ml of saline) was administered intravenously by the anesthesiologist at the request of the surgeon. Vessel illuminescence was observed within 10 s, for a duration of 1–2 min, at which time the video was being recorded. The first instance of ICG angiography was performed after the dura had been opened and the AVM exposed in order to properly identify the angioarchitechture of the AVM and to identify the feeding arteries and draining veins. ICG angiography was repeated when the AVM was believed to be totally resected. If the AVM was found to have a residual nidus, further resection was performed, after which ICG angiography was repeated to confirm total obliteration. The patient was carefully observed by the anesthesia team for any adverse reactions to the administration. Digital subtraction angiography All 32 patients underwent both intra- and postoperative angiography to confirm gross total resection of the AVM nidus and absence of AVM shunting. Intraoperative angiograms were performed using fluoroscopy with a C-arm, and postoperative angiograms were performed in the biplanar angiography suite. Both intra- and postoperative angiography were performed by the same interventional neuroradiology team. Assessments were made based on preoperative evaluation of arterial feeders and draining veins. Intra- and postoperative angiograms were compared in order to determine any residual AVM nidus or AV shunts. Both the interventional neuroradiologist and neurosurgeon interpreted all angiograms together at the time of intervention. If indicated, repeat angiograms were performed.
Patients Data were collected on 37 consecutive AVM patients within a single institution from January 2009 to July 2013 (Table 1). The Internal Review Board of Geisinger Medical Center approved the study. All patients underwent surgical resection of their vascular lesion by the senior author (ARD). ICG angiography was performed before the resection and after the completion of AVM resection. An intraoperative angiogram was performed after the last ICG angiography at the end of the resection. A postoperative angiogram was performed within 48 h to 2 weeks after surgery. Five patients in whom intraoperative ICG angiography, intraoperative DSA and/or postoperative DSA were not performed were excluded, leaving 32 patients included in the study. Fluorescent angiography All procedures were performed using a near-infraredequipped videoangiography microscope (Carl Zeiss Co.,
Results For the 32 patients included in the study, the mean age was 45 years (range: 2–75). There were 17 males (53 %) and 15 females (47 %). Fourteen of 32 (44 %) patients were neurologically completely intact (Table 1). There were 7 SpetzlerMartin grade 1 AVMs, 11 grade 2, 11 grade 3 and 3 grade 4 AVMs. Nineteen of 32 (59 %) AVMs were ruptured on presentation. Eighteen patients (56 %) underwent partial Onyx embolization prior to the planned surgical procedure. Intraoperative ICG Intraoperative ICG was performed in all 32 cases before and after AVM resection. No complications were encountered from the ICG dye administration. Nidus identification by ICG was determined to be helpful by the surgeon if it allowed for accurate visualization of the AVM nidus,
No No
No No
Yes
Yes
Yes Yes
(II) (II) (II) (II) (II) (II) (II) (II) (II) (II) (II) (III)
(III)
8 (F, 53) 9 (M, 28) 10 (M, 53) 11 (F, 43) 12 (M, 38) 13 (M, 68) 14 (F, 49) 15 (F, 18) 16 (F, 45) 17 (M, 38) 18 (M, 72) 19 (M, 63)
20 (M, 2)
21 (F, 58) (III) 22 (M, 42) (III)
23 (M, 39) (III) 24 (F, 62) (III)
25 (M, 26) (III)
26 (M, 75) (III)
27 (F, 59) 28 (F, 33)
(III) (III)
Yes
(I) (I)
6 (M, 50) 7 (M, 51)
No No No No No No Yes Yes Yes Yes Yes No
No No
No No
Yes Yes
Yes
Yes
Yes Yes
No Yes
No
No No No Yes Yes No No No No No No No
No No
No No
Yes Yes
Yes
Yes
No Yes
No Yes
No
Yes Yes No Yes Yes Yes No Yes No No Yes Yes
Yes Yes
Yes No
None (seizure) Yes (left-sided weakness and hemianopia) Yes (left-sided weakness, hemianopia, in pregnancy) Yes (cerebellar deficit, hydrocephalus) Yes (gait and cerebellar deficit) Yes (weakness and decreased mental status)
Yes (mild aphasia) None (hemorrhage) None (seizure) Yes (ataxia, hemorrhage) Yes (hydrocephalus) Yes (seizure, sensory deficit) None (seizure) Yes (hemiparesis) None (headache) None (seizure) Yes (mild aphasia) None (seizure, sensory symptoms) Yes (right-sided weakness), drowsiness None (headache, seizure) Yes (left-sided weakness)
Yes (hemianopia) None (seizure)
Yes (aphasia) None (seizure)
None (seizure) None (headache) None (incidental)
(I) (I)
No No No
4 (F, 62) 5 (F, 48)
No No No
(I) (I) (I)
1 (M, 42) 2 (F, 36) 3 (M, 50)
No No No
AVM grade Deep location Deep Ruptured Neurological deficit (S-M grading) (or some deep venous preoperative/ presentation components) drainage
Patient population and characteristics
Patient sex/age
Table 1
Yes/yes Yes/yes
Yes/yes
No/no
Yes/yes No/yes
Yes/yes No/no
Yes/yes
Yes/yes Yes/yes Yes/yes Yes/yes Yes/yes No/no Yes/yes Yes/yes Yes/yes Yes/yes Yes/yes Yes/yes
Yes/yes Yes/yes
Yes/no Yes/yes
Yes/ yes No/yes Yes/yes
No residual/Y No residual/N
No residual/Y
No residual/N
No residual/Y No residual/N
No residua/Y 1 residual shunt/N
No residual /N
No residual/N No residual/N No residual/Y No residual/N No residual/N No residual/N No residual/N No residual/N No residual/Y No residual/N No residual/N No residual/Y
No residua/N No residual/Y
No residual/N No residual/N
No residual/ N No residual/N No residual/N
No residual No residual
No residual
No residual
1 residual shunt (further resected) No residual No residual after further resection No residual No residual
No residual No residual No residual No residual No residual No residual No residual No residual No residual No residual No residual No residual
No residual No residual
No residual No residual
No residual No residual No residual
Intraoprtative ICGIntraoperative ICG Intraoperative AVM configuration residual?/en passage angio residual? helpful?*/1 or multiple vessel (yes/no) EDV identification
No residual No residual
No residual
No residual No residual done at 3 months# No residual
No residual No residual
No residual
No residual No residual No residual No residual No residual No residual No residual No residual No residual No residual No residual No residual
No residual No residual Done at 2 years no residual# No residual Done at 2 months no residual No residual No residual
Postoperative angio residual?
1 3
1
0
0 2
1 1
1
1 0 0 1 0 1 0 0 0 0 1 0
1 0
0 0
0 0 0
Outcome mRS
Acta Neurochir (2015) 157:351–359 353
No residual/N No residual/N Yes/yes Yes/yes No No (IV) (IV)
Yes Yes Yes Yes 31 (F, 27) 32 (F, 24)
arterial feeders, venous drainage and peripheral extension. In this regard, ICG was found to be useful in 26 patients (81 %). ICG was also helpful in determining en passage vessels from malformation-based pathological vessels in 9 patients out of 16 (56 %) in whom the preoperative angiogram had clearly identified an en passage vessel. In three cases with a very small subcortical nidus not clearly distinguishable at the surface, the ICG was able to illustrate the location of the nidus based on robust surface AV shunting. In only one case, the ICG demonstrated a small residual nidus, which was further resected and confirmed obliterated by repeat ICG and subsequently intraoperative DSA.
Intraoperative DSA Intraoperative digital subtraction angiography was performed in all 32 patients. There were no related complications. Thirty of the 32 patients did not have any residual AVM nidus or AV shunt. In two cases, a small residual nidus with an early deepseated shunt was identified. One of the patients had a grade 3 and the other a grade 4 Spetzler-Martin AVM. Further resections were performed and subsequent repeat intraoperative angiograms confirmed complete obliteration. #Three of the postoperative angiograms were performed outside the time frame of 48 h to 2 weeks
3
1 residual shunt No residual (further resected) No residual No residual No residual AV shunt residual No residual/Y Yes/no Yes Yes Yes 30 (M, 45) (IV)
*ICG helpful means the surgeon’s perception of being able to distinguish arterial (feeding arteries) from venous structures (early draining vein) and also able to distinguish the nidus peripheral extension and identify en passage vessels
2 No residual No residual No residual/N No/no Yes Yes Yes (III) 29 (F, 28)
Yes (hydrocephalus, cerebellar deficit, left-sided weakness) Yes (hemiparesis, decreased mental status) None (headache and ITP) Yes (seizure, left-sided sensory deficit)
Postoperative angio residual? Intraoprtative ICGIntraoperative ICG Intraoperative AVM configuration residual?/en passage angio residual? helpful?*/1 or multiple vessel (yes/no) EDV identification AVM grade Deep location Deep Ruptured Neurological deficit (S-M grading) (or some deep venous preoperative/ presentation components) drainage Patient sex/age
Table 1 (continued)
0 0
Acta Neurochir (2015) 157:351–359 Outcome mRS
354
Postoperative DSA All 32 patients had postoperative DSA. There were no related complications. Thirty-one of the 32 patients were found to have negative angiograms in terms of the AVM nidus or AV shunt. One patient with a grade 4 AVM was found to harbor residual deep-seated AV shunting, missed on both ICG angiography and intraoperative DSA. Overall, ICG angiography missed three out of four residual AVMs after initial resection, and the intraoperative angiogram missed one. Postoperative DSA confirmed preservation of en passage vessels in the nine patients in whom the ICG had identified the en passage vessel before surgical resection (Table 2).
Illustrative case 1 A 38-year-old female presents with new-onset seizure. Investigations revealed a right temporal Spetzler-Martin grade 2 AVM (Fig. 1a), with an en passage angular artery identified on the preoperative DSA (Fig. 1b). Intraoperative ICG angiography helped to distinguish the en passage vessel from the aberrant AVM-based vessels (Fig. 1c and d). The AVM was resected without residual nidus or shunts confirmed with ICG angiography, intra- and postoperative DSA. The en passage vessel was preserved (Fig. 1e).
Acta Neurochir (2015) 157:351–359 Table 2
355
Summary of relevant intraoperative ICG findings compared to the intra- and postoperative angiogram
Identification of en passage vessel intraoperative/en passage vessel identified on preoperative angiogram ICG helpful in identification of AVM angioarchitecture Missed residual nidus/total number of residual nidus
Intraoperative ICG
Intraoperative DSA
Postoperative DSA
9/16
Confirmation of preservation of 9 en passage vessels N/A 1/4
Confirmation of preservation of 9 en passage vessels N/A N/A
26/32 3/4
Illustrative case 2
Illustrative case 3
A 24-year-old female presented with intractable seizures and was found to have a large right parietal grade 4 AVM (Fig. 2a). Preoperative partial embolization with Onyx (Microtherapeutics, Irvine, CA) was performed. The patient then underwent surgical resection and both ICG angiography and intraoperative DSA failed to show a residual nidus/shunt (Fig. 2b). The postoperative angiography however revealed a deep-seated residual shunt to the vein of Galen with a very small nidus attached to it (Fig. 2c).
A 42-year-old male presented with a parietal hemorrhage and left-sided weakness. He was found to have a right temporal grade 3 AVM (Fig. 3a, b and c). The patient underwent surgical resection. Once it was felt, the AVM was completely resected; an ICG angiogram was done that demonstrated a residual nidus within the surgical field not clearly seen with gross microscopic inspection (Fig. 3d and e). A further resection was undertaken to achieve total removal confirmed with an intra- and postoperative DSA (Fig. 3f).
Fig. 1 a AP view of a right temporal AVM, Spetzler-Martin grade 2. b Lateral view of a right temporal AVM with visualization of an en passage vessel, SMII. c Intraoperative view of the nidus limits, draining veins and en passage vessel. d ICG view of the nidus, draining veins and en passage vessel. e Postoperative DSA demonstrating obliteration of the right temporal AVM and preservation of the en passage vessel
356
Acta Neurochir (2015) 157:351–359
Fig. 2 a AP view of the right parietal Spetzler-Martin grade 4 AVM. b ICG (not shown) and subsequently intraoperative DSA demonstrating obliteration of the right parietal AVM. c Postoperative
DSA showing a deep residual shunt with persistent early drainage to the vein of Galen missed by ICG and intraoperative angiography
Outcome
majority of patients and to distinguish en passage vessels from abnormal vessels in a subset of patients. This could potentially contribute to a reduction in morbidity from accidental sacrifice of an en passage vessel. In our study, the en passage vessel was recognized and preserved at surgery in 9 of the 16 patients in whom the vessel was identified on preoperative angiogram. This further emphasizes the importance of maintaining physiological vessel integrity to achieve a better neurological outcome as none of those patients had new neurological deficits postoperatively. Although one could argue that en passage vessels can be identified without ICG injection, we believe that the dye injection will help to confirm the vessel detection in a significant subset of patients. Hänggi et al. [11] reported on a series of 15 patients with AVM where ICG angiography was used during the surgery. They emphasized that the most frequent surgery-related complications resulted from injury to arterial branches of normal brain and incomplete resection of the nidus. Therefore, the accurate detection of a traversing branch of normal brain vasculature ensures a higher likelihood of preservation of that vessel. In their series, they concluded that ICG was able to identify en passage vessels and very useful in the detection of the residual nidus. The latter findings are in discordance with our results. Taddei et al. [16] reached the same conclusion based on a series of nine patients and even suggested that ICG can replace intra-/postoperative DSA. Takagi et al. [17] also reported on a case of a 2-yearold female where ICG was superior to postoperative angiography in detecting residual disease. Based on our results, we conclude that ICG angiography is indeed of value in smaller, lower grade and superficial AVMs where the integrity of the AVM and surgical resection cavity is readily accessible for ICG image acquisition. ICG videoangiography can only depict what is seen directly within the surgical field. Because of this premise, there are certain conditions in which the residual lesion may not be evident,
Twenty-nine of the 32 patients did not experience any new complication or postoperative deficit. One patient had a new postoperative hemianopia, which improved to quadrantanopia at follow-up. One patient experienced a new hemisensory deficit, which almost fully recovered. One patient had a bone flap infection. Outcome modified Rankin score (mRS) analysis revealed a score of 0 for 18 patients, 1 for 10, 2 for 2 and 3 for the other 2 patients (Table 1). There was no mortality in this series.
Discussion The use of intraoperative DSA for AVM surgery has documented benefits and become an important adjunct for confirmation of total obliteration [2, 3, 5, 6, 8, 10–18]. Fluorescent angiography with intraoperative visualization of cerebral arteries has achieved widespread use because of its ease of utilization, rapid delivery and patient safety [3–6, 9, 13]. The technique is safe, requires no additional equipment except for microscope software modification and can be administered multiple times throughout a surgical procedure. Most importantly, it potentially integrates vital information in real time in order to direct the surgeon as to the next course of action. For these reasons, it has become a recent contender in displacing DSA as the procedure of choice for detecting residual aneurysms [3, 4, 14] or AVMs intraoperatively. The present study is the second largest series of patients with AVM who had evaluation of intraoperative ICG angiography compared to intra- and postoperative catheter angiography. The results of our study demonstrate the ability of ICG angiography to elucidate AVM angioarchitecture in the
Acta Neurochir (2015) 157:351–359
357
Fig. 3 a AP view of the right temporal Spetzler-Martin grade 3 AVM. b Lateral view of the right temporal Spetzler-Martin grade 3 AVM. c Intraoperative view of the right temporal region showing the surface representation of the AVM. d Intraoperative view of the resection cavity demonstrating no gross residual AVM nidus. e ICG angiogram demonstrating a residual AVM nidus in an area that did not appear to have a residual nidus on gross inspection. f Intraoperative DSA demonstrating gross total resection of the AVM nidus
e.g., if it is outside of the visual field, deep in the surgical cavity or covered by brain tissue because of poor illumination. In our study, the ICG failed to detect a residual AV shunt in one Spetzler-Martin grade 3 and two grade 4 AVMs. This acquiesces with the recent criticism of ICG angiography that deeper, more extensive AVMs are discerned with less specificity by fluorescent angiography than digital subtraction angiography [3, 12, 13, 18]. The results of our study mimic the results of Killory et al. [12] as well as a very recent investigation by the same institution. In their initial study of ten consecutive patients, ICG angiography clearly revealed AV shunting and helped discern feeding arteries, draining veins and en passage vessels. In two
patients ICG failed to identify residual AVM and persistent early venous drainage that were subsequently detected by DSA. Clear advantages of ICG angiography were identified, including the speed with which it can be performed, the low cost associated with it and the low to no morbidity associated with performing the procedure. In their latest investigation of 130 cases reported over a 4-year period, Zaidi et al. [18] evaluated whether ICG performed in patients undergoing AVM surgery improves rates of resection and clinical outcome. They pre-selected the subgroup with superficial or lobar AVMs for whom they used ICG and another subgroup with no ICG. Both ICG and non-ICG groups had either intraor postoperative angiography performed to determine residual
358
AVMs and the need for revision or second surgery. As mentioned above, ICG was used mainly when the AVM nidus was close to the cortical surface or lobar. Approximately 14 % of patients had residual disease and reoperation rates, and mRS scores were not different between the two groups. They concluded that ICG may be beneficial in small superficial AVMs; however, it does not enhance the identification of residual disease or improve clinical outcomes for more complex and deeper lesions. In contrast to their study, all patients in our group (including superficial, lobar or deep AVMs) had the ICG technique evaluated; therefore, the weakness of ICG for deep and larger AVMs was directly demonstrated. With regard to the intraoperative angiogram, our series underlines the importance of high-quality angiography as one of the residual AVMs in this group was missed by standard intraoperative angiography using the C-arm. Ideally, a high-quality high-resolution angiogram in a hybrid endovascular neurosurgery suite can offer the same advantage as a routine postoperative angiogram [7]. This will not only increase the reliability of intraoperative assessment of AVM resection, but also obviates the need for a postoperative angiogram and potential second treatment if a residual is found after the operation. The timing of postoperative DSA is very crucial. We ideally do the postoperative DSA during the same admission, and, as described above, this was done between 48 h to 2 weeks. Except for three cases in which the postoperative angiogram was performed after 2 weeks, they all had postoperative DSA during the same admission. While there is a minute risk of a tiny residual AVM being missed by postoperative DSA, we believe this risk is minimal, and we suggest one more follow-up angiogram at 1–3 years for all patients with resected AVM to address that minute possibility. There is also the risk of false-positive findings, which is mainly a dysplastic neovascular network of feeding arteries around the resected nidus and which in general disappears after a few weeks [15]. In the only case in our series where a residual was found postoperatively, the residual was clearly a retained nidus with a draining vein and could not be attributed to a false-positive finding. Our report is subject to all the limitations and biases of a retrospective analysis. Although it confirms some of the benefits of ICG angiography and illuminates a distinct advantage over other techniques to identify nonpathological en passage vessels in brain AVM surgery, it also highlights its major limitations and suggests caution in the interpretation of important intraoperative findings based solely on ICG angiography.
Conclusion Although ICG angiography is a helpful adjunct in the surgery of small and superficial AVMs. Its yield in detecting residual
Acta Neurochir (2015) 157:351–359
AVM nidus or shunt is low, especially for deep-seated lesions and higher grade AVMs. Therefore, ICG angiography should not be used as a sole and/or reliable technique. A highresolution postoperative angiogram must be performed in brain AVM surgery and remains the best test to confidently confirm AVM total resection.
Conflicts of interest None.
References 1. Alander JT, Kaartinen I, Laakso A, Patila T, Spillmann T, Tuchin VV, Venermo M, Välisuo P (2012) A review of indocyanine green fluorescent imaging in surgery. Int J Biomed Imaging 2012:940585 2. Anegawa S, Hayashi T, Torigoe R, Harada K, Kihara S (1994) Intraoperative angiography in the resection of arteriovenous malformations. J Neurosurg 80:73–78 3. Balamurugan S, Agrawal A, Kato Y, Sano H (2011) Intra operative indocyanine green video-angiography in cerebrovascular surgery: an overview with review of literature. Asian J Neurosurg 6:88–93 4. Barrow DL, Boyer KL, Joseph GJ (1992) Intraoperative angiography in the management of neurovascular disorders. Neurosurgery 30: 153–159 5. Bauer BL (1984) Intraoperative angiography in cerebral aneurysm and AV-malformation. Neurosurg Rev 7:209–217 6. Dashti R, Laakso A, Niemela M, Porras M, Hernesniemi J (2009) Microscope-integrated near-infrared indocyanine green videoangiography during surgery of intracranial aneurysms: the Helsinki experience. Surg Neurol 71:543–550, discussion 550 7. Dehdashti AR, Thines L, Da Costa LB, terBrugge KG, Willinsky RA, Wallace MC, Tymianski M (2009) Intraoperative biplanar rotational angiography during neurovascular surgery. Technical note. J Neurosurg 111:188–192 8. Faber F, Thon N, Fesl G, Rachinger W, Guckler R, Tonn JC, Schichor C (2011) Enhanced analysis of intracerebral arterioveneous malformations by the intraoperative use of analytical indocyanine green videoangiography: technical note. Acta Neurochir 153:2181– 2187 9. Feindel W, Yamamoto YL, Hodge CP (1967) Intracarotid fluorescein angiography: a new method for examination of the epicerebral circulation in man. Can Med Assoc J 96:1–7 10. Ghosh S, Levy ML, Stanley P, Nelson M, Giannotta SL, McComb JG (1999) Intraoperative angiography in the management of pediatric vascular disorders. Pediatr Neurosurg 30:16–22 11. Hanggi D, Etminan N, Steiger HJ (2010) The impact of microscopeintegrated intraoperative near-infrared indocyanine green videoangiography on surgery of arteriovenous malformations and dural arteriovenous fistulae. Neurosurgery 67:1094–1103, discussion 1103–1094 12. Killory BD, Nakaji P, Gonzales LF, Ponce FA, Wait SD, Spetzler RF (2009) Prospective evaluation of surgical microscope-integrated intraoperative near-infrared indocyanine green angiography during cerebral arteriovenous malformation surgery. Neurosurgery 65:456– 462, discussion 462 13. Ng YP, King NK, Wan KR, Wang E, Ng I (2013) Uses and limitations of indocyanine green videoangiography for flow analysis in arteriovenous malformation surgery. J Clin Neurosci 20:224–232 14. Raabe A, Nakaji P, Beck J, Kim LJ, Hsu FP, Kamerman JD, Seifert V, Spetzler RF (2005) Prospective evaluation of surgical microscopeintegrated intraoperative near-infrared indocyanine green
Acta Neurochir (2015) 157:351–359 videoangiography during aneurysm surgery. J Neurosurg 103:982– 989 15. Solomon RA, Connolly ES Jr, Prestigiacomo CJ, Khandji AG, PileSpellman J (2000) Management of residual dysplastic vessels after cerebral arteriovenous malformation resection: implications for postoperative angiography. Neurosurgery 46:1052–1060 16. Taddei G, Tommasi CD, Ricci A, Galzio RJ (2011) Arteriovenous m a l f o r m a t i o n s a n d i n t r a o p e r a t i v e in d o c y an i n e g r e e n videoangiography: preliminary experience. Neurol India 59:97–100
359 17. Takagi Y, Kikuta K, Nozaki K, Sawamura K, Hashimoto N (2007) Detection of a residual nidus by surgical microscope-integrated intraoperative near-infrared indocyanine green videoangiography in a child with a cerebral arteriovenous malformation. J Neurosurg 107: 416–418 18. Zaidi HA, Abla AA, Nakaji P, Chowdhry SA, Albuquerque FC, Spetzler RF (2014) Indocyanine green angiography in the surgical management of cerebral arteriovenous malformations: lessons learned in 130 consecutive cases. Neurosurgery 10(Suppl 2):246–251
CLINICAL ARTICLE - VASCULAR
Comparison of indocyanine green fluorescent angiography to digital subtraction angiography in brain arteriovenous malformation surgery Christopher J. Bilbao & Tarun Bhalla & Shamsher Dalal & Hiren Patel & Amir R. Dehdashti
Received: 25 August 2014 / Accepted: 18 November 2014 / Published online: 10 December 2014 # Springer-Verlag Wien 2014
Abstract Background The potential utility of intraoperative microscope-integrated indocyanine green (ICG) fluorescence angiography in the surgery of brain arteriovenous malformations (AVMs) and evaluation of the completeness of resection is debatable. Postoperative catheter angiography is considered the gold standard. We evaluated the value of ICG and intraoperative catheter angiography in this setting. Methods Between January 2009 and July 2013, 37 patients with brain AVMs underwent surgical resection of their vascular lesions. ICG videoangiography and an intraoperative catheter angiography were performed in 32 cases, and a routine postoperative angiogram was performed within 48 h to 2 weeks after surgery. The usefulness of ICG findings and the ability to confirm total resection and to identify residual
Presentations: Oral Presentation at the AANS-CV section: February 2014 Clinical Trial IRB no. 2013–0455 C. J. Bilbao (*) : H. Patel Department of Neurosurgery, Philadelphia College of Osteopathic Medicine, 4190 City Avenue, Philadelphia, PA 19131, USA e-mail: [email protected] H. Patel e-mail: [email protected] T. Bhalla : S. Dalal Department of Neurosurgery, Geisinger Medical Center, 100 N. Academy Ave, Danville, PA 17822, USA T. Bhalla e-mail: [email protected]
nidus or persistent shunt were assessed and compared to intraoperative and postoperative digital subtraction angiography, respectively. Results There were 7 grade 1, 11 grade 2, 11 grade 3 and 3 grade 4 Spetzler-Martin classification AVMs. ICG angiography helped to distinguish AVM vessels in 26 patients. In 31 patients, it demonstrated that there was no residual shunting. In one patient, a residual AVM was identified and further resected. Intraoperative catheter angiography detected two additional small residuals that were missed by ICG angiography, both deep in the surgical cavity. Further resection of the AVM was performed, and total resection was confirmed by a repeat intraoperative angiogram. Postoperative angiography in a patient with a grade 4 lesion revealed one additional small deep residual AVM nidus with persistent late shunting missed on both ICG and intraoperative angiography. Overall ICG angiography missed three out of four residual AVMs after initial resection, while the intraoperative angiogram missed one. Conclusion Although ICG angiography is a helpful adjunct in the surgery of some brain AVMs, it’s yield in detecting residual AVM nidus or shunt is low, especially for deepseated lesions and higher grade AVMs. ICG angiography should not be used as a sole and/or reliable technique. Highresolution postoperative angiography must be performed in brain AVM surgery and remains the best test to confidently confirm complete AVM resection. Keywords Brain arteriovenous malformation . Fluorescence angiography . Indocyanine green videoangiography . Catheter angiography . Intraoperative
S. Dalal e-mail: [email protected] A. R. Dehdashti Department of Neurosurgery, North Shore University Hospital, 300 Community Dr., Manhasset, NY 11030, USA e-mail: [email protected]
Introduction The usefulness of intraoperative imaging during cerebrovascular surgery has been concretely elucidated since its
352
inception over 2 decades ago. It allows for localization of vascular lesions as well as confirmation of obliteration and cure. The application of digital subtraction angiography (DSA) allows for a detailed, dynamic view of the arterial and venous systems. However, there are several clinical and logistic limitations to DSA. Recently, the application of fluorescent videoangiography for visualizing neurovascular structures has become a safe and effective alternative to DSA. The role of this technique in brain AVM surgery and its usefulness in determining the completeness of arteriovenous malformation (AVM) resection as compared to DSA has been studied previously [3, 8, 10–18]. Its validity and weaknesses however need further investigation as there are some debates regarding the reliability of ICG angiography in brain AVM surgery. The purpose of this study was to retrospectively review our experience with indocyanine green (ICG) fluorescent angiography and determine whether (1) it is helpful during AVM surgery and (2) moreover equivalent to DSA for the detection of residual AVM nidus or AVM shunting.
Materials and methods ICG dye Indocyanine green (ICG) is a tricarbocyanine dye that has a molecular weight of 751.4 Da. It is a negatively charged ion that is stable at room temperature and is soluble in water but not readily soluble in saline. It is evident in the arterial phase within 10–20 s after a venous injection. An important clinical property of ICG is its fast binding to plasma proteins, particularly lipoproteins, which makes repeat administrations possible [1].
Acta Neurochir (2015) 157:351–359
Oberkochen, Germany). During the surgical procedure when fluorescence was desired, ICG (at a concentration of 25 mg in 5 ml of saline) was administered intravenously by the anesthesiologist at the request of the surgeon. Vessel illuminescence was observed within 10 s, for a duration of 1–2 min, at which time the video was being recorded. The first instance of ICG angiography was performed after the dura had been opened and the AVM exposed in order to properly identify the angioarchitechture of the AVM and to identify the feeding arteries and draining veins. ICG angiography was repeated when the AVM was believed to be totally resected. If the AVM was found to have a residual nidus, further resection was performed, after which ICG angiography was repeated to confirm total obliteration. The patient was carefully observed by the anesthesia team for any adverse reactions to the administration. Digital subtraction angiography All 32 patients underwent both intra- and postoperative angiography to confirm gross total resection of the AVM nidus and absence of AVM shunting. Intraoperative angiograms were performed using fluoroscopy with a C-arm, and postoperative angiograms were performed in the biplanar angiography suite. Both intra- and postoperative angiography were performed by the same interventional neuroradiology team. Assessments were made based on preoperative evaluation of arterial feeders and draining veins. Intra- and postoperative angiograms were compared in order to determine any residual AVM nidus or AV shunts. Both the interventional neuroradiologist and neurosurgeon interpreted all angiograms together at the time of intervention. If indicated, repeat angiograms were performed.
Patients Data were collected on 37 consecutive AVM patients within a single institution from January 2009 to July 2013 (Table 1). The Internal Review Board of Geisinger Medical Center approved the study. All patients underwent surgical resection of their vascular lesion by the senior author (ARD). ICG angiography was performed before the resection and after the completion of AVM resection. An intraoperative angiogram was performed after the last ICG angiography at the end of the resection. A postoperative angiogram was performed within 48 h to 2 weeks after surgery. Five patients in whom intraoperative ICG angiography, intraoperative DSA and/or postoperative DSA were not performed were excluded, leaving 32 patients included in the study. Fluorescent angiography All procedures were performed using a near-infraredequipped videoangiography microscope (Carl Zeiss Co.,
Results For the 32 patients included in the study, the mean age was 45 years (range: 2–75). There were 17 males (53 %) and 15 females (47 %). Fourteen of 32 (44 %) patients were neurologically completely intact (Table 1). There were 7 SpetzlerMartin grade 1 AVMs, 11 grade 2, 11 grade 3 and 3 grade 4 AVMs. Nineteen of 32 (59 %) AVMs were ruptured on presentation. Eighteen patients (56 %) underwent partial Onyx embolization prior to the planned surgical procedure. Intraoperative ICG Intraoperative ICG was performed in all 32 cases before and after AVM resection. No complications were encountered from the ICG dye administration. Nidus identification by ICG was determined to be helpful by the surgeon if it allowed for accurate visualization of the AVM nidus,
No No
No No
Yes
Yes
Yes Yes
(II) (II) (II) (II) (II) (II) (II) (II) (II) (II) (II) (III)
(III)
8 (F, 53) 9 (M, 28) 10 (M, 53) 11 (F, 43) 12 (M, 38) 13 (M, 68) 14 (F, 49) 15 (F, 18) 16 (F, 45) 17 (M, 38) 18 (M, 72) 19 (M, 63)
20 (M, 2)
21 (F, 58) (III) 22 (M, 42) (III)
23 (M, 39) (III) 24 (F, 62) (III)
25 (M, 26) (III)
26 (M, 75) (III)
27 (F, 59) 28 (F, 33)
(III) (III)
Yes
(I) (I)
6 (M, 50) 7 (M, 51)
No No No No No No Yes Yes Yes Yes Yes No
No No
No No
Yes Yes
Yes
Yes
Yes Yes
No Yes
No
No No No Yes Yes No No No No No No No
No No
No No
Yes Yes
Yes
Yes
No Yes
No Yes
No
Yes Yes No Yes Yes Yes No Yes No No Yes Yes
Yes Yes
Yes No
None (seizure) Yes (left-sided weakness and hemianopia) Yes (left-sided weakness, hemianopia, in pregnancy) Yes (cerebellar deficit, hydrocephalus) Yes (gait and cerebellar deficit) Yes (weakness and decreased mental status)
Yes (mild aphasia) None (hemorrhage) None (seizure) Yes (ataxia, hemorrhage) Yes (hydrocephalus) Yes (seizure, sensory deficit) None (seizure) Yes (hemiparesis) None (headache) None (seizure) Yes (mild aphasia) None (seizure, sensory symptoms) Yes (right-sided weakness), drowsiness None (headache, seizure) Yes (left-sided weakness)
Yes (hemianopia) None (seizure)
Yes (aphasia) None (seizure)
None (seizure) None (headache) None (incidental)
(I) (I)
No No No
4 (F, 62) 5 (F, 48)
No No No
(I) (I) (I)
1 (M, 42) 2 (F, 36) 3 (M, 50)
No No No
AVM grade Deep location Deep Ruptured Neurological deficit (S-M grading) (or some deep venous preoperative/ presentation components) drainage
Patient population and characteristics
Patient sex/age
Table 1
Yes/yes Yes/yes
Yes/yes
No/no
Yes/yes No/yes
Yes/yes No/no
Yes/yes
Yes/yes Yes/yes Yes/yes Yes/yes Yes/yes No/no Yes/yes Yes/yes Yes/yes Yes/yes Yes/yes Yes/yes
Yes/yes Yes/yes
Yes/no Yes/yes
Yes/ yes No/yes Yes/yes
No residual/Y No residual/N
No residual/Y
No residual/N
No residual/Y No residual/N
No residua/Y 1 residual shunt/N
No residual /N
No residual/N No residual/N No residual/Y No residual/N No residual/N No residual/N No residual/N No residual/N No residual/Y No residual/N No residual/N No residual/Y
No residua/N No residual/Y
No residual/N No residual/N
No residual/ N No residual/N No residual/N
No residual No residual
No residual
No residual
1 residual shunt (further resected) No residual No residual after further resection No residual No residual
No residual No residual No residual No residual No residual No residual No residual No residual No residual No residual No residual No residual
No residual No residual
No residual No residual
No residual No residual No residual
Intraoprtative ICGIntraoperative ICG Intraoperative AVM configuration residual?/en passage angio residual? helpful?*/1 or multiple vessel (yes/no) EDV identification
No residual No residual
No residual
No residual No residual done at 3 months# No residual
No residual No residual
No residual
No residual No residual No residual No residual No residual No residual No residual No residual No residual No residual No residual No residual
No residual No residual Done at 2 years no residual# No residual Done at 2 months no residual No residual No residual
Postoperative angio residual?
1 3
1
0
0 2
1 1
1
1 0 0 1 0 1 0 0 0 0 1 0
1 0
0 0
0 0 0
Outcome mRS
Acta Neurochir (2015) 157:351–359 353
No residual/N No residual/N Yes/yes Yes/yes No No (IV) (IV)
Yes Yes Yes Yes 31 (F, 27) 32 (F, 24)
arterial feeders, venous drainage and peripheral extension. In this regard, ICG was found to be useful in 26 patients (81 %). ICG was also helpful in determining en passage vessels from malformation-based pathological vessels in 9 patients out of 16 (56 %) in whom the preoperative angiogram had clearly identified an en passage vessel. In three cases with a very small subcortical nidus not clearly distinguishable at the surface, the ICG was able to illustrate the location of the nidus based on robust surface AV shunting. In only one case, the ICG demonstrated a small residual nidus, which was further resected and confirmed obliterated by repeat ICG and subsequently intraoperative DSA.
Intraoperative DSA Intraoperative digital subtraction angiography was performed in all 32 patients. There were no related complications. Thirty of the 32 patients did not have any residual AVM nidus or AV shunt. In two cases, a small residual nidus with an early deepseated shunt was identified. One of the patients had a grade 3 and the other a grade 4 Spetzler-Martin AVM. Further resections were performed and subsequent repeat intraoperative angiograms confirmed complete obliteration. #Three of the postoperative angiograms were performed outside the time frame of 48 h to 2 weeks
3
1 residual shunt No residual (further resected) No residual No residual No residual AV shunt residual No residual/Y Yes/no Yes Yes Yes 30 (M, 45) (IV)
*ICG helpful means the surgeon’s perception of being able to distinguish arterial (feeding arteries) from venous structures (early draining vein) and also able to distinguish the nidus peripheral extension and identify en passage vessels
2 No residual No residual No residual/N No/no Yes Yes Yes (III) 29 (F, 28)
Yes (hydrocephalus, cerebellar deficit, left-sided weakness) Yes (hemiparesis, decreased mental status) None (headache and ITP) Yes (seizure, left-sided sensory deficit)
Postoperative angio residual? Intraoprtative ICGIntraoperative ICG Intraoperative AVM configuration residual?/en passage angio residual? helpful?*/1 or multiple vessel (yes/no) EDV identification AVM grade Deep location Deep Ruptured Neurological deficit (S-M grading) (or some deep venous preoperative/ presentation components) drainage Patient sex/age
Table 1 (continued)
0 0
Acta Neurochir (2015) 157:351–359 Outcome mRS
354
Postoperative DSA All 32 patients had postoperative DSA. There were no related complications. Thirty-one of the 32 patients were found to have negative angiograms in terms of the AVM nidus or AV shunt. One patient with a grade 4 AVM was found to harbor residual deep-seated AV shunting, missed on both ICG angiography and intraoperative DSA. Overall, ICG angiography missed three out of four residual AVMs after initial resection, and the intraoperative angiogram missed one. Postoperative DSA confirmed preservation of en passage vessels in the nine patients in whom the ICG had identified the en passage vessel before surgical resection (Table 2).
Illustrative case 1 A 38-year-old female presents with new-onset seizure. Investigations revealed a right temporal Spetzler-Martin grade 2 AVM (Fig. 1a), with an en passage angular artery identified on the preoperative DSA (Fig. 1b). Intraoperative ICG angiography helped to distinguish the en passage vessel from the aberrant AVM-based vessels (Fig. 1c and d). The AVM was resected without residual nidus or shunts confirmed with ICG angiography, intra- and postoperative DSA. The en passage vessel was preserved (Fig. 1e).
Acta Neurochir (2015) 157:351–359 Table 2
355
Summary of relevant intraoperative ICG findings compared to the intra- and postoperative angiogram
Identification of en passage vessel intraoperative/en passage vessel identified on preoperative angiogram ICG helpful in identification of AVM angioarchitecture Missed residual nidus/total number of residual nidus
Intraoperative ICG
Intraoperative DSA
Postoperative DSA
9/16
Confirmation of preservation of 9 en passage vessels N/A 1/4
Confirmation of preservation of 9 en passage vessels N/A N/A
26/32 3/4
Illustrative case 2
Illustrative case 3
A 24-year-old female presented with intractable seizures and was found to have a large right parietal grade 4 AVM (Fig. 2a). Preoperative partial embolization with Onyx (Microtherapeutics, Irvine, CA) was performed. The patient then underwent surgical resection and both ICG angiography and intraoperative DSA failed to show a residual nidus/shunt (Fig. 2b). The postoperative angiography however revealed a deep-seated residual shunt to the vein of Galen with a very small nidus attached to it (Fig. 2c).
A 42-year-old male presented with a parietal hemorrhage and left-sided weakness. He was found to have a right temporal grade 3 AVM (Fig. 3a, b and c). The patient underwent surgical resection. Once it was felt, the AVM was completely resected; an ICG angiogram was done that demonstrated a residual nidus within the surgical field not clearly seen with gross microscopic inspection (Fig. 3d and e). A further resection was undertaken to achieve total removal confirmed with an intra- and postoperative DSA (Fig. 3f).
Fig. 1 a AP view of a right temporal AVM, Spetzler-Martin grade 2. b Lateral view of a right temporal AVM with visualization of an en passage vessel, SMII. c Intraoperative view of the nidus limits, draining veins and en passage vessel. d ICG view of the nidus, draining veins and en passage vessel. e Postoperative DSA demonstrating obliteration of the right temporal AVM and preservation of the en passage vessel
356
Acta Neurochir (2015) 157:351–359
Fig. 2 a AP view of the right parietal Spetzler-Martin grade 4 AVM. b ICG (not shown) and subsequently intraoperative DSA demonstrating obliteration of the right parietal AVM. c Postoperative
DSA showing a deep residual shunt with persistent early drainage to the vein of Galen missed by ICG and intraoperative angiography
Outcome
majority of patients and to distinguish en passage vessels from abnormal vessels in a subset of patients. This could potentially contribute to a reduction in morbidity from accidental sacrifice of an en passage vessel. In our study, the en passage vessel was recognized and preserved at surgery in 9 of the 16 patients in whom the vessel was identified on preoperative angiogram. This further emphasizes the importance of maintaining physiological vessel integrity to achieve a better neurological outcome as none of those patients had new neurological deficits postoperatively. Although one could argue that en passage vessels can be identified without ICG injection, we believe that the dye injection will help to confirm the vessel detection in a significant subset of patients. Hänggi et al. [11] reported on a series of 15 patients with AVM where ICG angiography was used during the surgery. They emphasized that the most frequent surgery-related complications resulted from injury to arterial branches of normal brain and incomplete resection of the nidus. Therefore, the accurate detection of a traversing branch of normal brain vasculature ensures a higher likelihood of preservation of that vessel. In their series, they concluded that ICG was able to identify en passage vessels and very useful in the detection of the residual nidus. The latter findings are in discordance with our results. Taddei et al. [16] reached the same conclusion based on a series of nine patients and even suggested that ICG can replace intra-/postoperative DSA. Takagi et al. [17] also reported on a case of a 2-yearold female where ICG was superior to postoperative angiography in detecting residual disease. Based on our results, we conclude that ICG angiography is indeed of value in smaller, lower grade and superficial AVMs where the integrity of the AVM and surgical resection cavity is readily accessible for ICG image acquisition. ICG videoangiography can only depict what is seen directly within the surgical field. Because of this premise, there are certain conditions in which the residual lesion may not be evident,
Twenty-nine of the 32 patients did not experience any new complication or postoperative deficit. One patient had a new postoperative hemianopia, which improved to quadrantanopia at follow-up. One patient experienced a new hemisensory deficit, which almost fully recovered. One patient had a bone flap infection. Outcome modified Rankin score (mRS) analysis revealed a score of 0 for 18 patients, 1 for 10, 2 for 2 and 3 for the other 2 patients (Table 1). There was no mortality in this series.
Discussion The use of intraoperative DSA for AVM surgery has documented benefits and become an important adjunct for confirmation of total obliteration [2, 3, 5, 6, 8, 10–18]. Fluorescent angiography with intraoperative visualization of cerebral arteries has achieved widespread use because of its ease of utilization, rapid delivery and patient safety [3–6, 9, 13]. The technique is safe, requires no additional equipment except for microscope software modification and can be administered multiple times throughout a surgical procedure. Most importantly, it potentially integrates vital information in real time in order to direct the surgeon as to the next course of action. For these reasons, it has become a recent contender in displacing DSA as the procedure of choice for detecting residual aneurysms [3, 4, 14] or AVMs intraoperatively. The present study is the second largest series of patients with AVM who had evaluation of intraoperative ICG angiography compared to intra- and postoperative catheter angiography. The results of our study demonstrate the ability of ICG angiography to elucidate AVM angioarchitecture in the
Acta Neurochir (2015) 157:351–359
357
Fig. 3 a AP view of the right temporal Spetzler-Martin grade 3 AVM. b Lateral view of the right temporal Spetzler-Martin grade 3 AVM. c Intraoperative view of the right temporal region showing the surface representation of the AVM. d Intraoperative view of the resection cavity demonstrating no gross residual AVM nidus. e ICG angiogram demonstrating a residual AVM nidus in an area that did not appear to have a residual nidus on gross inspection. f Intraoperative DSA demonstrating gross total resection of the AVM nidus
e.g., if it is outside of the visual field, deep in the surgical cavity or covered by brain tissue because of poor illumination. In our study, the ICG failed to detect a residual AV shunt in one Spetzler-Martin grade 3 and two grade 4 AVMs. This acquiesces with the recent criticism of ICG angiography that deeper, more extensive AVMs are discerned with less specificity by fluorescent angiography than digital subtraction angiography [3, 12, 13, 18]. The results of our study mimic the results of Killory et al. [12] as well as a very recent investigation by the same institution. In their initial study of ten consecutive patients, ICG angiography clearly revealed AV shunting and helped discern feeding arteries, draining veins and en passage vessels. In two
patients ICG failed to identify residual AVM and persistent early venous drainage that were subsequently detected by DSA. Clear advantages of ICG angiography were identified, including the speed with which it can be performed, the low cost associated with it and the low to no morbidity associated with performing the procedure. In their latest investigation of 130 cases reported over a 4-year period, Zaidi et al. [18] evaluated whether ICG performed in patients undergoing AVM surgery improves rates of resection and clinical outcome. They pre-selected the subgroup with superficial or lobar AVMs for whom they used ICG and another subgroup with no ICG. Both ICG and non-ICG groups had either intraor postoperative angiography performed to determine residual
358
AVMs and the need for revision or second surgery. As mentioned above, ICG was used mainly when the AVM nidus was close to the cortical surface or lobar. Approximately 14 % of patients had residual disease and reoperation rates, and mRS scores were not different between the two groups. They concluded that ICG may be beneficial in small superficial AVMs; however, it does not enhance the identification of residual disease or improve clinical outcomes for more complex and deeper lesions. In contrast to their study, all patients in our group (including superficial, lobar or deep AVMs) had the ICG technique evaluated; therefore, the weakness of ICG for deep and larger AVMs was directly demonstrated. With regard to the intraoperative angiogram, our series underlines the importance of high-quality angiography as one of the residual AVMs in this group was missed by standard intraoperative angiography using the C-arm. Ideally, a high-quality high-resolution angiogram in a hybrid endovascular neurosurgery suite can offer the same advantage as a routine postoperative angiogram [7]. This will not only increase the reliability of intraoperative assessment of AVM resection, but also obviates the need for a postoperative angiogram and potential second treatment if a residual is found after the operation. The timing of postoperative DSA is very crucial. We ideally do the postoperative DSA during the same admission, and, as described above, this was done between 48 h to 2 weeks. Except for three cases in which the postoperative angiogram was performed after 2 weeks, they all had postoperative DSA during the same admission. While there is a minute risk of a tiny residual AVM being missed by postoperative DSA, we believe this risk is minimal, and we suggest one more follow-up angiogram at 1–3 years for all patients with resected AVM to address that minute possibility. There is also the risk of false-positive findings, which is mainly a dysplastic neovascular network of feeding arteries around the resected nidus and which in general disappears after a few weeks [15]. In the only case in our series where a residual was found postoperatively, the residual was clearly a retained nidus with a draining vein and could not be attributed to a false-positive finding. Our report is subject to all the limitations and biases of a retrospective analysis. Although it confirms some of the benefits of ICG angiography and illuminates a distinct advantage over other techniques to identify nonpathological en passage vessels in brain AVM surgery, it also highlights its major limitations and suggests caution in the interpretation of important intraoperative findings based solely on ICG angiography.
Conclusion Although ICG angiography is a helpful adjunct in the surgery of small and superficial AVMs. Its yield in detecting residual
Acta Neurochir (2015) 157:351–359
AVM nidus or shunt is low, especially for deep-seated lesions and higher grade AVMs. Therefore, ICG angiography should not be used as a sole and/or reliable technique. A highresolution postoperative angiogram must be performed in brain AVM surgery and remains the best test to confidently confirm AVM total resection.
Conflicts of interest None.
References 1. Alander JT, Kaartinen I, Laakso A, Patila T, Spillmann T, Tuchin VV, Venermo M, Välisuo P (2012) A review of indocyanine green fluorescent imaging in surgery. Int J Biomed Imaging 2012:940585 2. Anegawa S, Hayashi T, Torigoe R, Harada K, Kihara S (1994) Intraoperative angiography in the resection of arteriovenous malformations. J Neurosurg 80:73–78 3. Balamurugan S, Agrawal A, Kato Y, Sano H (2011) Intra operative indocyanine green video-angiography in cerebrovascular surgery: an overview with review of literature. Asian J Neurosurg 6:88–93 4. Barrow DL, Boyer KL, Joseph GJ (1992) Intraoperative angiography in the management of neurovascular disorders. Neurosurgery 30: 153–159 5. Bauer BL (1984) Intraoperative angiography in cerebral aneurysm and AV-malformation. Neurosurg Rev 7:209–217 6. Dashti R, Laakso A, Niemela M, Porras M, Hernesniemi J (2009) Microscope-integrated near-infrared indocyanine green videoangiography during surgery of intracranial aneurysms: the Helsinki experience. Surg Neurol 71:543–550, discussion 550 7. Dehdashti AR, Thines L, Da Costa LB, terBrugge KG, Willinsky RA, Wallace MC, Tymianski M (2009) Intraoperative biplanar rotational angiography during neurovascular surgery. Technical note. J Neurosurg 111:188–192 8. Faber F, Thon N, Fesl G, Rachinger W, Guckler R, Tonn JC, Schichor C (2011) Enhanced analysis of intracerebral arterioveneous malformations by the intraoperative use of analytical indocyanine green videoangiography: technical note. Acta Neurochir 153:2181– 2187 9. Feindel W, Yamamoto YL, Hodge CP (1967) Intracarotid fluorescein angiography: a new method for examination of the epicerebral circulation in man. Can Med Assoc J 96:1–7 10. Ghosh S, Levy ML, Stanley P, Nelson M, Giannotta SL, McComb JG (1999) Intraoperative angiography in the management of pediatric vascular disorders. Pediatr Neurosurg 30:16–22 11. Hanggi D, Etminan N, Steiger HJ (2010) The impact of microscopeintegrated intraoperative near-infrared indocyanine green videoangiography on surgery of arteriovenous malformations and dural arteriovenous fistulae. Neurosurgery 67:1094–1103, discussion 1103–1094 12. Killory BD, Nakaji P, Gonzales LF, Ponce FA, Wait SD, Spetzler RF (2009) Prospective evaluation of surgical microscope-integrated intraoperative near-infrared indocyanine green angiography during cerebral arteriovenous malformation surgery. Neurosurgery 65:456– 462, discussion 462 13. Ng YP, King NK, Wan KR, Wang E, Ng I (2013) Uses and limitations of indocyanine green videoangiography for flow analysis in arteriovenous malformation surgery. J Clin Neurosci 20:224–232 14. Raabe A, Nakaji P, Beck J, Kim LJ, Hsu FP, Kamerman JD, Seifert V, Spetzler RF (2005) Prospective evaluation of surgical microscopeintegrated intraoperative near-infrared indocyanine green
Acta Neurochir (2015) 157:351–359 videoangiography during aneurysm surgery. J Neurosurg 103:982– 989 15. Solomon RA, Connolly ES Jr, Prestigiacomo CJ, Khandji AG, PileSpellman J (2000) Management of residual dysplastic vessels after cerebral arteriovenous malformation resection: implications for postoperative angiography. Neurosurgery 46:1052–1060 16. Taddei G, Tommasi CD, Ricci A, Galzio RJ (2011) Arteriovenous m a l f o r m a t i o n s a n d i n t r a o p e r a t i v e in d o c y an i n e g r e e n videoangiography: preliminary experience. Neurol India 59:97–100
359 17. Takagi Y, Kikuta K, Nozaki K, Sawamura K, Hashimoto N (2007) Detection of a residual nidus by surgical microscope-integrated intraoperative near-infrared indocyanine green videoangiography in a child with a cerebral arteriovenous malformation. J Neurosurg 107: 416–418 18. Zaidi HA, Abla AA, Nakaji P, Chowdhry SA, Albuquerque FC, Spetzler RF (2014) Indocyanine green angiography in the surgical management of cerebral arteriovenous malformations: lessons learned in 130 consecutive cases. Neurosurgery 10(Suppl 2):246–251
