EC-IC and IC-IC Bypass for Giant Aneurysms Using the ELANA Technique

Published on 08/03/2015 by admin

Filed under Neurosurgery

Last modified 22/04/2025

Print this page

This article have been viewed 2223 times

28 EC-IC and IC-IC Bypass for Giant Aneurysms Using the ELANA Technique

Albert van der Zwan, Tristan P.C. van Doormaal, Luca Regli, Cees A.F. Tulleken

Natural history

Giant intracranial aneurysms, approximately 5% of all intracranial aneurysms, are one of the most challenging lesions in neurosurgery. By definition their fundus size is 2.5 cm or larger, and they are predominantly located in those proximal parts of the vascular tree with high blood flow, such as the internal carotid artery (cavernous and supraclinoid ICA), the proximal middle cerebral artery (MCA), the basilar artery (BA), and vertebral artery (VA). Two-thirds of these aneurysms are located in the anterior circulation and are slightly more common in the female population. Approximately 8% present in the pediatric age, but most commonly they become manifest during the fifth to seventh decades. Most of these aneurysms (20% to 70%) present with rupture and they have the same re-rupture rate as other smaller aneurysms.^1,² The second most common presentation is the mass effect of the aneurysm causing ocular movement disturbances due to cranial nerve palsies, vision or vision field loss and focal weakness/gait disorders, headaches and seizures. Third, approximately 8% of giant aneurysms present with ischemic syndromes most likely due to embolic phenomena originating from the aneurysm sac to more distally located vascular territories. The natural history of these aneurysms is very poor with high morbidity and mortality rates between 65% to 85% within 2 years after discovery.^1,²

Therefore, effective treatment of these lesions is life-saving. However, this asks tremendous and intensive investment in pretreatment diagnostics and considerations regarding the two aims that are the most important in the treatment of giant aneurysms: the permanent exclusion of the aneurysm from the circulation and the relief of mass effect that in some cases also plays a role in symptomatology.

Therapeutic modalities

Many treatment modalities are extensively discussed in literature like Hunterian ligation, endovascular coiling, gluing, stenting, coiling, reconstructive clipping, trapping, and parent vessel replacement bypass surgery, or combinations of these treatment modalities. Reliable comparative analysis of all these treatment modalities, however, is difficult due to small numbers, variable secondary circumstances, and no standardization of results or outcomes.

Endovascular Treatment

As primary clipping can be technically very difficult or impossible, the endovascular approach to these aneurysms has been increasingly popular. Although in smaller aneurysms endovascular treatment has proven to be effective in many cases, the results of endovascular treatment of giant aneurysms are still very disappointing.^3,⁴ Endovascular Hunterian obstruction in the case of a giant ICA aneurysm of the parent vessel seems to be the most effective and simplest method of treatment. The functionality of the ICA and its intracranial collaterals as the circle of Willis and the leptomeningeals should be thoroughly evaluated before definitive occlusion. Therefore, at our center in patients harboring these aneurysms, first a balloon test occlusion (BTO) of the ICA is performed with venous outflow study to observe if the patient tolerates definitive ICA occlusion. If the patient clinically does not tolerate the occlusion, or if asymmetry in venous outflow is over 1 second, a replacement bypass procedure is planned before definitive ICA occlusion. In more distal giant aneurysms, however, BTO is less reliable due to the complexity of the aneurysm and its afferent arteries. In the explicit review of Parkinson et al., a total of 316 patients with giant aneurysms treated endovascularly using coils, parent vessel occlusion, Onyx®, stenting with or without coil or Onyx® were reviewed. In only 57% of patients was a complete occlusion or cure achieved, with 7.7% mortality and 17.2% major neurological morbidity.² In addition, in distally located giant aneurysms the results of endovascular treatment are poor.⁴ Endosaccular coiling proves to be an ineffective long-term treatment of these aneurysms due to recanalization and further aneurysm growth. Parent vessel occlusion can be considered, but risks of late ischemia remain relatively large. Selective embolization of these distal giant aneurysms harbors high complication risks (20% mortality). In those cases, especially in more distal giant aneurysms of the MCA, ACA, and PCA where mass effects also play a role in symptomatology, endovascular treatment seems to be less effective.

Parent Vessel Replacement Bypass Surgery

No major reviews on results of flow replacement bypass surgery are available but smaller series are available.^5,⁶ From these studies the results of parent artery bypass surgery using mostly VSM or radial arterial grafts with consequent trapping or proximal occluding with or without additional endovascular therapy seem to be more promising than the results of endovascular treatment alone.

Before planning bypass surgery it is mandatory to determine the amount of flow to be replaced by the bypass. As most giant aneurysms are located in the more proximal cerebropetal arteries conducting high flows like the ICA, MCA, ACA, BA, and VA, the flow to be replaced by any bypass is most highly dependent on the functional capacity of the Circle of Willis and collaterals as the leptomeningeal vessels, ophthalmic artery, and choroidal plexus. Exact determination of flow needed to replace is therefore hard to make. However, in case of BTO failure in a giant ICA aneurysm, it seems to be safe to choose for a bypass type with potential high flows. In cases of giant MCA, ACA, BA, or VA aneurysms the flows to be replaced in the parent arteries should be determined using single vessel flow MRA flowmetry or multivessel NOVA technique (VasSol®).⁷^–⁹Consequently, partially based on these data a custom made plan for an EC-IC or IC-IC bypass has to be defined to rebuild vascular architecture. The anatomy must be studied in detail to determine which vessels are approachable for bypass grafting. Currently, conventional angiography, and CT perfusion and MRA are used to study this.

Before bypass attachment, intraoperative quantitative flow measurements are performed on different intracerebral arteries using a dedicated flow meter (Transonic®, Ithaca, NY), again to assess the flow, which has to be replaced by the bypass. If we use the superficial temporal artery (STA) as a donor, cut flow is measured before bypass completion to use in the cut flow index.¹⁰

In aneurysms located more distally, like on the P2 or M3, we measured medium-high flows and replaced the flows using a conventional STA-MCA or STA-PCA bypass based on pre- and intra-operative flow measurements where the aneurysm was trapped. The flows in the bypasses ranged between 30 and 60 cc/min after trapping. In more proximally located aneurysms like the ICA, MCA, BA, and VA, the flows to be replaced can measure up to 180 cc/min. In those cases, a high-flow bypass on the proximal arteries to replace these amounts of flow has to be considered.

It is important here to recall Poiseuille’s law, which says that flow varies proportionally to the pressure drop over the bypass, the length of the bypass, and the fourth power of the radius: Q = π r4 P/8μL (Q = flow rate; r = radius of pipe; P = pressure drop over the pipe; μ = viscosity of fluid; L = length of the pipe). For example, because cortical vessel and the STA have a relatively small diameter, the possible flow through the conventional STA-cortical MCA bypass is limited to 10 to 40 cc/min. Increased flows of 84 to 100 mL/min are possible when the STA is connected to the larger M3.¹¹ High flows of 130 to 200 cc/min can be achieved when a venous or arterial interponate is used to connect the ECA with the intracranial ICA. So, in cases where higher flows are needed, a more proximal anastomosis should be made. However, anastomosing techniques on more proximally located arteries are technically demanding for surgeons and anesthesiologists. The conventional anastomosing technique with temporal occlusion of the major cerebral arteries theoretically provides an extra risk of ischemia during the classical procedure, and measures to prevent hypothermia and cardiac arrest and pentothal protection are described to minimize this risk. To circumvent this risk, the ELANA technique was developed by our senior author (Tulleken).

The elana technique

The ELANA technique (Figure 28–1) facilitates the construction of an end-to-side anastomosis without temporary occlusion of the recipient artery. It was primarily developed for cerebral augmentative revascularization (1992); however, it turned out that the technique was also quite feasible for creating protective or replacement bypasses.¹²^–¹⁵ A vein (mostly the vena saphena magna, VSM) or artery (radial artery, RA) can be used as the donor vessel. The inner diameter should be minimally 2 mm to enable the laser catheter to pass through the donor vessel. First, depending on the size of the recipient vessel, a platinum ring of 2.6 mm or 2.8 mm is attached to the distal segment of the donor vessel using eight microsutures (see Figures 28–1A through 28–1C). The ring with the attached distal donor segment is subsequently stitched end-to-side to the recipient using again eight microsutures (see Figure 28–1D). This is followed by the passing of the ELANA 2.0® laser suction catheter down the lumen of the open donor vessel. The tip of the catheter is placed against the sidewall of the recipient vessel (see Figure 28–1E and Figure 28–2). After 2 minutes of active suctioning from the dedicated inside portion of the catheter, the laser fibers on the outside of the catheter are activated within 5 seconds. The laser broaches the recipient arterial wall and separates an arteriotomy flap from the recipient (see Figure 28–1F). The suction portion of the catheter maintains contact with the small arteriotomy flap, thus preventing its migration into the lumen of the recipient.

Figure 28–1 Schematic drawing of ELANA anastomosis. A to C, Vein attached to platinum ring using eight ethilon 8/0 stitches. D, Attachment of vein/ring to intracranial artery. E, Introducing laser catheter through vein and lasing out flap. F, Withdrawal of catheter with flap.

(Redrawn from van Doormaal TP, van der Zwan A, Verweij BH, Langer DJ, Tulleken CA: Treatment of giant and large internal carotid artery aneurysms with a high-flow replacement bypass using the excimer laser-assisted nonocclusive anastomosis technique, Neurosurgery 2008;62:1414, with permission.)

Figure 28–2 ELANA laser catheter tip.

A major advantage of this procedure is the nonocclusive character of the anastomosis, and to our knowledge this is the first nonocclusive anastomosis technique in neurosurgery. Second, for creating this nonocclusive type of anastomosis less intracranial arterial exposure is needed as no temporal clips are used. Third, no measures as hypothermia are needed, as the complete procedure is nonischemic. Fourth, as there is no exposure to the ring and maximally four through-the-vessel-microsutures are needed, adequate re-endothelialization is achieved.¹⁶ Due to these characteristics safe anastomoses can be made on proximal recipients (ICA, proximal MCA, proximal ACA, P1, and BA) (Figure 28–3A through 28–3E) and various bypass types have been constructed (Figure 28–4). At our institute more than 300 patients and worldwide a large cohort of more than 400 patients were treated using the ELANA technique between 1992 and 2010 with good results and good long-term patency rates (up to 93%). After finalizing the ELANA technique in the form described, we treated 265 patients with large and giant aneurysms since 1999. These aneurysms were located in the cavernous ICA, supraclinoid ICA, ICA bifurcation, proximal MCA, proximal ACA, BA, and at the VA-BA junction. Although this technique introduces the major advantage of its nonocclusiveness and therefore the absence of intraoperative stress moments of temporal occlusion of major arteries, the technical challenges are mainly formed by the stitching of the donor vessel with the ring on the recipient vessel, most often deep intracranially. Mainly as the result of this part of the procedure, one ELANA anastomosis takes between 90 to 120 minutes, depending on its location and the neurosurgeon’s experience.

Figure 28–3 Angiography of ELANA anastomoses. A, On ICA. B, ICA bifurcation. C, A1. D, MCA bifurcation. E, Basilar artery.

Figure 28–4 Different types of ELANA bypasses.

Therefore, ELANA training sessions organized at our institute are mandatory for future ELANA neurosurgeons. In addition, for every vascular neurosurgeon, regular lab training sessions are obligatory for maintaining and developing adequate microvascular skills using ELANA techniques. Further developments of the ELANA technique will focus on safe, sutureless, and minimally invasive techniques to make this technique easier and consequently shorten the procedure.

Perioperative preparation and monitoring

Preoperative Flow Measurements

For preoperative planning of surgery, bypass indication, type of bypass and location, and an estimation of the flow to be replaced can be made by using single-vessel flow MRA flowmetry or multivessel NOVA technique (VasSol®). At our institute, only single-vessel MRA flowmetry is available, and this technique is also used for postoperative evaluation of bypass flow.⁹

Anticoagulation Strategy

Although in the early years of our experience we did not administer anticoagulation medication preoperatively to avoid hemorrhagic complications during surgery, over 4 years ago we began to premedicate patients with acetylsalicylic acid (80 mg/day) 3 days before surgery. Heparin is administered intraoperatively intravenously by bolus injection (1000 IU) at the moment that the bypass is opened. During the rest of the procedure, an extra bolus of 1000 IU heparin is administered approximately every 2 to 3 hours. At the end of the surgery, the median quantity of intraoperatively administered heparin is 2000 IU (range, 1000 to 5000 IU). Postoperatively, all patients receive bypass thrombosis prophylaxis. From 1999 to 2002, this was achieved with 50 g/day dextran 40 (Rheomacrodex; Medisan Pharmaceuticals, Parsippany, NJ) and, from 2002 on, with 15,000 IU/day of heparin. Both of these medications were administered during the 3 days in the intensive care unit. After 3 days, they were stopped, and one daily dose of acetylsalicylic acid (80 mg) was administered. We changed this postoperative protocol over 4 years ago, however, and only acetylsalicylic acid is continued directly after surgery.

Intraoperative Flow Measurements

In all procedures since 1998 we intraoperatively measured flow through parent vessels, distal efferent vessels, and the bypass using a dedicated flow meter (Transonic®, Ithaca, NY).⁹ Especially in flow replacement bypass surgery it is important to measure the flow in the artery that is to be replaced. In addition, after bypass completion, it is important to monitor bypass flow during parent vessel occlusion and aneurysm trapment to measure the increase in the flow through the bypass during these maneuvers. Any complication, such as kinking of the bypass, compression of the bypass during closure, or thrombosis of the bypass, can be detected by flow monitoring during the rest of the procedure until the end of closure.

Intraoperative Near-Infrared Indocyanine Green Video Angiography (ICGA)

Before and after bypass construction as well as during aneurysm trapment or reconstruction, the use of indocyanine green video angiography (ICGA)¹⁷ can be very useful to monitor involvement of perforating arteries. Although this method is not replacing intraoperative DSA, which is not available in our hospital, it can be very helpful as long as it is taken into account that this method can only be used on those vessels in the field of surgery that are visible, and that deeper vessels are missed.

ELANA EC-IC bypass for giant aneurysms of the ICA

Between 1999 and 2009 we operated on 45 patients with giant aneurysms of the cavernous and supraclinoid ICA ¹⁸ (Figure 28–5). Mean age was 53 years (standard deviation [SD] 15) and predominantly female (75%). Thirty-two patients (71%) failed occlusion tests, and the remaining patients lacked sufficient collaterals on imaging. Presentation of symptoms included cranial nerve compression signs exclusively in 27 patients (75%) and SAH, with or without cranial nerve compression signs, in 14 patients (31%). In the remaining four patients, no objective symptoms were registered. In all patients the VSM was used for bypass. The vein was cut in half and firstly the ELANA anastomosis was created on the intracranial ICA, ICA bifurcation, or proximal MCA. In a second phase the other half of the vein was used to make a conventional end-to-side anastomosis on the ECA and finally both vein pieces were connected in an end-to-end anastomosis. The mean surgical time for the bypass procedure of all patients was 443 minutes (range, 300 to 750 min). The mean time until catheter vacuum suction—meaning the anastomosis between recipient, platinum ring, and donor vessel was made, but the recipient artery wall was still intact—was 246 minutes (range, 150 to 330 min). During surgery, the recipient artery was never occluded as part of the bypass procedure. With an intact bypass, we proceeded with intraoperative ligation of the ICA if an acute danger of aneurysm bleeding existed or if the compression symptoms increased rapidly before the operation. In the first patients of our series, the ICA was left patent during surgery to prevent ischemic complications. In these patients, we occluded the ICA after surgery with detachable balloons after confirmation that the bypass remained patent for at least 24 hours as demonstrated on contrast imaging. However, at present, we choose for immediate intraoperative occlusion of the ICA, because in the early years five out of 23 patients (22%) had to be re-operated on. In these cases, the bypass occluded within a day due to low bypass flow because the ICA was left open postoperatively, waiting for endovascular occlusion.

Figure 28–5 Preoperative (A) and postoperative (B) angiography of a giant aneurysm of the left ICA after unsuccessful previous endovascular treatment. During surgery, an EC-IC bypass from the ECA to the ACA (ELANA) was performed and the ICA proximally occluded.

A successful patent EC-IC bypass was constructed in 44 out of 45 patients (98%). After ligation of the ICA, a mean intraoperative bypass flow of 102 mL/min (range, 65 to 170 mL/min) was recorded using a flow probe (Transonic®, Ithaca, NY). In 15 patients, magnetic resonance angiography bypass flow measurement was performed after ICA occlusion. A mean flow of 138 ml/min (range, 70 to 180 mL/min) was recorded, demonstrating a powerful replacement EC-IC bypass. Control imaging from all patients in whom we performed intraoperative ICA ligation or postoperative ICA balloon occlusion showed ICA aneurysm thrombosis.

Complications of ELANA Bypass for Giant Aneurysms of the ICA

Two fatal complications (4%) occurred, neither directly related to ELANA application. The first patient died of an air embolism from a central line 1 day postoperatively and the second patient died of an aneurysm rupture 2 days after surgery. Both patients waited for postoperative endovascular final treatment. The last complication again underlines the importance of direct surgical treatment by ligating the ICA after the bypass procedure.

Non-fatal complications, including ischemia due to thrombosis or manipulation of the aneurysm (two patients), aneurysm bleeding before definitive occlusion of the ICA (one patient), ischemia due to thrombosis of the bypass (two patients), occurrence of transient ischemic attacks (TIAs) due to stenosis of the conventional anastomosis in the ECA (one patient), and progressive cranial nerve palsy due to aneurysm thrombosis (two patients) occurred in eight patients (18%).

Bypass Patency of ELANA Bypass for Giant Aneurysms of the ICA

As there is heterogeneity of surgical strategy and indication in the pursuing years, patency data on the whole group are not informative. As in the early years, for safety reasons, we planned to wait with definitive treatment of the aneurysm 1 or more days after bypass surgery. The patency rate in this group was only 71%. The competition between bypass flow and parent artery flow plays an important role in this. Changing our strategy to immediate (partial) trapping of the parent artery and aneurysm increased patency rates to 93%.

Functional Outcome of ELANA Bypass for Giant Aneurysms of the ICA

Long-term follow-up outcome was defined as favorable if the modified Rankin Scale (mRS) was found to be equal to or higher than the preoperative mRS (mean long-term follow-up: 3.3 years; range, 0.6 to 5.6 years). This was found in 35 patients (78%). Defining patients with a mRS score of 3 (moderate disability, needing help in daily life, but walking without assistance) or worse, 37 patients (82%) of our patient group were independent in the long term. Cranial nerve recovery occurred in 42% of the patients.

Discussion of ELANA Bypass for Giant Aneurysms of the ICA

Comparing these data with the results of currently available endovascular treatment modalities, our results strongly suggest that neurosurgical treatment using ELANA EC-IC bypass neurosurgery is favorable above endovascular treatment. Our changing strategy of directly, intraoperatively ligating the ICA seems to obtain even better results.

ELANA bypass for giant aneurysms of the MCA

As MCA giant aneurysms are even more life-threatening lesions prone to rupture compared to giant aneurysms of the ICA, an aggressive therapeutic strategy is even more warranted ¹ (Figure 28–6). The preferred treatment is to take the aneurysm completely out of circulation and in some cases even debulking the mass of the aneurysm. The results of filling these MCA aneurysms with coils or glue are mostly disappointing by recanalization, impaction, or sustaining mass effect. Although literature provides many case reports of flow replacement bypasses using the STA as donor end-to-side connected to a small distal MCA branch for reversal of flow in these aneurysms, this technique still demands a temporal MCA occlusion with a theoretical risk of ischemia. In addition there is also a risk of ischemia due to insufficient flow replacement. According to the mathematical model of Hillen et al., there are hemodynamic arguments that, for replacing higher flows in the MCA (up to 120 ml/min), larger diameter bypasses (>2 mm) should be preferred.¹⁹ An intracranial IC-IC bypass should be able to replace the MCA in an orthograde, and therefore, more physiological direction; this approach also minimizes surgical exposure, thus avoiding neck carotid surgery for the donor part of a regular conventional EC-IC bypass.

Figure 28–6 Preoperative (A) and postoperative (B) angiography of patient with a giant aneurysm of the left MCA. ELANA IC-IC bypass (type I) from the ICA to the M1 bifurcation.

Therefore, at our institute, we prefer to treat these aneurysms using IC-IC ELANA replacement bypass followed by trapment of the complete aneurysm after determining the flow to be replaced by combining preoperative (MRI) and intraoperative flowmetry data (Transonic Flow Probe and Flow Meter, Transonic Systems, Ithaca, NY).

Since 1999, we have treated 25 patients with giant aneurysms with a mean diameter of 38 mm (SD, ±12 mm) measured on imaging.²⁰ The mean age was 45 years (SD ±16 years) and 12 patients were female (48%). All aneurysms were assessed as being not safely clippable or coilable. Preoperative flow measurements were performed as far as possible, as coils previously unsuccessfully delivered into these aneurysms (in six patients, 24%) disturb flow measurements.

In 11 patients (44%), the aneurysm presented with SAH, in three patients (12%) with ischemic stroke, in three patients (12%) with epilepsy, in three patients with TIA (12%), and in four patients with no objective symptoms. Three types of ELANA bypasses were performed in this group of patients, all using the VSM as the bypass graft.

Type I, ICe-ICe: Type I is defined as an IC-IC bypass with proximal and distal anastomoses constructed both using the ELANA technique (ICe-ICe). This type of bypass was performed in 12 patients (36%). The proximal anastomosis was made on the proximal ICA (n = 8) or MCA (n = 4) depending on reach ability and quality of these proximal arteries. The VSM was cut in half and the anastomosis was constructed on the ICA or MCA. After completion of this anastomosis a temporal clip was placed on the vein. The distal anastomosis was mostly made on the distal MCA (M2, n = 10; M3, n = 2) using the second half of the VSM, again followed by temporally clipping this vein. The two vein halves were then end-to-end anastomosed and just before finalizing flushed by removing both temporal clips. An example of this type of bypass is shown in Figure 28–6. As no temporal occlusion of any of the vessels is necessary, we prefer to construct this type of bypass.

In three patients the configuration of the aneurysm was very complex with an extra MCA branch originating from the aneurysm. Therefore an extra EC-IC bypass was constructed considering the size of these smaller branches using conventional anastomosing techniques (end-to-side STA-MCA) (Figure 28–7).

Type II, ICe-ICc: Type II is defined as an IC-IC bypass with an ELANA anastomosis on the proximal side and a conventional anastomosis on the distal side, as the distal MCA branch is not always large enough (>2.6 mm) to construct an ELANA anastomosis on (ICe-ICc). This type was constructed in 14 patients (56%). The proximal ELANA anastomosis was made on the ICA (n = 10) and on the MCA (n = 4). Distal conventional anastomoses were made on the M2 (n = 4) and on the M3 (n = 10).

Type III, EC-ICe: This type is defined as an EC-IC bypass with an ELANA anastomosis on the distal side (EC-ICe). In two patients (8%) the intracranial ICA or MCA proximal to the aneurysm was not surgically reachable due to aneurysm size. Therefore an EC-IC bypass with a distal ELANA anastomosis on the distal MCA and a proximal anastomosis on the ECA was constructed. This type of flow replacement bypass was described previously in the treatment of internal carotid aneurysms.

Figure 28–7 Seventeen-year-old boy with giant MCA aneurysm (A). An IC-IC ELANA bypass (B) and an extra STA-MCA bypass (C) were constructed. The MCA aneurysm was completely trapped.

Of 59 anastomoses, 40 (68%) were ELANA. In three patients, a second bypass procedure was initiated after the first bypass attempt failed. All 40 ELANA attempts resulted in a patent anastomosis with a strong backflow directly after ELANA catheter retraction. In six ELANA anastomoses (15%), the disk of arterial wall, which is normally attached to the tip of the catheter after laser application and catheter retraction (the flap), was not found on the tip. In these patients no adverse events occurred and bypasses were functioning well.

Mean intraoperative flow (±SD) through the 23 IC-IC bypasses before any part of the MCA or aneurysm was occluded was 15 (±10 cc/min). The mean intraoperative flow (±SD) through the IC-IC bypasses after partial or full MCA occlusion, measured at the end of the operation, was 53 (±13 ml/min). The two EC-IC bypasses had a flow before MCA occlusion of 35 (±50 ml/min). These aneurysms were endovascularly treated the next day, and MR flow measurement on the first EC-IC bypass showed that bypass flow had increased from 35 cc/min (intraoperatively) to 100 cc/min postoperatively and after endovascular occlusion of the parent artery.

In the early phases of our experience, for safety reasons we decided to construct the bypass first and planned the definitive aneurysm endovascular treatment 1 or 2 days postoperatively. However, as we found out, the risks of this strategy are twofold—thrombosis of the bypass and postoperative preendovascular treatment aneurysm bleeding—so our present strategy is total exclusion of the aneurysm from the circulation during the same procedure. This should be performed with clips placed on the MCA proximal of and distal from the aneurysm, thus trapping the aneurysm completely. We succeeded in this in only 11 patients (44%), because in most cases total trapping can be impossible if the complexity of the aneurysm base (e.g., calcification or vessel incorporation) does not allow clip application on the MCA nearby. In that situation, we tried different partial trapping configurations with clips on the proximal and/or distal parts of the MCA, leaving a selection of proximal and/or distal MCA branches patent. We repeatedly measured bypass flow and MCA flow (Transonic flow probe and flow meter; Transonic Systems, Inc., Ithaca, NY) after every temporary and definitive clip placement to monitor sufficient bypass flow. This partial trapping approach was used in 14 patients (56%). In two of these patients, there was still a considerable flow in the aneurysm after distal partial trapping. These remnants, including the parent vessel, were successfully coiled after surgery. However, partial trapping in which the aneurysm itself partially filling also comes with a risk of bleeding of the aneurysm. This occurred in four out of 14 aneurysms (29%) that were partially trapped. In one patient with a large 60-mm MCA fusiform aneurysm where partial trapping was impossible due to calcifications, a balloon occlusion was planned 3 days postoperatively. One day postoperatively, the aneurysm bled and the patient died.

Of special interest were our findings on one patient who was provided with an IC-IC bypass. At that time in the early phase of our experience we decided to plan the definite endovascular treatment 1 day after surgery using detachable balloon as described previously. Surprisingly, the next day, the aneurysm spontaneously thrombosed with the bypass still open. Currently, this is not our strategy, as we consider the risk of thrombosis of the bypass with the parent artery still open too high.

Complications of ELANA Bypass for Giant Aneurysms of the MCA

Hemorrhagic: Besides the fatal aneurismal hemorrhage (n = 1, 4%), there were 11 nonfatal complications in nine patients (36%). In three patients, a nonfatal postoperative SAH occurred before endovascular treatment was performed.

Ischemic: Eight ischemic complications occurred in eight patients (32%). Three of these (12%) were due to failure of the distal conventional anastomosis in type II treatment modality. In the first of these three patients, the thrombosis began at the site of the distal conventional anastomosis and the proximal ELANA anastomosis remained patent as was clearly visible on angiography. We decided not to perform a second bypass procedure because of the relatively small neurological deficits (right-hand paresis), which was probably attributable to a large intracerebral and leptomeningeal collateral network also visible on angiography. In the other two patients, an ischemic region was observed at the site of the temporarily occluded recipient distal M2 and M3 branch.

Four patients (16%) experienced minor ischemic symptoms because of the occlusion of small MCA branches leaving the aneurysm as a result of aneurysm thrombosis. However, none of these patients experienced a poor functional outcome (postoperative mRS score higher than preoperative mRS score).

Finally, in one patient (4%) an intraoperative infarct occurred. She was a 22-year-old patient with a left-sided giant MCA aneurysm with only headache symptoms (see Figure 28–6). An IC-IC bypass (type I) was planned and during surgery, when both proximal and distal ELANA anastomoses were constructed and only the final stage of end-to-end anastomosing of the two venous bypass segments was to be finished, a spontaneous hemorrhage of the (until this part) asymptomatic aneurysm occurred at the neck of the aneurysm. The massive bleeding that followed urged us to temporally trap the MCA and the aneurysm to make further finishing of the bypass possible. This turned the procedure from a nonocclusive one into a temporal ischemic procedure and it took us 12 minutes to finish the bypass and remove the temporary clips on the bypass. During this time we checked collateral leptomeningeal capacity by removing the distal clip on the MCA, and there was no backflow indicating that in this patient collateral capacity was minimal. Although the occlusion time was only 12 minutes, which is much shorter than the occlusion time usually necessary for a conventional anastomosis on larger proximal arteries, the patient woke up with right-sided hemiparesis and mild dysphasia. Postoperative CT showed a left-sided putamen infarct, and 2 months were required for full recovery. This incident, together with the three ischemic incidents of the distal conventional anastomoses, clearly shows that conventional anastomosing techniques with temporal occlusion bear significant risk of infarction and that the nonocclusive character of the ELANA anastomosing technique does prevent infarcts in certain patient groups. As there are currently no reliable methods to test leptomeningeal capacity in the treatment of giant MCA aneurysms, we prefer to use the ELANA nonocclusive anastomosing technique in these patients as much as possible.

Bypass Patency of ELANA Bypass for Giant Aneurysms of the MCA

In total, four bypasses occluded because of thrombosis at the distal conventional anastomosis site (16%), and one bypass thrombosed because of a postoperative aneurysm hemorrhage (4%), making short-term bypass patency in 20 (80%) of 25 bypasses. Although long-term follow-up studies are preferred, we could not manage to have a 100% long-term follow-up MRA in all patients. However, of the remaining 22 short-term patent bypasses, two patients after 2 and 3 years show symptomless occlusion of the bypass but with extensive collaterals between the distal MCA branches and the ACA and PCA arterial networks. Long-term patency in this group therefore amounted to 72%.

Functional Outcome of ELANA Bypass for Giant Aneurysms of the MCA

On long-term follow-up (mean, 3.6 years; range, 0.2 to 7.7 years after surgery), 20 patients (80%) had a favorable outcome, meaning that their postoperative mRS score was equal to (10 patients) or higher (10 patients) than their preoperative status. Five patients (20%) had a long-term unfavorable outcome. In four of these patients, the unfavorable outcome was related to surgery (16%). One patient (5%) died 2 months postoperatively of unknown causes. The family refused a postmortem study, but the patency of the bypass of this patient was angiographically confirmed 2 weeks before death.

Defining patients with a mRS score of 3 (moderate disability, needing help in daily life, but walking without assistance) or worse as dependent, long-term independence totaled 16 patients (64%).

Discussion of ELANA Bypass for Giant Aneurysms of the MCA

Comparing the results of these treatments with literature is problematic, as different patient groups and outcome measurements are used. Using GOS scale in our patients we had a good or excellent outcome in 84%. However, this does not take the preoperative status into account. In addition the challenges in the treatment of giant MCA aneurysms are unique and should not be compared with results in the treatment of giant ICA or posterior circulation aneurysms.

It is noteworthy that, in 14 patients in whom the aneurysm was not completely trapped intraoperatively, four aneurysms bled postoperatively (29%). For one patient the bleeding was fatal and in one patient the bypass occluded as the result of this. The outcome of these patients was poor. On the other hand, in 11 patients where complete trapping was possible the minor ischemic complications attributable to trapping or manipulation of the aneurysm did not cause any decrease in long-term functional outcome. The only bypass that on long term occluded after trapping also did not cause any decrease in functional outcome. Retrospectively, this bypass was probably not needed because of the preexisting large collateral flow through intracerebral and leptomeningeal collaterals, as shown on imaging. These differences in the groups with and without direct total trapping suggest that direct total trapping of the aneurysm after bypass construction during the same procedure should be pursued whenever possible. Prevention of hemorrhagic complications probably outweighs the functional decrease caused by ischemic complications.

Although numbers are too small to make reliable comparison possible, it is remarkable that three bypasses occluded because of thrombosis at the site of the conventional anastomoses, and another two patients had an ischemic complication with neurological sequelae owing to temporary occlusion time, which was necessary to construct a peripheral conventional anastomosis. This clearly shows that temporary occlusion of arteries to the brain is accompanied by higher risks.

The phenomenon of the long-term bypass occlusion in two patients with extensive leptomeningeal collaterals with previous successful and patent bypasses reveals that the human cerebral vasculature is dynamic and that it can adapt itself to changing circumstances. After bypass construction, a new equilibrium is slowly established in the MCA flow territory, which varies inter-individually but also intra-individually in time depending on preexisting collaterals and bypass location. The bypass could then function as a slowly fading, nonphysiologic bridge to reach a new physiologic situation with the aid of leptomeningeal and intracerebral collaterals. This hypothesis is supported by earlier reports on leptomeningeal function after MCA occlusion.²¹Although in literature numbers of bypass patencies are often used in the discussions of treatment modalities in giant aneurysms, this finding of long-term occlusion of bypasses without symptomatology, due to extensive collaterals, clearly demonstrates that follow-up in terms of functional outcome as the mRS is much more important.

ELANA bypass for giant aneurysms of the ACA

Although we frequently see patients with giant aneurysms of the ACA, it is rarely necessary to use ELANA technique for bypassing these lesions. An example of this is a 61-year-old patient with an SAH and a giant aneurysm of the ACA that was not coilable and because of multicalcifications not easily clippable (Figure 28–8). As there are no possibilities of measuring distal leptomeningeal anastomoses, we decided to perform a IC-IC bypass from the A1 to the proximal A2, and then clipping the aneurysm, including the distal A1 and ACA. Intraoperative flow measurements on the A2 showed a flow of 30 cc/min to be replaced. As the STA in this case was very small we decided to construct an IC-IC bypass and to trap the aneurysm. According to intraoperative flow measurements, initial flow in the bypass with the A1 still open was 15 cc/min but went up to 30 after trapping the aneurysm, including the A1-A2 segment, demonstrating that the bypass functioned as a 100% replacement bypass. The patient recovered very well and is at work (follow-up 40 months), and CTA control after 2 years showed the bypass to be fully patent.

Figure 28–8 Giant aneurysm of the ACA (A) and complete trapping of the aneurysm after construction of an ELANA IC-IC bypass from the A1 to the A2 (B).

ELANA bypass for giant aneurysms of the posterior circulation

Giant aneurysms of the posterior circulation are even more challenging than those of the anterior circulation. This is because of the large size of the aneurysm and the little space there is left for the thrombosing and consequently swelling aneurysm. In addition, thrombosing perforators coming out of the aneurysm often cause devastating infarcts in the brainstem with consequent sequelae. Finally, complete trapping is seldom possible, leaving the risk of bleeding of the aneurysm.

We treated 10 patients with a giant aneurysm of the posterior circulation with an ELANA bypass on the BA (Figures 28–9A and 28–9B), P1 (Figure 28–9C), P2, or the SCA.²²^–²⁴ An example of progressive brainstem ischemia after reversal of flow is the case of a 55-year-old male who suffered a SAH from a giant fusiform BA aneurysm. We constructed an IC-IC bypass from the ICA to the BA, thus creating a large artificial posterior communicating artery (PCom) (Figure 28–9A). Both anastomoses were made using the ELANA technique. The next day both VAs were endovascularly occluded and the flow through the bypass, which first was directed from the P1 to the ICA, now reversed to the posterior circulation supplying the BA vascular territories. The aneurysm slowly thrombosed, but the next day a progressive brainstem ischemia developed with the bypass still patent and the patient died.

Figure 28–9 A, EC-IC bypass from the ECA (conventional anastomosis) to the BA (ELANA anastomosis). B, IC-IC bypass from the ICA (ELANA anastomosis) to the BA (ELANA anastomosis). C, EC-IC bypass from the ECA (conventional anastomosis) to the P1 (ELANA anastomosis).

A clear example of the risk of aneurysm rupture after reversal of flow treatment and therefore not complete trapping of the aneurysm is a 44-year-old male patient with a giant BA aneurysm and progressive brainstem symptomatology (see Figure 28–9B). He was treated with an EC-IC bypass from the external carotid artery (ECA) to the BA using the ELANA technique for the latter anastomosis. Both VAs were occluded in the same session. Postoperative angiography showed partly thrombosis of the aneurysm due to reversal of flow. The patient was slowly recovering but died after 2 months as the result of a massive SAH from the aneurysm remnant.

Of the group of 10 patients, only two had a good functional outcome (20%). One of these two cases is a 31-year-old female with a history of SAH from a giant basilar aneurysm and unsuccessful coiling of the aneurysm 2 months before. An EC-IC ELANA bypass was constructed from the ECI to the P1 (ELANA anastomosis) (see Figure 28–9C). The intraoperative flow in this bypass was 70 cc/min with the VAs still open, and we planned an endovascular occlusion of the VAs and aneurysm the next day. The aneurysm occluded and the bypass completely took over posterior circulation. MR flowmetry after 2 weeks showed that the flow in the bypass increased to 120 cc/min, and on follow-up (12 years) she had maximum outcome scores and is fully active in daily life.