A case report co-authored by my recently graduated trainees, Drs. Roy Miler and Eric Shang. An update to be presented at Midwest Vascular Surgery Society Meeting in Chicago.
The patient is an 80 year old woman with lung cancer who was getting a port placed at her home institution. It was to be a left subclavian venous port, but when access was not gained, a left internal jugular venous port was attempted, but after the intitial stick and sheath placement, pulsatile bleeding was recognized and the sheath removed. Hemostasis was achieved with clips and the wound closed and a right internal jugular venous port was placed. The postprocedural CXR shown above showed tracheal deviation and numerous clips from the initial port placement attempt, and a CT scan with contrast (unavailable) showed a carotid pseudoaneurysm of 3cm projecting posteriorly behind the pharynx/esophagus. She was kept intubated and sedated, and transferred for management.
On examination, her vital signs were stable. She had 2cm of tracheal deviation and swelling was apparent at the base of the neck. While my trainees may be better versed at this than I at the particulars of this, my old general surgery trauma training kicked in, as she had a Zone I neck carotid injury, which in my experience is highly morbid despite how stable the patient was. Point again to trainees, this is no different from someone having stabbed this patient with a knife at the base of the neck. My options were:
- Open repair
- Endovascular repair from femoral access
- Hybrid repair
Open repair is the approach of choice for zone 2 injuries because aerodigestive tract injuries can also be addressed and the exposure is straightforward. For Zone 1 injury, the exposure is potentially possible from a neck exposure, but in my experience, jumping into these without prepping for a sternotomy puts you into a situation without a plan B. The exposure of the carotid artery at this level becomes challenging with hemorrage from the artery once the compression from the hematoma or pseudoaneurysm is released. A sternotomy in this elderly woman, while not optimal, may be necessary if open control is required, but the best plan is to avoid this.
This should be a straightforward repair from an endovascular approach, even with the larger sheath required for the covered stents. A purely endovascular approach is problematic for two reasons. One, cerebral protection devices are built for bare carotid stents and not peripheral stent grafts, but this is not prohibitive -it should be fine. Without a planned drainage, the hematoma would be left behind which could cause prolonged intubation and problems with swallowing -both an issue for an elderly patient battling lung cancer. Endovascular access could provide proximal control for an open attempt from above, but instrumenting from the arch in an 80 year old has a known 0.5-1% stroke rate.
A hybrid open approach with exposure at the carotid bifurcation offers several advantages. With control of the internal carotid artery, cerebral protection is assured while the carotid artery is manipulated. At the end of the procedure, the internal carotid can be backbled through the access site with the common carotid artery clamped. The hematoma could be avoided until the stent graft is deployed. An unprotected maniplation in the arch can be avoided. Once the stent graft is deployed, drainage of the hematoma can be performed.
This required setting up a table off the patient’s left that allowed the wire to lie flat to be manipulated by my right hand. The carotid bifurcation was accessed through a small oblique skin line incision and the common, internal, and external carotid arteries, which were relatively atherosclerosis free, were controlled with vessel loops. The patient was heparinized. The internal carotid was occluded with the loop, and the common carotid below the bifurcation was accessed and an 8F sheath with a marker tip inserted over wire. Arteriography showed the injury and pseudoaneurysm.
The location of the injury based on CT and on this angio would have baited a younger me into directly exposing it, but experience has taught me that which occasionally you can get away with it, the downsides -massive hemorrhage, stroke, need for sternotomy, just aren’t worth it. The sheath was brought across the injury and a Viabahn stent graft was deployed across the injury.
The hemorrhage was controlled and the hematoma was then exposed and drained -the cavity was relatively small and accepted the tip of a Yankauer suction easily. A Jackson-Pratt drain was placed. The access site was repaired after flushing and retrograde venting as described.
She recovered rapidly after extubation postop. She was able to breath and swallow without difficulty and had suffered neither stroke nor cranial nerve injury. The drain was removed on postop day 2.
The patient recently returned for a 6 month followup. Duplex showed wide patency of her stent.
More gratifyingly, her port was removed as her cancer was controlled with an oral regimen.
Let me start with my bias that all penetrating trauma should be approached in a hybrid endovascular OR. It is a natural setting for trauma and this case illustrates that. In a hybrid operating room, central aortic and venous injuries can be controlled endovascularly while open repair, including salvage packing, can be done. Excess morbidity of central vascular exposures can be avoided. Temporary IVC filters can be placed if indicated (becoming rarer and rarer). Cardiopulmonary bypass can be started.
In this patient, hybrid therapy brought the best of both techniques and avoided many of the pitfalls of the purely open or endovascular approach. For stable zone I penetrating injuries of the neck, it is clear that this is a reasonable approach.
Leriche Syndrome -one of those disease names that adds to our work in a way that an ICD codes and even the “aortoiliac occlusive disease” fails to describe. When I hear someone described as having Leriche Syndrome, I think about a sad, chain smoking man, unmanned, complaining of legs that cramp up at fifty feet, pulseless.
The CT scan will occasionally show an aorta ringed by calcium in the usual places that are targetrs for clamping below and above the level of the renal arteries. Even without the circumferential calcium, a bulky posterior plaque presages the inability to safely clamp the aorta. Woe to the surgeon who blithely clamps a calcified lesion and finds that the rocky fragments have broken the aorta underneath the clamp! The first way to deal with this is to look for ways not to clamp the aorta, by planning an endovascular procedure, but circumstances may necessitate the need to control the aorta despite the unclampability.
The traditional methods of avoiding clamping the calcifed peri-renal aorta are extra-anatomic bypasses including femorofemoral bypass and axillo-femoral bypass. I propose these following options for the consideration when the patient needs a more durable solution while avoiding a heavily diseased aorta.
Not Clamping I:
EndoABF does work to avoid clamping -these are common femoral endarterectomies supplemented by stenting of the aortoiliac segment, including in those with appropriate anatomy, a bifurcated aortic stent graft. This is often not possible to treat both sides, but one side is usually more accessible. Often, people will compromise and perform an AUI-FEM-FEM, but I have found the fem-fem bypass to be the weak link, as you are drawing flow for the lower half of the body through a diseased external iliac artery. The orientation of the proximal anastomosis is unfavorable and in the instance of highly laminar or organized flow, the bypass is vulnerable to competitive flow on the target leg, leading to thrombosis.
The femorofemoral bypass is the option of patients whose options have largely run out. It is made worse when fed by an axillofemoral bypass. Sometimes, you have no choice, but in the more elective circumstance, you do.
Not Clamping II:
The second method is performing a aorto-uni-iliac stent graft into a conduit sewn end to end to the common iliac aftery, oversewing the distal iliac bifurcation.
The conduit is 12mm in diameter, the key is to deliver the stent graft across the anastomosis, sealing it. The conduit is then sewn to the side of a fem-fem bypass in the pelvis, maintaining antegrade flow to both legs. The other option is to sew the conduit to a 14×7 bifurcated graft. Illustrated above is this 12mm conduit sewn end to end to the diseased common iliac artery with wire access into the aorta and a aorto-uni-iliac device. Typically, a small AUI converter (Cook, Medtronic) can be used, but the aorta is often too small even for a 24mm device, and an iliac limb with a generous sized docking segment (Gore) ending in a 12mm diameter fits nicely. Below is a CTA from such a case, where the stent graft is deployed across the anastomosis, sealing it off from anastomotic leaks (exoleaks).
Not Clamping III:
Often, the infrarenal aorta is soft anteriorly and affected only by posterior plaque at the level of the renal arteries. While a clamp is still not entirely safe (I prefer clamping transversely in the same orientation as the plaque with a DeBakey sidewinder clamp), a balloon is possible. I do this by nicking the aorta -simple application of a finger is sufficient to stop the bleeding if you have ever poked the ascending aorta to place cardioplegia line.
A Foley catheter is inserted and inflated. The Foley’s are more durable and resist puncture better than a large Fogarty. This is usually sufficient for control, although supraceliac control prior to doing this step is advised. The aorta can be endarterectomized and sewn to the graft quite easily with this non-clamp.
This has worked well, Although pictured above with an end-to end anastomosis planned, it works just as well end-to-side. I actually prefer end to side whenever possible because it preserves the occluded native vessels for future intervention in line.
The Non-Thoraco-Bi-Femoral Bypass
The typical board answer for the non-clampable aorta is taking the inflow from the thoracic aorta or from the axillary artery -neither of which are good options. The first because the patient is positioned in right lateral decubitus and tunneling is not trivial. The second because of long term durability. The supraceliac aorta, technically it is the thoracic aorta, is often spared from severe plaque and clampable. Retropancreatic tunelling is straightforward, and a 12 or 14mm straight graft can be tunelled in this fashion from the lesser sac to the infrarenal retroperitoneum. It then sewn to the supraceliac aorta and then anastomosed to a 12x6mm or 14x7mm bifurcated aorto-bifemoral bypass, of which limbs are tunneled to the groins.
This worked very well recently, allowing a middle aged patient with severe medical problems, occluded aorta and iliac arteries, with critical limb ischemia, survive with minimal blood loss and home under 5 days. It delivers excellent flow to both legs in an antegrade fashion. Dr. Lew Schwartz gave me a list of references showing that this is not novel, but represents a rediscovery as the papers were published in the 80’s [reference], and buttresses the principle that innovations in open vascular surgery are exceedingly rare, largely because we have been preceeded by smart people.
Conclusion: All of these come about through application of some common sense and surgical principles. The most important this is that the aorta is the best inflow source and reconstructing it with the normal forward flow of down each leg and not reversing directions as in a fem-fem bypass gives each of these options a hemodynamic advantage.
References for Supraceliac Aorta to Lower Extremity Bypass
The principles of salvage are in rescuing valuable undamaged goods in the setting of catastrophe. This guided me when a patient was flown in from an outside institution to our ICU with a saline soaked OR towel in his right groin -he had had an aorto-bifemoral bypass for aorto-iliac occlusive disease a year prior, but had never properly healed his right groin wound which continued to drain despite VAC therapy and wound care. On revealing his groin, this is what I saw:
A CT scan was sent with the patient but has been lost to time, and it showed a patent aorto-bifemoral bypass send flow around an occluded distal aorta and iliac arteries. The graft did not have a telltale haze around it nor a dark halo of fluid which signaled to me that it was likely well incorporated and only sick in the exposed part. The patient was not septic, but had grown MRSA from the wound which was granulating from the extensive wound care that had been delivered.
I felt that it would be possible to move his anastomosis point more proximally on the external iliac in a sterile field (figure above), and then close, then endarterectomize the occluded external iliac artery after removing the distal graft, then after vein patching, cover the repair with a sartorius muscle flap. It would salvage the remaining graft and avoid a much larger, more intense operation which was plan B. To prepare for that, I had his deep femoral veins mapped.
The patient was prepped and draped, the groin was excluded by placing a lap pad soaked in peroxide/betadine/saline solution (recipe for “brown bubbly” liter saline, a bottle of peroxide, a bottle of betadine), and covering with an adesive drape. The rest of the abdomen was then draped with a second large adhesive drape. A retroperitoneal (transplant-type) right lower quadrant incision was made (below) and the external iliac artery and graft were exposed. As predicted on CT, the graft was well incorporated.
The external iliac artery was opened and focally endarterectomized of occlusive plaque (image below). The adventitia had good quality despite the longstanding occlusion.
The graft was mobilized and transected and anastomosed end to side to this segment of artery (below). Dissecting was made difficult by how well incorporated it was.
The wound was irrigated (with brown bubbly) and closed, dressed, and sealed over with the adhesive drape. The groin wound was then revealed and the graft pulled out (below).
Remote endarterectomy using a Vollmer ring was used -in this case I didn’t use fluoroscopy given the short distance to the terminus of the plaque which i had mobilized in the pelvis.
The plaque came out easily and was not infected appearing. It is shown below ex vivo.
A segment of saphenous vein was harvested from the patient medially and the arteriotomy was patched. The sartorius muscle was mobilized and applied as a flap over this. The wound was irrigated with brown bubbly and packed open with the intention of VAC application.
The patient healed very rapidly and remains infection free. I had used this approach on several occasions in the past and twice more recently. It truly is salvage as it preserves the uninfected graft while never exposing it to the infection in the process of operating. It avoids having to remove the whole graft which then damages the left side -I have seen other surgeons take this approach elsewhere taking a all-or-nothing approach to graft infection to considerable morbidity to the patient. It avoids having to harvest deep femoral vein -another large operation to which the body responds truculently. The patient recently came by for his 4 year followup, still smoking, but legs preserved.
Followup At 4 Years
The patient came back in followup -it has been 4 years since his infection was repaired. He was complaining of short distance claudication. His wound healed well and remains closed. CTA shows along with his short segment SFA occlusion which we will treat, a widely patent R. EIA (below).
The remote endarterectomy of the external iliac artery remains patent. Compare this to the preop CTA which I found and wasn’t available when I posted this case originally:
The chronically occluded EIA can be readily seen. The artery shown in the current CTA is that recanalized artery.
Patient is a 77 year old man with history of HTN, hyperlipidemia, former smoking, and CAD with CABGx5 and bilateral lower extremity bypasses who developed unstable angina consisting of neck and throat pain. He underwent catheterization at an outside hospital and found to have 100% LAD occlusion, a diseased, small patent left main and left circumflex (the profunda femoral artery of the heart!), 100% RCA occlusion, a patent but diseased SVG to distal RCA, and a patent LIMA graft to distal LAD but with severe plaque and near occlusion of his proximal left subclavian artery.
He had an NSTEMI. His vitals signs stabilized in the coronary care unit and he was sent to a telemetry floor. Whenever he walked, he would get the jaw pain, and this would also occur sporadically while recumbent.
On examination, he had no left brachial pulse, only a monophonic signal there, and bounding femoral pulses where there were the origins of bilateral femoral-tibial bypasses. His radial artery pulse was diminished on the right and absent on the left. Both saphenous veins had been harvested as were arm veins for the left leg bypass.
CTA shows the left subclavian artery to be occluded at its origin.
Cardiac surgery, interventional cardiology, and vascular surgery were called in for consultation. Cardiology consultation (Drs. Kapadia and Shisheboor) felt, and I agreed, that the left subclavian lesion was a poor candidate for recanalization and stenting. CT Surgery (Dr. Faisal Bakaeen) and I had a long discussion regarding alternate conduits, as he had unknown radial but likely radial artery disease, and had all usable veins previously harvested. I brought up a free RIMA graft -I had worked with Dr. Daniel Swistel, in NYC as a resident, who was Dr. George Green’s protege, and as a medical student at P&S I scrubbed Dr. Green’s final cardiac case. He routinely performed bilateral ITA bypasses decades before all-arterial revascularizations were routine. I get enthusiastic talking about cardiac disease! Walking through all the options -does anyone use deep femoral vein as coronary bypass conduit -we agreed ultimately that the best option would be a carotid-subclavian bypass with plenty of backup.
At its heart, it would be this vascular surgeon’s attempt at an off-pump single vessel CABG (above). Preparations were made with cardiac anesthesia and cardiac surgery to place an IABP (intra-aortic balloon pump) if he became unstable. For my part, the operation was straightforward, but I was going to have to go about it efficiently. I also figured that with a clamp beyond the LIMA takeoff, no significant change would occur to the coronary flow from the LIMA graft. So I hoped as I worked very deliberately. We kept him on the hypertensive side during the case.
The operation went well. The patient’s angina resolved and a followup CT showed the patent bypass feeding the LIMA and LAD.
His resting angina resolved. He followed up a month later and was very pleased. Moreover, he had a brachial and radial artery pulse and a general weakness of the left arm that he never complained about before lifted.
The carotid subclavian bypass is something that really needs to be in the armamentarium of a modern vascular surgeon. Though out of print, Wylie’s Atlas (the unabridged, multivolume version) is available used through online sellers, and is useful for elucidating the anatomy which boils down to avoiding cutting the important structures -the phrenic nerve, the vagus nerve, the brachial plexus, branches of the subclavian including the vertebral artery, while cutting away muscles -lateral head of sternocleidomastoid, any part of the omohyoid, the anterior scalene muscle. And dividing the lymphatic duct if encountered. And tunneling under the jugular vein. And minding the buttery fragility of the SCA. The best technical paper out there is by Dr. Mark Morasch and it mostly deals with carotid-subclavian transposition (reference 1) but has excellent figures on bypass as well. I do both transposition and bypass, but for brevity, I prefer bypass.
This is not a unique problem, having been reported in the literature. An unusual variant of this is coronary sbuclavian steal syndrome (reference 2), which refers to reversal of flow in the LIMA bypass in the setting of subclavian artery occlusion and left arm exertion -which was not the case here, but interesting enough to mention. Here, it was a straightforward case of managing the hemodynamics. The key point of operating on such a patient was having the surety of quick response in the case of ischemic heart failure -we operated in the cardiovascular operating rooms with rows of perfusion pumps and balloon pumps and VADs and ECMOs at the ready. Indeed, this result could not have been so straightforward and routine seeming without the combined effort and experience of the whole Heart and Vascular Institute from nursing to consultant staff.
- Morasch MD. Technique for subclavian to carotid transposition, tips, and tricks. J Vasc Surg 2009;49:251-4.
- Cua B et al. Review of coronary subclavian steal syndrome. J Cardiol. 2017 Apr 14. pii: S0914-5087(17)30090-4. doi: 10.1016/j.jjcc.2017.02.012. [Epub ahead of print]