Salvage treatment of trigeminal neuralgia by occipital nerve stimulation

Abstract

Background

Although most patients suffering from trigeminal neuralgia (TN) respond to medical or surgical treatment, nonresponders remain in very severe painful condition.

Case result

We describe for the first time a case of severe refractory classical TN treated successfully (follow-up one year) by chronic bilateral occipital nerve stimulation (ONS), because other classic medical and surgical options failed or could not be performed.

Conclusions

This single case suggests that ONS might be offered to TN patients refractory both to standard drugs and interventions, with a favorable risk/benefit ratio, although its long-term efficacy remains unknown.

Introduction

Trigeminal neuralgia (TN), has a prevalence of 0.1–0.2 per thousand and an incidence of 12.6/100,000/year (1). Although most patients initially respond to carbamazepine and/or other antiepileptic drugs, a high proportion of them become drug resistant with time and finally need surgical treatment (2). Several surgical procedures may be performed, including microvascular decompression (MVD), gasserian ganglion percutaneous techniques and/or radiosurgery, but these do not provide long-term pain relief for all patients. Continuous electrical occipital nerve stimulation (ONS) has been proposed for a decade to treat occipital neuralgia and primary headaches (3), but has never been proposed to treat TN. We describe for the first time a case of severe refractory TN treated successfully with ONS, because no other classic surgical options could be performed.

Case report

A 76-year-old woman suffered from classical TN for 14 years, initially treated medically, then needing several successive surgical treatment, including one gammaknife radiosurgery (GKS) in 2000 and six gasserian rhizotomies (two radiofrequency thermocoagulations (RFTs) in 2000 and 2005, and four balloon compressions in 2006, 2007, 2008 and 2011). Each of these procedures had achieved pain control for six to 12 months, but TN had recurred then. She was referred to our center in December 2011 for severe TN worsening lasting for two months. She complained of very frequent brief paroxysms of electric shock-like pain in the maxillar (V2) territory of the right trigeminal nerve, triggered by touching the right upper lip, speaking and swallowing, leading to mutism and severe denutrition (loss of 6 kg during the last two months). Despite the numerous previous surgical procedures, she still had classical TN fulfilling the International Classification of Headache Disorders criteria (4) with no characteristic of neuropathic pain or anesthesia dolorosa, except mild hypoesthesia in the mandibular (V3) territory of the trigeminal nerve, sequel of a previous RFT. Magnetic resonance imaging (MRI) clearly showed a vascular compression on the trigeminal nerve (Figure 1(a)). The pain attacks’ intensity and frequency, and the consequent impairment of her general condition, were consistent with a refractory neuralgia status despite medical treatment at the highest tolerated dose, i.e. carbamazepine 600 mg/d associated with baclofene 30 mg/d and phenytoin 300 mg/d. She was also resistant to fosphenytoin and lidocaine intravenous administrations.

Figure 1.

(a) Preoperative high-resolution magnetic resonance imaging (MRI) (three-dimensional fast imaging employing steady state acquisition (3D FIESTA) sequence) showing a neurovascular conflict between the root entry zone of the trigeminal nerve (white arrow) and the antero-inferior cerebellar artery (arrow heads). (b) Postoperative skull radiograph (antero-posterior view) showing the location of the two horizontal occipital octopolar electrodes crossing the great occipital nerves paths.

An additional percutaneous rhizotomy was proposed but two experienced neurosurgeons failed to penetrate the foramen ovale. We hypothesized that this was because of the obliteration of the foramen by excessive scar tissue related to the previous procedures. An MVD was proposed but because of age, severe denutrition and poor general condition, the risks of a posterior fossa approach were estimated to be too high.

Because the patient was still suffering and there was no other solution, we proposed ONS. We dutifully informed the patient about the technique, its potential risks and benefits, and the fact that this technique was not validated and not even reported in TN.

Two subcutaneous electrodes (Octrode, St Jude-ANS) were implanted bilaterally, under local anesthesia, perpendicular to the great occipital nerves in the occipital region (Figure 1(b)), according to the technique previously described (5,6), and then connected to an external generator. Continuous electrical stimulation (frequency 60 Hz, pulse width 120 ms, intensity 4.5 mA) was delivered during the next two weeks; medical treatment was not modified.

After the surgery, the patient reported a progressive but dramatic improvement (Figure 2). The pain intensity of the “tics douloureux” progressively decreased during the first two postoperative days and their frequency, initially unchanged, decreased after one week of stimulation. The gingival trigger zone progressively lost its triggering properties. Consequently, she became able to speak and to eat and began to regain weight. After 10 days of continuous stimulation, she was pain free. The externalized electrodes were removed to avoid infection and the stimulation was consequently stopped. TN attacks recurred two weeks after the stimulation was stopped and became progressively as severe as before surgery. Consequently, subcutaneous ONS hardware was re-implanted and chronic stimulation (bilateral bipolar stimulation, frequency 30 Hz, pulse width 350 ms, 1.8 mA, unchanged during the one-year follow-up) induced a progressive improvement of TN attacks following the same pattern as previously. One month after surgery she reported a 50% improvement of TN attack intensity and frequency, and had regained weight (1 kg). A maximum improvement of 90% was reached 1.5 months after surgery and then maintained (follow-up one year). Phenytoin was stopped but two attempts to decrease the dose of carbamazepine below 600 mg/d (Figure 2) and one attempt to stop the stimulation provoked the recurrence of severe TN.

Figure 2.

Evolution of the global pain score (out of 10) over time, during the trial stimulation and the chronic stimulation (black lines) and two attempts to decrease the medical treatment (white lines).

Discussion

To our knowledge, this case is the first report of severe intractable classical TN treated successfully by ONS, although ONS can alleviate trigeminal neuropathic pain (7). The fact that ONS induces perceptible paresthesias in the occipital region prevents blind evaluation of the ONS effect. Consequently, a placebo effect cannot be discarded but is unlikely because placebo effect is rare in cases of TN and, if present, is usually mild (8) while the therapeutic effect of ONS was dramatic and maintained for months. Spontaneous improvement of this acute exacerbation of TN, with no relation to ONS, was also very unlikely because TN systematically recurred after stopping ONS and ceasing medical treatment and because the therapeutic effect followed the same chronology after ONS transient trial and after definitive ONS implantation.

Although ONS has not been validated for TN, we proposed this treatment to our patient considering the severity of her painful condition leading to a life-threatening impairment of general health, partially due to severe denutrition as a consequence of the triggering of pain by mastication. On the another hand, ONS has been proposed for a decade to treat occipital neuralgia and facial pain. Although needing surgical hardware implantation, ONS is a minimally invasive technique and its main risks are limited to subcutaneous infection, electrode migration and hardware dysfunction, nevertheless frequent in some series. Consequently we considered that the risk/benefit ratio of ONS was favorable in this case, but we decided to perform first an ONS trial, the electrodes being implanted under local anesthesia, which was successful.

Although therapeutic management of TN is well established (9), a proportion of TN patients continue to suffer despite medical and/or surgical treatment. Pain-free survival rates at three years are 54%–64%, 52% and 75% after RFT, GKS and MVD, respectively (9). Consequently, a proportion of patients with medically and surgically refractory TN need additional surgical procedures that may be combined to obtain pain relief (10). This single case suggests that ONS might be offered to TN patients refractory both to standard drugs and interventions, with a favorable risk/benefit ratio, although its long-term efficacy remains unknown.

ONS appeared efficient in various primary headache and facial pain conditions with different patho-physiological mechanisms, including cluster headache, chronic migraine, occipital neuralgia and facial neuropathic pain. Although the mechanisms of action of ONS are still unclear, ONS could act through stimulation of afferent fibers of the great occipital nerve projecting to the convergent neurons of the trigeminocervical complex (TCC), which is involved in the patho-physiology of several primary headaches (11). Neuroimaging studies have also suggested the existence of upper central mechanisms, nondisease specific, including activation of the thalamus and pain matrix cortical regions (12,13). As the patho-physiology of TN is completely different from that of primary headaches, our case adds to the arguments in favor of a common nonspecific mechanism of action of ONS on these different painful conditions. The fact that occipital nerve block (ONB) can transiently improve classical TN (14) suggests that ONB or ONS might share similar mechanisms of action by inducing changes in cervical afferent sensory inputs to the TCC and might modulate the hyperexcitability of trigeminal neurons (15). Lastly, we cannot exclude that ONS increased the sensitivity to antiepileptic drugs, as association of both treatments seemed necessary to obtain improvement of TN pain. Such a synergic effect has been shown in the past with spinal cord stimulation (16,17), but has not been suggested yet with ONS. This synergic effect could result from the activation of descending inhibitory pathways and the consequent reduction of the TCC activity, as previously suggested (18). Moreover, our case is interesting because it suggests that ONS could be used to treat painful conditions of the V2 territory, not only facial pain located in the V1 territory. Indeed, functional convergence of trigeminal and occipital afferents in the TCC has been mainly studied through stimulation of the V1 fibers, using stimulation of the dura mater in animal models (19,20) or nociceptive blink reflexes (21).

Clinical implications

Occipital nerve stimulation might be an option to treat refractory trigeminal neuralgia.