The Complete Line of Reasoning Behind the Low Volume Brachial Plexus Nerve Block (not for the Faint of Heart)
The idea of the Low Volume Brachial Plexus Block came out of the early work on the Phrenic Sparing Interscalene Block PSIB (link on this site) project. Early on it became apparent that the phrenic nerve was coming into contact with the local anesthetic because the volume of local anesthetic injected for an interscalene was more than could be kept in the interscalene space after injection. Many times, on ultrasound, we have seen the local anesthetic volume spreading out over the “tops” of the scalene muscles while we were still injecting. Performing an ultrasound scan a few minutes after completing the block would reveal that any volume remaining between the scalenes at the end of the block would have dissipated. This was presumably not because it had all absorbed into the plexus structures. Again, most of the volume of the block would have found a natural plane around the most superficial fascial borders of the scalenes to be the easiest path and would have “leaked” out that way. The block would have been better called the “circum-scalene” block.
It seemed obvious that since the majority of people carry their phrenic nerves on the anterior side of the anterior scalene, the normal path of local anesthetic volume spread was going to block the phrenic nerve as well.
Well that’s one aspect; the corollary to that conclusion is that if the majority of the brachial plexus anatomically passes between the scalenes, then the volume that ran off over the scalene tops was not participating in establishing or continuing the block. In other words, that extra volume was not just wasted in terms of the block; it was causing undesirable side effects.
So the lesson there is easy enough, find a way to lower the volume injected and keep it between the scalenes. A little thought resulted in the block technique in the Neuraxiom article “The Perfect Interscalene Block” (Link here), in which the following principles guided the block;
- 1. Minimize the number of needle passes to keep the spots for leakage to a minimum
- 2. Lower the volume injected in any given spot to that amount which seemed to fit comfortably
- 3. Surround the brachial plexus nerve bundle with small volumes
- 4. Place a small amount directly into the nerve bundle between the nerves
In the “Perfect Block” technique, 5mls were placed between the bundle and the anterior scalene, 5mls between the bundle and the middle scalene, 3mls were placed directly into the nerve sheath. The block worked well and there was a minimum of extra volume to be displaced, no phrenic involvement, we examined the diaphragms with M-mode ultrasound, everything was good. Still I’m sure some of the 5ml aliquots placed on either side did spread out around the scalenes because the scalenes are muscles and when they contract, they squeeze the interscalene space and force out any left-over volume.
It was during the “Perfect Block” that the little light-bulb came on, with the thought, “Why are we injecting around the bundle at all?”
Okay, new project; lower the volume and inject directly into the brachial plexus bundle sheath under ultrasound. We’d already done this to some degree in the supraclavicular block with somewhat larger volumes. How low a volume was appropriate? Fortunately, a few days later I was called to assist with an interscalene block in recovery room for a patient who had refused the block pre-operatively. The anesthesiologist was amenable to trying the “Perfect Block” technique described above (5ml front, 3ml in the sheath, 5ml behind). The approach was fairly low interscalene. The 3mls was injected slowly with the needle tip positioned visibly among the nerves in the sheath, we could easily see the sheath swelling up with the local anesthetic and the nerves moving away from each other as the space expanded. Within perhaps 5 seconds, the patient remarked that her eye felt funny. A quick look revealed a Horner’s syndrome had started.
This seemed very significant as it indicated the volume capacity and state of compliance of the brachial plexus sheath in the neck of this particular patient. It meant that if you can inject 3mls of solution into the brachial plexus sheath at the base of the neck and the solution reached the sympathetic nerves at the cervical nerve roots, it was carried there by a fairly low volume conduit with little compliance. It also meant that the volume spread proximally at least as much as it did distally and probably actually favored proximal spread. This could indicate that the brachial plexus sheath dead-ended distally. (I’m speculating that it attaches to the individual nerves as they approach the terminal branches of the brachial plexus.)
If this all were true then 3mls could create the block. If you’ve looked at a few brachial plexuses with ultrasound you’ve noticed that there is a great variation in the way they can appear. Some seem to have large diameter nerves and some seem to be made up of very small nerves, divisions and branching can occur higher or lower than expected. In short they’re not all alike. So I’m guessing there is nothing magical about the number 3 in terms of sheath capacity in milliliters. Besides I always liked 5 better.
So I approached a few anesthesiologists with the idea of performing an interscalene block with 5mls and you can probably guess what their first reaction was. Fortunately I am in collaboration with an anesthesiologist in France, Dr. Andre Ceccoli who brought the Phrenic Sparing Interscalene Block -- PSIB Project to me. It is standard practice for him to place a peri-neural catheter near the plexus following the single-shot interscalene block in order to provide continuous infusion post-operatively. I explained my thoughts to him regarding the low-volume intra-sheath approach and he immediately agreed to try it. (With the catheter in place he could re-establish the block and “rescue” the patient if my idea failed or was very short in duration.)
The next week he performed 5 of these blocks in one day and every one of them was successful. He use 5mls of ropivacaine with clonidine 3 units/kg for the blocks and he reported that they lasted between 8 and 12 hours.
Increasing the Block Duration
I took this as a “proof of concept” at least for the theory, but I was hoping for a longer duration. So I was faced with this situation; 5mls of local in the right place creates a solid block, Looks just like a block created by 30mls, but the duration seems to last about half as long. What is about the other 25mls that makes the block longer? That extra volume doesn’t change the onset time. That extra volume does not hang around the injection site in a liquid form creating a fluid reservoir from which the block is kept renewed as the original blocking drug is lost. From serial surveys after injection it appeared to me that most of the solution from large volume (30ml) injections dissipated away from the site quickly (5 mins), squeezed out of the injection site along naturally occurring planes. It is unlikely that portion of the local contributed to the block duration. But some of the block “overfill” undoubted was bound to the tissues immediately surrounding the target nerves, and it was this fraction of the injectate that must be responsible for the longevity of the block.
How could this tissue bound drug surrounding the target act to lengthen the duration? Again it seemed unlikely that the drug would unbind later and serendipitously head for newly opened sites in the sodium channels of the target nerve at the right time. The excess drug bound up in the surrounding tissue might then create a condition or environment that caused the block to last longer.
I thought this problem (of making a longer lasting block) over for the next few days, entertaining ideas about additives and raising the local anesthetic concentration. The block solution already contained epinephrine or clonidine, and raising the concentration of the local probably wasn't a good idea because local anesthetic can become toxic to nerve tissue when the concentration is too high. Then one morning while waiting in a dark car for my children’s daycare to open, I decided to imagine what this looked like from a molecular viewpoint. To make this thought process work, I threw out all variables which did not relate to diffusion. Therefore I did not consider lipid solubility, protein binding, pKa, metabolism of the drug or local tissue conditions. I viewed the process simply as solute movement according to concentration gradients.
I color-coded the local anesthetic molecules and flooded the area around the nerves with it. I imagined the local “staining” the outer layers of the nerves and slowly the anesthetic stain spread into the nerve. As the solution spreads around and along the nerve, the staining starts flowing into the periphery of the nerve from other places. The staining continues from the outside to inside until the tint reaches the center of the nerve’s mass and it continues until the nerve is evenly colored. The block is now set.
Considering only diffusion, the drug molecules begin leaving the nerve. They leave from the periphery of the "stained" area first. That is, it is improbable for drug molecules to move from the central mass of the stained area to the surrounding "unstained" area and be carried away from the site. Instead the drug must follow the gradient of concentration. Where there is no gradient, there is no net movement. Where there is the greatest gradient, there is the greatest net movement.
So the "block" recedes from the nerve area from edges, and the edge of the concentration gradient moves toward the center of the block from all edges, more or less simultaneously. When the edges meet in the center, the drug has diffused completely away and the block is completely dissipated. (even though clinically the block has ended long before this)
One implication of this is; the larger the "blocked" or "stained" area, the farther away from the center the edges are and the greater the surface area which must recede to center, therefore the longer it takes the block to diffuse out and recede completely. So this might be the key to understanding the relationship between local anesthetic volume injected and block duration.
This end understanding is a problem to finding a solution to this particular problem because increasing the volume of the local anesthetic leads to the side effect originally addressed in modifying the block, that is, excessive volume also leads to collateral block of the phrenic nerve.
So I tried to understand the other method of making a block last longer, vasoconstrictive additives.
Vasoconstrictive Additives to the Local Solution
In thinking about this method we can see that the edges of the block remain close to the center but the rate at which the edges recede to center is made slower by slowing the rate at which diffusion of the local occurs across the gradient. The epinephrine added to the solution constricts vessels around and within the injected site reducing blood-flow, reducing the effusion of the fluid serum portion of the blood into the tissue interstices which then circulates by diffusion amongst the tissue boundaries and is picked up by the lymph and carried away. By slowing the "turnover" of interstitial fluid at the edges of the block, you slow the creation of the concentration gradient at the "edge", which in turn slows the overall diffusion of local away from the center mass of the nerve and so prolongs the block.
The two methods both prolonged the block at the receding edge, higher volumes created a larger perimeter, vasoconstrictors maintained a smaller perimeter longer. Higher volumes created more phrenic blocks, therefore I needed some way to maintain the smaller perimeter longer, in excess of that created by the vasoconstrictor.
More vasoconstrictor seemed like a bad idea, invites more ischemic issues. How about cold? Cold pack applied locally at the site of the injection could slow the movement of interstitial fluid and thereby slow the turnover of the fluids at the edge of the block. Seemed like a good idea and as long as the cold pack didn't get too cold it didn't seem to have much downside. Well one way to find out, the next low volume interscalene I helped with I explained my idea to the anesthesiologist and he agreed to try it out. We finished the block, 5mls of bupivacaine 0.5% with epinephrine 1:200,000 injected very slowly into the sheath holding the brachial plexus nerves and placed a standard disposable ice-bag over the injection site. The ice-bag was off during the surgery and back on as soon as the procedure was over and the patient was in the recovery room.
I spoke with the patients before surgery about keeping the bag on after surgery, specifying that the bag was not to be placed over the surgery site but over the block injection site as long as possible after the surgery and through the night if possible. I said probably 30 minutes out of every hour would be enough to prolong the block. I tried this routine with the next few patients that received low volume interscalene blocks, the length of the block lengthened slightly from 8-12 hours to 15-20 hours. This result seemed underwhelmingly reassuring about the ice-bag protocol. But I also wasn't sure of the patient compliance with the protocol. Still it seemed enough to keep on trying.
The Low-ish Volume Interscalene Block
A compromise in the design of the technique seems to be the best way to arrive at the optimal end-point.
Putting together the thoughts from the 2 subjects discussed above then;
- The interscalene/supraclavicular brachial plexus block can be accomplished with a very low volume (5mls) of local anesthetic, well below the established average volumes of 30mls or greater
- Using the same local anesthetic mixtures, the block performed with larger volumes (30mls or greater) last significantly longer than the low volume blocks (5mls)
- Almost all interscalene/supraclavicular brachial plexus blocks performed with larger volumes (30mls or greater) produce a collateral block of the phrenic nerve on the side of the block
- Almost none of the interscalene/supraclavicular brachial plexus blocks performed with low volumes (5mls) produce a collateral block of the phrenic nerve on the side of the block
- There is no difference in the onset time or intensity of the 2 techniques (low vs. high volume). They are identical in all ways other than the duration and the occurrence of the concomitant phrenic block.
While the conclusions I draw next concerning the incidence of concomitant phrenic blocks by volume are premature because the PSIB (Phrenic Sparing Interscalene Block) project is ongoing, at this time it looks like about 92% of blocks done with greater than 20mls result in phrenic block, while only about 23% of those blocks done with less than 10mls produce the paralysis of the hemidiaphragm. These numbers 23% vs. 92% can represent the irreducible minimum for phrenic avoidance, and current standard rate of involvement of the phrenic nerve in the block, respectively.
Therefore, by increasing the volume of the low volume blocks we will incrementally increase the duration of the resulting blocks and, at some point in this escalating volume, the incidence of phrenic involvement will begin to rise above the 23% average for the low volume technique. The volume point at which phrenic inclusion rises above the 23% level, will mark the highest desirable volume if one hopes to keep the lowest possibility of phrenic involvement. If the block duration at that volume is reasonable, then the technique modification is finished. If the block duration at that volume is still below what one considers desirable, then one must abandon the lowest possible phrenic involvement level (23%) and accept a slighter higher rate in trade-off for the increase in duration OR change some other variable in the technique (needle placement, injection rate, local anesthetic, adjuncts and additives, etc.)
The Next Step
We are currently performing a modification of the block as described in the Neuraxiom article "The Perfect Interscalene Block". The goal of the modification of the block is intended to created an area of local anesthetic "soaked" non-nervous tissue surrounding the anesthetized nerve bundle. This pushes the perimeter of the block area further away from the center (which is hopefully the center of the target nerve bundle). Diffusion of the drug from the area must take place from the edges and since the edges are further from the center, it will take longer for the entire block to dissipate.
The modification we first try is the "3-5-3" block. The block needle is guided to the spot between the brachial plexus bundle and the anterior scalene, at about the level of the 2nd visible trunk. There we inject 3mls of the local solution. The block needle is then moved to a spot between the brachial plexus bundle and the middle scalene muscle (again taking care to stay out of the bundle), moving to the level of the 2nd trunk, 3mls are injected there. Then the needle is directed INTO the midst of the nerve bundle and 5mls are injected SLOWLY there.
The idea here is that if a small amount of local, in this case 6mls, is injected into the interscalene space (in 2 - 3ml aliquots) it won't be enough to overwhelm the compliance of the space and so won't be forced out to possibly reach the phrenic. It should be absorbed fairly quickly into the fascia and muscle surrounding the bundle (providing the tissue reservoir to extended the perimeter of the block).
The 5mls placed into the sheath provides the actual block.