Thanks for the post. Nice case. It is not uncommon to see the outflow cephalic occlusion and flow maintained via the collaterals to the basilic system. Agree that as long as the flow is maintained, the AVF does not need to be abandoned.

Do you have the forearm picture to show the occlusion?
Also, you know that the central stenosis will return soon. What would be your approach then- stent?
thanks
Anil Agarwal

The central stenosis as shown has been described as vascular thoracic outlet syndrome at the costoclavicular junction. Stents have been shown to fracture or be occluded due to the bony structures comprising the CCJ. First rib resection followed by stenting when necessary has been well-described for this kind of lesion.

These are all pictures saved by the technician. In the first picture the vein in the middle is the cephalic vein occluded at the level of the elbow.
In central lesions if the re-stenosis post dilation is less than 30% and symptoms improve I don't place a stent. Sometimes I bring the patient the following month for angiography to reevaluate the lesion. If it has more than 30% restenosis and symptoms are severe I place a bare metal stent.
Thoracic outlet syndrome is most common on the left side and tends to affect more the nerve and left subclavian vein. Sometimes it affects the left subclavian artery. Common causes are presence of a cervical rib or hyperthrophy of scalenes muscles. I don't think this the case of this patient

I could see the upper arm cephalic vein occluded at the elbow.
You are not completely accurate as far as Vascular TOS. There is neurologic TOS and there is Paget-Schroetter Syndrome. But, there is a body of literature looking at this in dialysis patients. 3D CT central reconstructions have shown compression of the CCJ due to the clavicle and the first rib (not a cervical rib).
Check out this paper. I can email it to you as I cannot attach it.
Semin Vasc Surg. 2011 Jun;24(2):113-8.
Management of central vein stenoses and occlusions: the critical importance of the costoclavicular junction.
Illig KA.

TOS is something that we have to think about it, but the frequency of venous TOS is about 2-3% of the total of cases of TOS. I think that most of the central lessions in HD patient are secondary to catheters. In venous TOS the abnormality is the costoclavicular space ( extrinsic compression under the costoclavicular ligament). In arterial TOS is the presence of cervical or anomalous ribs.
If we think that the patient has venous TOS we can do a venography with hyperabduction maneuver and confirm it

I am not certain where the 2-3% frequency of venous TOS has been seen. There have been no good studies in dialysis patients for the VTOS at the CCJ as far as the frequency. It is difficult to extrapolate from the rest of the population. When pta is performed at the CCJ, the compression can many times be seen. We have been finding that first rib resection has much improved symptoms and fewer interventions. Stents placed at the same area are not effective. Of course, we need a multi-center trial to look at more patients.
Ann Vasc Surg. 2015;29(4):698-703. doi: 10.1016/j.avsg.2014.11.021. Epub 2015 Feb 24.
Aggressive Costoclavicular Junction Decompression in Patients with Threatened AV Access.

Interesting case with good discussion here – it’s great to have this forum up and running.

The reality here is that you got a great result with resolution of arm edema post angioplasty; however, this will likely be a recurrent lesion. Continued angioplasty will be reasonable, until you no longer have durable results.

While 1st rib resection is a good idea in cases of outlet syndrome, the practicality of actually putting an ESRD patient with multiple comorbidities through this surgery is not without its drawbacks. Additionally, most communities do not have access to surgeons with expertise in dialysis access associated thoracic outlet syndrome.

Stenting is a reasonable choice when angioplasty fails. While there is a chance of stent-fracture with thoracic outlet syndrome, we’ve never experienced any significant consequences with this nature of stent fracture. We use the Bard Lifestar stent when using bare metal centrally – and I believe it is more flexible than some of the older, more rigid bare metal stents.

When aggressive, recurrent bare stent restenosis occurs despite angioplasty, reinforcing the initial bare stent with a 2nd bare stent often helps to provide a more durable result.

Alternatively, we have seen excellent long-term patency with more flexible stent-grafts (particularly the Viabahn) at this junction. The only concern of a stent-graft here is need to jail off the IJ vein – but this may be warranted if there are no other adequate options to manage arm edema.

We’ve had several cases where stent-grafting here has led to permanent results (over 4-6 years), with no recurrent arm edema and maintained access patency. This has not been studied, but I suspect the same advantages of stent-grafts at the venous anastomosis and at the cephalic arch also apply to this type of lesion.

Also, we have never seen a Viabahn stent-fracture here, likely due to its relative flexibility.

Admittedly, we have not always teased out the etiology of these lesions as thoracic outlet vs catheter-associated stenosis – but with endovascular stent-graft management, we can usually achieve long-term, durable results.

*One caveat on central stenting – give yourself enough of a landing zone for a stent or stent-graft here, and go big (usually 13mm Viabahn / 13.5 Fluency, or 12-14mm bare stent if you’re doing bare metal) with a wire secured at the IVC. You don’t want any migration here.

I agree with the follow up. I have had many bare metal stents (BMS) fail in the central veins. Even with re-stenting them with a BMS, I have had failures. The Viabahn does work better, but I have had it fail too. I also caused a symptomatic RIJ vein DVT once using a Viabahn across the CCJ. The patient did well overall and did not lose his access (the centrals veins were occluded despite prior placement of a BMS). Also, I totally agree with an escape wire in the inferior vena cava. I once had a stent migrate when I tried to place it across a completely occluded SVC despite prior angioplasty. The stent fell into the patient's right atrium on the guidewire. I was able to move it to the IVC with much effort. Then, I stented the SVC again (making certain to deploy the stent a bit more cephalad). The patient did well.

dear Dr Gabbard,
thanks for the info and a wonderful discussion about the VTOS.. i went through both the papers you had recommended. It looked like subclavian vein is the one affected by the VTOS whereas the lesion in this case is at the innominate. would you still consider VTOS here ? in other words is it the location of stenosis at/near CCJ which defines VTOS or is the vein affected (subclavian)?
thanks

Great discussion. Have ahad a few of these and 1st rib resection has only given temporary relief.
Re-fibrosis around the site I suspect causes the issue.
What was the flow in the AVF?
We have found reduction of inflow is helpful in higher flow cases.
If the thoracic inlet "stenosis" is actually a decent size (it could be 6mm and still be a relative stenosis of <50%) then reducing inflow might be the answer rather than focussing on the TO/TI

This has been a good discussion.
I agree with the flow mismatch. I have found that banding or surgical plication of the "mega-fistula" to be quite useful in this group of patients.
As far as where the lesion is with CCJ compression and VTOS, I have found it to be variable and depending on the anatomy of the patient. I have had some broad-shouldered, stout patients to have compression quite lateral (surprising, but was shown to be the case with imaging and at the time of surgery). I have also seen cephalic arches terminate medially at the CCJ and be compressed so severely that the fistula would thrombose (despite stenting). Decompression of the CCJ ceased a q6 week thrombosis rate to an angioplasty every six to nine months. I also agree that many times the area is heavily fibrosed and damaged due to the compression. Many times, stenting is required in the future and appears to have longer patency rate (although we are still looking at this). One issue I have seen that is quite interesting.... That of pain. When the CCJ is stented, I have found patients develop significant pain if there was severe compression. When the CCJ is decompressed, the pain resolves. Some patients have said that the pain had resolved in the PACU. Others had said that for the first time in years, they had been able to raise the arm over his/her head to comb his/her hair. Also, Dr. Illig noted that the morbidity was much less with the dialysis patients. Whereas patients with Paget-Schroetter Syndrome required weeks of physical therapy, the dialysis patients usually had improved symptoms within days.

Great case and interesting discussion - I could spend all evening .....
(1) I would like to see more images. I am not sure I see the "basilic system" - the runoff seems to be to the brachial and deep brachial veins.
(2) mention is made of the innominant as the "culprit lesion", but the pre-dilation image doesn't show much in the way of collaterals, and is not much different than the post dilation image. Where is the culprit? We have all seen an impressive "Medusa's head" of collaterals disappear when a critical stenosis is opened, but I don't see that here.
(3) regarding stenting versus first rib resection - my local thoracic surgeon is just not enthusiastic about a first rib resection in the face of venous hypertension and dilated fragile collaterals veins, I don't know why - except that I remember feeling shaken and lucky to have a living patient after changing my scrubs twice doing a graft revision in the face of thin-walled collaterals. My experience with over 200 central venous stents has been presented at VASA, showing that long term secondary patency (some tending is required) is 90 percent at five years. I have never (NEVER) seen a Viabahn stent fracture. I have seen all kinds of BWS fracture, and I will say that the only reason to use a BWS is if it is the only thing you have at hand - then you go back and re-line it with a covered stent.
(4) as for worrying about jailing out the IJ, most often the IJ is already walled out by the obstructive process - restoring any drainage is beneficial.
(5) the remark about flow is very cogent - we get patients all the time who have been dilated over and over again, with fairly unimpressive images (no Medusa), and who prove to have a dilated feeding artery, an over large anastomosis, and flows from 3000 to over 5000 cc/min. The problem is not outflow stenosis, it is a firehose fistula. I suspect this might be the problem here. Flow reduction is mandatory for acceptable longterm results.
Plication is temporary, as the fistula quickly regrows. Banding is tricky to calibrate, and prone to failure. Most oftern, we start with an aneurysmoplasty of the leading segment to debulk the fistula, which does not affect flow enough, then do a 8mm PTFE banding, tightening it in stages until flows in the 1200-1500 cc/min are attained.
Often, the only answer to a firehose fistula is to take down the fistula with its 1 cm diameter anastomosis and proximalize or distalize the inflow with a piece of tapered PTFE.
(6) Finally, flow reduction has it's risks - dropping demand from 5000 to 1200 cc/min in a hypertrophied artery can get you a thrombosed artery ,- I have seen several cases in flow reduction or after ligation post transplant. That's why I prefer distalization to proximalization.
Interesting conversation - thank you - I have enjoyed exposing my ignorance and arrogance

These are great points. I think the best statement is to what is best for your patient in the area where you work. Dr. Illig literally wrote the book on VTOS. His first rib resections usually take about 45 minutes with a one or two day hospital stay. Collaterals can be a problem but also provide vein for a patch angioplasty. The surgery has to be done with imaging during the operation. Stenting is many times still necessary, and I think the Viabahn is the best.
What is interesting here is that a long-segment plication is done for as much of the entire length of the access as can be accessed. Also, if needed the anastomosis is also plicated or reimplanted. We monitor blood flow volumes to look at the difference. Using the graft segment is fraught with problems from our local surgeons. Unless a PAI type of arterial inflow is used, the small segments of PTFE have continued issues with anastomotic stenoses and graft narrowings. It has not been a good fit. I find that the surgeons do it to not perform the extensive plication that is needed.

Also, I have seen a Viabahn fracture. Usually, this is due to repeated trauma or pta. But, at the CCJ when there is severe VTOS, I have seen the Viabahn fail even with re-stenting with a second Viabahn. Now, I am not advocating first rib removal for all patients. I have had some patients with CCJ stenoses that responded to pta alone with great long-term results. Additionally, no one has discussed claviculectomy for VTOS. Dr. Illig has done this when we have had to rebuild the central veins due to subclavian vein, innominate vein, or CCJ occlusions that could not be crossed endovascularly. Sometimes, the internal jugular vein was turned down. Other times, prosthetic was used. And other times, the guidewire was physically pulled through the occlusion after patch angioplasty when the vessel could be seen after claviculectomy. Usually, the vessel was stented after such a procedure. The results have been quite good. Several patients have dialyzed with their dialysis accesses for years after the central reconstructions. Cosmetically, I found that the patients really looked no different. I actually had to palpate the area of the clavicle resection to notice the difference. What was most important was that these patients did not need to have a thigh access created. The outcomes were durable. But, they did not continued follow up. Still, I found the maintenance procedures usually were only two to three times per year.

Great discussion
1. I dont have more pictures. I don't know why there are no more collateral veins. It could mean that the swelling was detected quickly and there was not enough time for them to develop. If you have enough collaterals the swelling will disappear by itself.
2. I think that there is significant improvement of the lesion after angioplasty. You can compare how the contrast runs before and after angioplasty and you can see it flows much better.
3. I don't think that this patient has firehose fistula, but even if he does how can you have 3-5 liters/min going throught collaterals vein? , when you have big and multiple collateral veins , it can be view as capillaries ( increase cross sectional area decrease velocity/pressure of blood flow) . The collateral veins develop to decrease pressure and restore flow (like varicose veins)
4. I talked with 2 of the vascular surgeons in my hospital about central veins stents versus angioplasty. They told me that they have been doing central vein angioplasty alone in HD patients for a long time with great results. Is everyone placing stents in all central lessions? In other words, is angioplasty alone not correct?

I think if you have a patient with arm and face swelling and central vein stenosis, then that is the culprit lesion almost invariably (collaterals or no collaterals). I have had one patient who the surgeon ligated the lymphatics during graft creation. Despite wide open central veins, I could never get the swelling to resolve. Even when the fistula thrombosed, the swelling never resolved.
As far as stenting in the central veins, it depends on the lesion. If the narrowing is occluded or almost occluded, and the operator has to use a great deal of effort to cross the lesion, then stent it. Otherwise, angioplasty does work. Many times there are diminished returns. If the patient requires angioplasty more and more frequently or the symptoms recur quickly, then stenting or CCJ release (for those particular lesions) may be needed. Unfortunately with stenting or pta, we will lose the battle with some of these patients. The lesions will occlude and cannot be crossed. Follow up is a must.