Indications

The usual indications for PNS in neuropathic pain are very similar to those used for SCS procedures. The pain is chronic and severe, negatively affects the patient’s functionality, and is refractory to the usual medical treatment, including medications, physical therapy, and less invasive interventions, such as local anesthetic and sympathetic blocks, application of transcutaneous electrical nerve stimulation (TENS), and injections of botulinum toxin preparations, if these interventions are even considered by the treating physicians. For the purposes of this chapter, the neuropathic pain is broadly defined as “pain initiated or caused by a primary lesion or dysfunction in the nervous system,”⁶⁶ and this wide definition includes, among other syndromes, such debatable nosologic conditions as fibromyalgia, migraines, and coccydynia.

For classic indications in neuropathic pain treatment, the nerve involvement is usually suggested by concordant sensory loss, partial or complete, and may be confirmed by a local anesthetic block. This approach has been recommended for treatment of occipital neuralgia, trigeminal neuropathic pain in infraorbital and supraorbital distribution, and inguinal pain in postherniorrhaphy pain syndromes, although, as far as we know, there are no literature data about any predictive value of blocks on PNS success. Interestingly enough, success or failure of TENS does not appear to have any predictive value for PNS whatsoever, and the patient’s failure to improve with TENS should not prevent the physician from trying PNS for pain control (personal experience).

A somewhat different approach is chosen for treatment of emerging indications, such as refractory migraines and fibromyalgia, for which there is no obvious injury to the underlying peripheral nerve and the usefulness of diagnostic nerve blocks is rather doubtful. The same holds true for conditions that probably represent a combination of neuropathic and nociceptive pain, such as axial back and neck pain.

Two other criteria are considered before implantation of a PNS system: the patient’s psychological evaluation and results of a short-term trial with externalized electrodes. Both these steps are taken as the standard approach to SCS; the psychological evaluation details and their predictive values in SCS were recently reviewed.⁶⁷ Outpatient stimulation trial is also a routine step in SCS surgical procedure, and the trial length usually varies between 2 and 14 days. Usually, a 50% improvement in pain intensity serves as a cutoff limit for considering the trial successful. In some places, the patient satisfaction with pain relief, even if the pain intensity decrease is less than 50%, justifies proceeding with permanent implantation.

In terms of specific indications, there are at least 11 distinct pain conditions for which PNS has been recently reported: (1) postherpetic neuralgia^40,45,64; (2) posttraumatic or postsurgical neuropathic pain that is related to underlying dysfunction of a certain nerve, including the infraorbital, supraorbital, or occipital nerves^45,48,52; (3) classic migraine, “transformed migraine” presenting with occipital pain and discomfort, and hemicrania continua^{39,43,46,47,53,54,56,61}; (4) occipital neuralgia or cervicogenic occipital pain^{34,38,47,49,51}; (5) complex regional pain syndrome^28,30,68; (6) cluster headaches^54,58,59,69; (7) chronic daily headaches^39,53,60; (8) inguinal pain after herniorrhaphy⁶²; (9) coccydynia⁶⁵; and (10) fibromyalgia.^57,70

Because of the relatively simple and nontraumatic nature of PNS, the list of contraindications is short and is based predominantly on common-sense considerations. For example, PNS is contraindicated in patients with bleeding disorders and active anticoagulation that cannot be stopped for a few days around the surgical procedure; in patients with active infection, particularly if there is bacteremia or direct involvement of the surgical region; in patients with major cognitive impairment, untreated depression, or malingering; and in patients with unsuccessful PNS trial. In addition, because all devices on the market have not been cleared for routine magnetic resonance imaging (MRI), patients who require follow-up MRI (e.g., patients with tumors) should not be implanted with PNS.

Surgical Technique

The hardware for PNS is usually implanted with a straightforward set of guidelines. The electrode (lead) is implanted in the vicinity of the stimulated nerve. The approach may include either direct exposure of the nerve,³² so that the electrode may be placed next to it (or even wrapped around it—similar to vagal nerve stimulation system),^7,8,71 or percutaneous electrode implantation through a needle inserted perpendicular to the nerve course.^34,48,52,59 Between these two approaches is the technique of surgical lead insertion toward the nerve without actual nerve exposure.^{44,47,49,58,61}

The electrode is then anchored to prevent future migration or displacement. In most cases, the anchors are placed over the electrode stem (in case of wire-type electrodes) or its tail (in case of surgical leads–paddles) and sutured to the nearby firm and relatively immobile tissue (fascia, periosteum). Depending on the area of stimulation, this may be done either in the incision (e.g., in the occipital area) or at a distance from the insertion site to avoid cosmetic disfigurement (as in infraorbital and supraorbital electrodes, where the anchors are placed in a separate incision behind the ipsilateral ear).

The electrodes are inserted for the trial in sterile conditions either under local anesthesia or under sedation augmented by infiltration of the insertion site with local anesthetic. Because the procedure is short and the surgical site is superficial, general anesthesia is almost never needed.

The direction of electrode insertion may be chosen based on implanter’s preference; for example, we routinely insert electrodes from lateral to medial not only in the supraorbital and infraorbital regions (where it is probably the only way to place them) but also in the occipital area,^51,52,55 whereas others prefer to insert electrodes from medial to lateral.^41,43 Standard 4-, 8- or even 16-contact electrodes⁶¹ are used (Figs. 166-1 to 166-3); the electrodes are passed in the epifascial plane under the skin but above the muscles. Our general approach is to have the electrode cross the path of the nerve chosen as a stimulation target. As long as this nerve happens to be either under one of the electrode’s contacts or between the two contacts, the stimulation can be steered toward it to achieve adequate coverage. For the trial insertion, we do not implant any deep anchors or extensions. The electrodes are sutured to the skin with plastic anchors and fine nylon, and a strain-relief loop is created around the insertion site to avoid inadvertent electrode pullout.

FIGURE 166-1 Anteroposterior radiographs of unilateral (A) and bilateral (B) 8-contact occipital nerve stimulation electrodes.

FIGURE 166-2 Anteroposterior radiograph of bilateral 4-contact supraorbital nerve stimulation electrodes.

FIGURE 166-3 Anteroposterior radiograph of unilateral 4-contact infraorbital nerve stimulation electrode.

The electrodes are almost always inserted under fluoroscopic guidance. In treatment of craniofacial pain, standard landmarks are used for the insertion: the infraorbital foramen and the floor of the orbit for the infraorbital nerve; the supraorbital groove or foramen and the supraorbital ridge for the supraorbital nerve; and the C1 arch and radiographic midline for the occipital nerves. When we started doing this procedure we tested each patient for stimulation-induced paresthesias in the operating room so that the position of the electrode could be adjusted if needed. Lately, however, we rely exclusively on anatomic electrode positioning owing to its high reliability in achieving appropriate coverage. This has resulted in significant improvement in patient comfort associated with deeper sedation that may be used now because the patient’s cooperation in the procedure is not needed. The electrode is covered with sterile dressing and attached to the external stimulation system; the initial programming is performed to produce adequate paresthesias in the painful area; the patient is instructed on adjustment of the stimulator depending on activity and pain level; and antibiotics are prescribed for the duration of the trial to avoid development of superficial infection (although this particular step is not supported by any clinical evidence).

When the trial is completed, the temporary system is replaced with the permanent one. We prefer to remove the temporary electrode in the operating room after confirming its position with fluoroscopy, and then we insert a new permanent electrode that is connected either directly to the generator or to an extension cable that connects to the generator. The electrodes that we use are cylindrical wire type (e.g., Quad, Octad, Quad Plus, or Quad Compact, Medtronic; Qattrode, Octrode, or Axxess, ANS; and Linear, Advanced Bionics). Other groups reported using plate-type (paddle-type) electrodes (e.g., Resume, Resume II, or Resume TL, Medtronic; or Artesan, Advanced Bionics) for stimulation of the occipital nerves. Implantation of such electrodes may be preceded by a trial with wire-type electrodes,⁴⁹ or the trial may be done with plate-type electrodes connected to a temporary extension.⁴⁷ In all cases, the electrodes are placed over the course of the peripheral nerve that supplies the painful area and may be involved in the generation of pain.

The electrodes or extension cables are tunneled toward the generator pocket. The tunneling step is painful and, in our practice, necessitates the use of general anesthesia. Location of this pocket is chosen based on the patient’s and surgeon’s preference. Placement of the generator into the gluteal area,^49,50 abdominal wall,^38,44 and infraclavicular areas^{34,45,47,48,51,55,58} has been described. In our opinion, the infraclavicular area, which is routinely used for placement of deep brain stimulation generators, is the preferred location for both trigeminal and occipital nerve stimulation systems (as documented by a recent analysis of clinical experience of patients with infraclavicular generators that found extremely high levels of patient satisfaction with this particular location).⁷² Independently of location, the pocket should satisfy certain requirements: it has to be deep enough to avoid hardware erosion; it should not be too deep to interfere with reprogramming or, in the case of rechargeable devices, their regular charging; and it should be located in a relatively immobile region because the hardware may fail if subjected to repetitive mechanical stress.

A completely different setup is used with a new revolutionary device, BION (Advanced Bionics), that integrates the generator and electrode into a single (and very unique) apparatus.⁷³ Initially developed for functional electrical stimulation, the BION has been tried for PNS in the treatment of overactive bladder⁷⁴ and headaches^54,56 in multiple clinical centers. Several published reports have been made of its clinical use,^54,56,74 although the device has not yet received formal FDA approval. In essence, BION is a miniature leadless implantable neurostimulator that is designed to be implanted through a small incision as a part of a minimally invasive procedure. Because the stimulating electrodes are mounted directly on the neurostimulator, the possible complications associated with the use of leads (e.g., migration, disconnection) are avoided. The currently tested device is the shape of a cylinder measuring 3.3 mm in diameter and 27 mm in length. It has a cathode on the tip of the stimulator and an anode in its distal surface; it contains a programming and telemetry module and a rechargeable battery. BION allows delivery of a wide range of stimulation parameters with pulse width up to 1000 µsec, stimulation rate up to 1000 Hz, and electrical current amplitude up to 12 mA. Similar to other devices on the market, BION is programmed by trained clinicians; a remote control is given to the patient, allowing adjustment of the stimulation parameters within a preset range. The charging process takes less than 2 hours, and the interval between charging depends on the device use, ranging from daily to every other week, with total battery life expectation of up to 20 years.⁷⁵

Results and Future Directions

So far, the use of PNS for neuropathic pain has been associated with consistently positive results.53–60,62–64 With the only exception of postherpetic trigeminal neuralgia, in which PNS has shown only minor improvement in less than half of patients,⁴⁵ the other indications have either been shown to respond to PNS or are currently investigated in prospective fashion.

Interestingly enough, the older and more established indications, such as posttraumatic and postsurgical neuropathy, occipital neuralgia, and complex regional pain syndromes, are likely to be overshadowed by the relatively new indications for PNS, such as migraines and daily headaches, cluster headaches, and fibromyalgia. Even if the effectiveness of PNS for newer indications is less prominent than it is for older and more established ones, the significantly higher prevalence of these newly investigated conditions is likely to raise further interest in PNS in the spectrum of pain-relieving interventions. Obviously, the low invasiveness of PNS and its reversibility, testability, and adjustability make it a preferred choice, particularly in comparison to the destructive surgical options, but the need for implanted hardware and the current high cost of the devices will undoubtedly require more studies and larger numbers of patients before this option is accepted by the medical and third-party payer communities.

The complication rate of PNS is generally low, but both minor and major complications, such as local infections, hardware erosions, component disconnections, electrode fractures and displacements,^32,48,51,55 and even sepsis,^18,22 have been reported in the literature. The complications of perineural fibrosis, described in the past with the use of plate-type or wrap-around electrodes,^18,24,33 are essentially unheard of with the use of modern wire-type electrodes. Most complications require system revision or replacement; lasting side effects are overall unlikely.

The mechanism of PNS is still largely unknown. Several animal studies proposed certain explanations that are related to direct excitation of the central pain-processing system and increase in the excitability of the system,^76,77 although limited human research indicated activation of the dorsal rostral pons, anterior cingulate cortex, and cuneus in response to PNS in the suboccipital area.⁴⁶ Better understanding of the PNS mechanism may result in refinement of surgical indications, individual tailoring of appropriate treatment approach, and perhaps optimization of the hardware choice.

Development of special devices for PNS is yet another potential direction of progress in this rapidly growing area. New electrodes with different spacing options and lower profile may be particularly useful for PNS. Significant research in the area of nerve-electrode interface is taking place,⁷⁸ but the devices for PNS have not been developed or approved for widespread clinical use. It is possible that newly developed electrodes will be used not only for PNS in the strict sense of its definition but also in the neighboring area of stimulation of spinal nerves⁷⁹ and the gasserian ganglion.⁸⁰ Undoubtedly, BION^73,75 and similar devices, dedicated to PNS, may widen the indications and further decrease the rate of complications and side effects. Getting approval for these devices to use with PNS procedures may facilitate the acceptance of the approach and better reimbursement of procedures that are done now only as part of research protocols or on an off-label basis.

The future research and growing clinical experience will help in identifying the best candidates for PNS, choosing the best procedure and best hardware for the individual patient, and defining adequate expectations for patients and pain specialists. The opportunities for growth are endless—it appears that the current state of PNS represents only the tip of the iceberg and that its true role in the pain management continuum is still to be discovered.