Introduction

One of the less common areas of neuromodulation, peripheral nerve stimulation (PNS) is probably its fastest growing direction, at least in the field of pain treatment. The approach, which is still considered novel and experimental by many, is in fact neither – PNS was introduced before the much more accepted modality of spinal cord stimulation (SCS) and there are devices on the market today that are fully approved for PNS applications.

The history of PNS goes back to early 1960s. Few cases of electrical device implantation next to a nerve for control of neuropathic pain were performed even before the groundbreaking ‘gate-control’ theory of pain was proposed by Melzack and Wall . Shelden and colleagues implanted PNS devices around the trigeminal and other nerves for control of pain as early as 1962 . Since then, however, PNS has been through a period of relative growth followed by almost complete abandonment and then was reborn with the introduction of the percutaneous implantation approach in the late 1990s .

Today, PNS is used for variety of painful conditions in almost every part of the human body. Interestingly enough, the rapid growth in the number of reports and studies on PNS occurred despite the lack of dedicated and approved devices and, in almost every case, PNS is performed using hardware designed for SCS applications.

General principles of PNS

The basis of PNS is considered a reversible suppression of pain due to production of concordant paresthesias. Similar to SCS, this effect may be supported by the above-mentioned ‘gate-control’ theory of pain. PNS does not alter sensation in the zone supplied by the stimulated nerve – but it usually does not work when such sensation is already altered or absent, meaning that it is pretty much impossible to obtain a pain-relieving effect in the area of complete numbness, such as in extreme cases of diabetic neuropathy or true anesthesia dolorosa.

There are two distinct technical principles of PNS use and they dictate the choice of equipment and surgical approach. Both of them follow the same goal of delivering repetitive electrical stimulation to the nerve that is involved in pain production or transmission. For this to happen, one has to know which nerve or nerves are involved in that particular patient, and the best way to confirm PNS usefulness is to perform a trial of stimulation. Although nerve blocks and transcutaneous electrical nerve stimulation (TENS) were considered as possible predictors of PNS success, their effects did not correlate with PNS results.

For the first PNS approach, the nerve in question is surgically exposed and the electrode is placed directly over or under it (or around it in the case of wrap-around electrodes). This is an older technique – the first clinical series of the successful application of this approach was published in early 1970s . The use of flat (paddle-type) electrodes allowed the elimination of some concerns of nerve injury from the scar around the electrode, and the subsequent suggestion of putting a thin layer of fascia between the electrode contacts and the nerve itself was aimed at further reduction in nerve irritation from the presence of the large electrode nearby. Right around that time, a special electrode was developed and approved for use in PNS – a paddle electrode with a mesh attached to it (OnPoint, Medtronic, Minneapolis, MN). Also, even though not a single implantable pulse generator (IPG) is officially approved for PNS, the radiofrequency-coupled systems manufactured by Medtronic and St Jude Medical (Plano, TX) have PNS among their approved uses. The use of paddle-type electrodes for PNS continues to be an accepted means of PNS delivery and multiple recent reports document its success and reliability in a variety of clinical settings .

The second way of PNS application involves percutaneous insertion of stimulating electrodes. The initial technique of percutaneous PNS was used to prove the concept when Wall and Sweet stimulated their own infraorbital nerves to confirm development of analgesia during the stimulation . This approach, however, was not used clinically until the mid-1990s when Weiner and Reed described their technique of percutaneous insertion of PNS electrodes for treatment of occipital neuralgia . Since then, this approach has been successfully used in many anatomical locations and for different clinical conditions. Although associated with a high rate of complications and frequent need of surgical revisions , the percutaneous PNS approach is very appealing due to its technical ease and low invasiveness. A recent suggestion to use ultrasound guidance for location of the nerve to be stimulated now allows implanters to target many peripheral nerves along their subfascial or epifascial course. This includes occipital nerves and nerves in the trunk and extremities .

Somewhat similar to the percutaneous PNS approach is the so-called peripheral nerve field stimulation (PNFS). The difference in PNS and PNFS is the substrate of stimulation . Although both modalities definitely stimulate the nerve fibers that carry nociceptive information from the periphery to the central processing areas, PNS works with visible and identifiable nerves, usually the named ones, whereas PNFS works on unnamed, frequently multiple, smaller nerves that are hard to identify within subcutaneous tissues.

Overall, however, all these approaches are quite similar to each other. The goal of stimulation remains the same as it is necessary to produce non-painful sensation – paresthesias – in the area of pain, and maintain these paresthesias, usually on a continuous basis, for the pain to subside.

From a technical point of view, PNS implantation somewhat resembles the SCS procedure as it is usually performed in stages. During the first stage, the electrodes are implanted for trialing purposes. If the plan is to keep the trialing electrode in place for subsequent permanent use, the electrode has to be anchored and connected to a temporary extension cable that is then tunneled away from the insertion point. Care is taken to avoid damaging the nerve during electrode insertion. If the electrode is inserted with an open technique, the nerve is identified and dissected so that the electrode can be placed in its immediate vicinity. Anatomical location of the stimulation site should take into consideration the size of the electrode and the room for anchors, connectors and extensions. There is usually no need for any adjunctive imaging technique as the large nerves are easily identifiable in their expected anatomical locations.

With the use of the percutaneous technique, the trial electrodes are frequently discarded upon completion of the trial, so different electrodes can be implanted during the second stage of the PNS procedure. Here, there is no need to visualize the stimulated nerve directly. Instead, one may use standard anatomical landmarks, trusting limited variability in the nerve course and an ability to capture the nerve with multiple contacts of the stimulating electrode, particularly if the electrode is placed perpendicular to the course of the nerve. In addition to this, both fluoroscopy and ultrasound have been used intraoperatively to check the position of the nerves (ultrasound) and direction of the electrode path (fluoroscopy).

Following trial electrode insertion, the patient goes through a testing period that varies from 2–3 days to a week or even longer. During this time, the patient is encouraged to evaluate the effectiveness of PNS in terms of pain suppression, and to note any of the side effects that PNS may produce (pain, discomfort, spasms, etc.) so the decision can be made on whether the overall benefits of PNS justify the trauma and expense associated with permanent implantation.

At the time of permanent implantation, an IPG is implanted away from the area of stimulation, and the electrode(s) may be either directly connected to the IPG or connected to it via extension cables of appropriate length. As opposed to SCS, where electrodes are almost uniformly inserted into posterior epidural space, PNS electrodes may be implanted anywhere in the body (face, head, neck, trunk and extremities) and therefore one has to be creative in choosing the IPG site and the path for the electrodes and/or extension cables in order to minimize the chance of hardware migration (if the anchors are loose) or fracture (if the anchors are too tight) whenever excessive mobility is encountered. The thickness of tissues overlying the electrodes, anchors, connectors and generators should also be taken into consideration so that erosion and infections are avoided.

Since the decision on PNS effectiveness and side effects may only be made by the patient, it is important to go through all the logistics of appropriate patient selection. In addition to confirmation of severity and chronicity of the pain, it is important to check whether less invasive modalities have already been tried, and whether the patient’s psychological condition makes him/her an appropriate surgical candidate. The importance of psychological evaluation has been shown from the very beginning of PNS use as those with untreated depression, psychosis, major secondary gains and somatization disorder had overall unsatisfactory results in the long term . And similar to all other neuromodulation applications, it is also important to set realistic expectations as PNS does not cure the underlying pain syndrome and rarely eliminates pain completely but, with appropriate use in selected patients, it decreases pain levels and improves or normalizes their functionality. The patients also have to be prepared for a high chance of needing some kind of reoperation during the follow up as statistics show that, although most PNS complications are minor, they appear to be much more common compared to other neuromodulation procedures.

Specific PNS applications

The indications for PNS evolved over time. If the initial indications concentrated around isolated peripheral neuropathies, post-traumatic and post-surgical, and complex regional pain syndromes type 2, the more current ones include craniofacial pain syndromes, such as occipital neuralgia, transformed migraines, cervicogenic headaches, trigeminal neuropathic pain, post-herpetic neuralgia, and various localized pain syndromes in the trunk and extremities.

The immediate and long-term results of PNS in the extremities were encouraging in the earlier series. The success rate was between 50 and 60% in most series and the complications were rare. But the procedure did not become universally accepted for two main reasons. First was the lack of appropriate, specially designed electrodes for this application, and the ones that were used were usually custom made as small cuffs or buttons. Second was the need to expose the nerve for electrode implantation and only a few centers had enough interest and expertise to do it on regular basis. In addition to this, multiple reports indicating development of perineural fibrosis after a long-term PNS use cooled down the enthusiasm for PNS application. Most importantly, however, was the wide acceptance of SCS as a pain-relieving surgical procedure, particularly for those very indications where PNS was used in the past. Therefore, a change in practice came with the change in indications and PNS was tried for those conditions where SCS had not been successful.

Following the pioneering work of Weiner and Reed , multiple other centers started using PNS for stimulation of the greater and lesser occipital nerves. The electrodes are usually implanted perpendicular to the course of these nerves on one or both sides of the patient’s head at the level of craniocervical junction. The direction of insertion and the anchoring points vary from center to center. We routinely prefer anchoring the occipital PNS electrodes in the retromastoid region and tunneling them toward IPG located in the infraclavicular region . The results of our early experience with occipital PNS have been repeated in many other centers . Moreover, after the initial suggestion of Popeney and Aló , occipital PNS has been explored as a treatment for drug-resistant migraines . With very high prevalence of migraine headaches and significant proportion of drug- and treatment-resistant cases, occipital PNS for migraines may be one of the most common applications of neuromodulation.

The other, much less prevalent but perhaps even more debilitating than severe migraine condition, is the cluster headache. Occipital PNS has been used with a great degree of success in these patients and even those centers that use hypothalamic deep brain stimulation for treatment of this condition would frequently prefer occipital PNS as a first surgical option .

In addition to using percutaneous cylindrical electrodes for occipital PNS, paddle-type electrodes have been used for similar indications. The benefit of using paddle-type electrodes is their better stability and lower risk of migration. On the other hand, the insertion of this type of electrode is more invasive as it requires tissue dissection, whereas percutaneous electrodes are inserted through a spinal needle. Another difference that may be of particular importance in the case of occipital PNS is the direction of stimulation: paddle-type electrodes provide a unidirectional stimulation compared to circumferential in cylindrical wire-like percutaneous electrodes. This difference positively differentiates paddles in the case of SCS where the stimulation field is aimed at the posterior columns of the spinal cord. But, in the case of semi-blind positioning of PNS electrodes, this circumferential stimulation may be of benefit and the cylindrical electrodes may be tried in some cases of paddle electrode ineffectiveness.

Trigeminal branch stimulation is somewhat technically similar to occipital PNS. Electrodes are placed based on anatomical landmarks crossing the course of the supraorbital, infraorbital or auriculotemporal nerve. The epifascial location of the electrode makes it important to keep it at a sufficient depth – primarily to avoid electrode erosion. The anchoring point for these electrodes is usually placed in the retroauricular region and, from there, the electrode or extension cable is tunneled toward the infraclavicular IPG. Trigeminal PNS seems to work best for post-traumatic neuropathic pain but is not so good for post-herpetic neuralgia . It is also now being explored for treatment of migraines and cluster headaches . PNS for pain outside the craniofacial region may be divided into several subgroups. Back pain, by far the most prevalent chronic pain location, has been successfully treated with PNFS and the location of electrodes included all kinds of combination of vertical and horizontal paraspinal positions . In addition to this, there are now reports of a so-called ‘cross-talk approach’ with stimulating electrodes placed far from each other and stimulation delivered from anodes to cathodes located on the opposite sides of the lumbar region , and of a ‘hybrid stimulation’ where PNFS is used in conjunction with PNS .

Similar to lower back pain, PNS has been used for the treatment of neck pain , chest wall pain , inguinal pain and abdominal pain . In each of these scenarios, PNFS electrodes are placed into the middle of the painful area or around it, and PNS electrodes are inserted in the vicinity of the nerve responsible for the pain. Finding inguinal nerves with ultrasound may be helpful for correct electrode placement but, in other parts of trunk (back, neck, chest wall, etc.), PNS electrode positioning is performed based on the patient’s description of the painful area: the area of pain is usually drawn on the patient’s skin and the electrode position is chosen in such a way that the entire painful area is covered with a stimulation field from its center or from the edges.

Finally, pain in the extremities may be successfully treated with PNS. In the original series of the 1970s and 1980s, PNS of the upper extremities resulted in better pain relief compared to that of the lower extremities – partly due to the mixed nature and larger size of the commonly stimulated sciatic nerve. Recently, however, percutaneously inserted PNS electrodes have been used for the control of pain in both upper and lower extremities .