Blood Pump Speed versus Fistula Flow

Dear John,

Before I get to the questions, first some history…my wife has been undergoing haemodialysis for just over 2 years with the last 11 months on HHD using the NxStage Cycler. She has an upper left arm AVF which has been generally performing very well except that she has a problem with a recurring stenosis in her shoulder area, the first occurrence being just before she started HHD. This has been subjected to a fistulaplasty procedure on three occasions to date and she is now awaiting an appointment for a fourth. Following her last fistulaplasty in December last year her Transonic result showed 1190 ml/min but this has now reduced to 760 ml/min (36 % reduction) last month.
Last August I read with interest your article on Blood Pump speed and your Dialysis Fistula dated 05-02-2008 and despite the NxStage training focusing on achieving 450 ml/min I have, since December, limited the pump speed to 1/3 X fistula flow and have been running at 380 ml/min as I felt that this may have had some bearing on the stenosis issue. Based on the current situation, however, this does not appear to be the case. Can you add any comment about this ?
I have continued to run at 380 ml/min because, as you are probably aware with the NxStage, and running with renal unit type speeds of around 300 ml/min results in unacceptably long sessions. Interestingly, even with her current fistula flow and running at 380 ml/min my wife’s arterial pressures remain at -120 /130 mm Hg and her venous pressures at 180/190 mm Hg. Prior to her fistulaplasty last December a clot was daignosed within the fistula and since that time her Consultant Nephrologist has prescribed Warfarin (to maintain an INR of 1.5 - 2.5) and I can only believe that this has prevented yet another clot formation resulting from the lowered “flow between needles”. 380 ml/min results in a dialysis time of around 3.5 hours with excellent urea and creatinine clearance plus good phosphate and potassium control on a 5 days per week schedule. Hence I’m a big fan now of slower (within reason) dialysis. We never remove any more that 1.5 L of fluid and typically <1.0 L on successive day sessions (dry weight = 79.5 kg) as my wife is very careful with her fluid intake.
My question is…have you conducted any more recent investigations into “flow between needles” and do you have any indications as to the absolute minimum acceptable ?

Dear Michael …
I hope you don’t think I am ignoring you question - I will answer it early this coming week … i am rather flat out this weekend, so I hope you can be patient another day or two

Absolutely not a problem John, I appreciate your quick response anyway.

Dear Michael

Thank you for your excellent question and the detail you have provided – it makes it easier to think through an answer.

The trite response to the last two questions is ‘no’ and ‘no’ … sadly, our knowledge of fistula flow has advanced little (if at all) and I really don’t know the answer to the second question.

That said, there are still some useful observations I can make: I note that your wife’s fistula has needed several (now 3) angioplasties (or fistuloplasties) for a recurring shoulder-region stenosis. This is the crux of her problems. Though not true of all venous stenoses, the vast majority of veins, once they have stenosed, will always re-stenose after angioplasty. Though the angioplasty may provide temporary respite and re-open the vessel, the fibrous scar tissue in the wall of the vein that induced the initial stenosis will, in the main, contract and narrow the vessel again. The mean functional window is usually measured by a few months with the interval between each successive ‘plasty’ and the subsequent re-stenosis shortening with each attempted repair.

Stents can be used … but our (and others) experience of stents – especially in the shoulder or within-chest-cavity narrowings – has been broadly dismal as re-stenosis (either within the stent or at one or both ends of it) is all too common. One simply tends to end up with a chain of stent upon stent … never a nice result. The cardiologists use what are called ‘drug-eluting’ stents – stents that have been impregnated with a variety of agents (drugs) designed to inhibit endothelial cell proliferation … NB: a stent commonly excites a proliferative response in the cells (they are called endothelial cells) that line the inside wall of the blood vessel (in this instance the arterialised vein) which leads to vessel narrowing and re-stenosis. While there has been some recent experience with this type of stent in AV fistula stenoses, I must say that it has been - to the best of my knowledge - with rather disappointing results to date.

Our preferred ‘fix’ – wherever possible – has been to opt for a surgical fix. We are blessed, here at our service, with some ‘inventive’ and excellent vascular surgeons who have made it their business to be helpful and thoughtful about surgical approaches and vessel options.

That said, it is clearly easier for a surgical resolution if the stenosis is outside the chest cavity. When this is the case, an elliptical length covering the area of the stenosis is excised from the vein wall and is ‘matched’ by a subsequent in-sewing of an vessel-expanding (larger) elliptical patch using a piece of vein that has been ‘harvested from elsewhere (eg the leg). This commonly offers a very good (and permanent) outcome.

Sometimes, a gortex graft interposition or a mobilised external jugular bypass may be feasible alternative (or better) options.

One thing we do believe in here is that surgical solutions, when and where possible, seem to offer better long-term outcomes for fistula patency and functional integrity than do repeated angiolasties and/or stents. In our view, a fistuloplasty merely offers breathing space to consider if and what more permanent surgical solutions can be found. And remember … there is usually the other arm for a new AVF or the potential for a leg AVF … … oh, and yes, we have had one patient (now in his 80s) who has been self-dialysing at home (alone and without a carer) via a femoro-femoral AVF for over 7 years now.

You have noted that a recent month-on-month Transonic measurement has shown (again) a fall in fistula flow from 1190 to 760 ml/min … a 36% reduction in one month. As a 25% reduction over 3 months is viewed as significant, this is a fast fall indeed. It is, I am afraid to say, a clear indication – along with the venous pressure you mention of 180-190 – of a re-stenosing vessel … and fast. Another angioplasty will only temporise. Given this data, I would be seeking a surgical resolution, not a radiological one.

While I think that high in-fistula flows may have some bearing on stenosis/re-stenosis rates - as was discussed in the article I wrote here at the HDC site on ‘Blood Pump speed and your Dialysis Fistula’ on 05-02-2008 - in the case of your wifes’ AVF, the ‘once it’s stenosed, it’ll re-stenose’ rule takes clear precedence over an potential protective effects that might accrue from lowering blood pump speeds. Lowering the pump speed may buy a smidge of time – but, if at all and at best, only a smidge. It’s not the answer … see above re a surgical solution!

It s true that not all stenoses are in easy (or safe) reach of the vascular surgeon without creating major in-cavity chest surgery. However … and you don’t give a clear indication of the actual site of your wifes’ stenotic lesion … most are ‘reachable’ with thoughtful planning and fall-back options.

As a general comment, I think that lower blood flow rates may play a role in reducing initial stenosis rates, particularly when the stenosis is at or near the venous needling site(s) where flow-return turbulence disturbs/distorts the normal laminar flow of blood within the vein and the eccentric sheer force of the returning blood flow jets against the vein wall and acts on the vessel wall to create local damage to the lining endothelial cells. Importantly, there is some research data to support this view.

HOWEVER, back to the central point: once a stenosis has occurred, the ensuing trouble then is most commonly (1) recurrent and (2) likely independent of flow and (3) dependent on local factors like ‘shrinking’ or ‘scarring’ fibrosis in the vessel wall. Researchers have looked at heaps of potentially helpful agents: fish oil, anticoagulants, aspirin, ACE inhibitors, anti-fibrin agents - the list is long - but none reliably help.

In addition, while stenoses distant from needling areas are less likely to be flow-induced by the return needle, stenoses are actually quite common in the shoulder region where the veins must pass under or through the bony structures of the chest wall to become ‘intra-thoracic’. Kinking or sharp angling of veins in this region, especially if accompanied by high fistula-driven flows, may again play some role in the genesis of stenoses at this more distant site from the needling area through the disturbance of flow patterns and turbulence created by a mixture of high flow within a ‘geometrically challenged’ vessel architecture.

As for your point about the potential for clotting in the interval section of the AVF ‘between’ the arterial and venous needles, and the relationship of this clotting potential to any reduction of flow in this interval-section … well, that’s a mighty interesting issue which I discussed in the same HDC article from 2008 you mentioned in your question. That clotting might potentially be more likely when a high flow ‘draw’ from the arterial site is demanded by a machine set with a high arterial flow (Qb) … is largely conjecture – but it is conjecture that makes some sense.

If the AVF has an natural flow rate of, say, 800 ml/min and the ‘arterial’ blood flow (Qb) is set at 450 ml/min, then 800 – 450 = 350 ml/min should remain to flow through the interval section of the AVF while 450 ml/min is ‘drawn’ from the fistula to provide the Qb of 450 that has been set on the machine. At the venous return needle, 450 ml minus any volume that is UF’ed would be returned to then reconstitute full fistula flow toward the heart with the 350 ml/min that has traversed the interval segment of the fistula.

I know that reads messily … but I hope you get the concept.

If natural fistula flow is 800 ml/min (ie: the same as in the example as above) but the machine is set to draw only a Qb of 225 ml/min, 800 – 225 = 575 ml/min would flow through the interval segment to reconstitute with a venous return of 225 ml/min minus any UF … and then to flow on to the heart.

Common sense dictates that if clots were to form in the interval segment, they might be more likely to do so in the lower interval segment flow example. It’s a matter of flow physics and probability.

Lowering the blood flow rate (Qb) ‘stolen’ by the machine from the fistula’s natural flow should increase the interval segment flow and diminish the chance for clot formation. Hence the ‘theory’ that clots are less likely to form in the AVF interval segment when machine blood flow rates are low and interval segment flows are high – or, conversely, that they are more likely to form when the machine Qb is high and interval segment flows have been diminished.

This vessel ‘physics’ excludes any consideration of other possible contributors: like ESAs, blood viscosity, medications etc – all of which (and more) can play a role. While I understand her having been put onto warfarin (and, yes, we have done that on occasion too), we try to avoid warfarin in dialysis patients as much as possible – wary as we are (and should be) of bleeding issues, herald bleeds, calciphylaxis etc. Nevertheless, I can sympathise with the thinking of the nephrologist who added it.

Michael – I have ‘rabbited’ on here quite a bit … and, despite that, I am not sure that I have answered your questions.

Have ‘we’ done further research in this fraught area … no!

Should ‘we’ – or someone … yes!

And, as regards a minimum acceptable flow through the interval segment … no, I do not have a rule of thumb for this either.

They are great questions, Michael … but little knowledge + even less science + a bunch of waffly answers makes for one very embarrassed and inadequate ‘expert’, who wished he could answer you better.

Hopefully, someone will take up these issues and sort it all out before too long. Sadly, I doubt it will be me as my time for such painstaking and careful work is long past.

Many thanks for your very comprehensive reply John. Based on my engineering background, I fully comprehend your assertions and have supported them from the first time I read your article. It’s unfortunate that there continues to be a dearth of controlled procedure test data but changing local protocol without such data is an impossible task. I will, however, continue to press whoever I can with the need to have a proper study conducted in the UK.

Your comments on recurring stenoses and stenting generally agrees with what we have been told by our nephrologist and I am attempting now to escalate the discussion to vascular surgeon level in order to have a resoloution going forward.