Changes in peripheral and central sensitization in patients undergoing occipital nerve stimulation

Abstract

Introduction:

Central sensitization and impaired conditioned pain modulation (CPM) response have been reported to contribute to migraine progression. Migraine patients can present with allodynia possibly attributed to increased sensitivity of peripheral ends of nociceptors with both peripheral and central sensitization. Occipital nerve stimulation (ONS) works by stimulating the distal branches of C1, C2 and C3 possibly altering the nociceptive traffic to the trigemino-cervical complex, brainstem and supranuclear connections.

Aims:

This observational study explores peripheral and central sensitization in patients undergoing percutaneous ONS.

Methods:

Following local regulatory approval, 13 patients undergoing ONS with dual Octrode 90 cm leads and rechargeable implantable pulse generator (IPG) (St Jude) were recruited to have quantitative sensory testing (QST) pre- and post-procedure 2 weeks, 1, 3, 6 and 12 months.

Results:

Patients with intractable migraine demonstrated impaired CPM (mean baseline pressure pain thresholds (PPTs): 61.98 kPa vs 48.01 kPa cuff inflated) prior to ONS, reverting to an efficient CPM response within 2 weeks following ONS implant (68.9 kPa vs 104.5 kPa cuff inflated) and continuing positively over the next 12 months. In contrast, no statistical difference was observed in PPTs.

Conclusion:

This is the first reported observation highlighting the effects on central sensitization following ONS. A consistent and sustained improvement in CPM was observed in contrast to PPT’s where there was no difference. Normalisation of the CPM response following ONS indicates that the treatment may reduce central sensitization in the migraine population.

Introduction

Headache is one of the most common neurological problems accounting for 4% of primary care consultations and up to 30% of neurology appointments.¹ Headache disorders are a significant cause of pain and disability to individuals and are also a major societal burden.² Even though considerable advances in the understanding of the pathogenesis of migraine, new pharmacologic treatments and the emergence of new innovative interventions for difficult cases, in many patients migraine remain intractable to medical therapy.

The pathophysiology of migraines remains unclear; however, a role for the trigeminocervical complex (TCC) has been reported. TCC is a common convergent pathway by which both trigeminal (dural) and cervical (mainly via the greater occipital nerve) afferent inputs project into higher centres in the thalamus and cortex. This afferent pathway is further influenced by the conditioned pain modulation (CPM) pathway originating from the brainstem structures including periaqueductal grey matter, nucleus raphe magnus and the rostroventral medulla.³ There is some evidence to suggest that increased peripheral sensitization of the afferent inputs of the TCC, generalised central sensitization of the trigemino-spinal or the second order trigeminal neurons and impaired descending pain inhibitory pathway may all contribute to the development of the chronic intractable or treatment resistant primary headache disorder.⁴,^5,20

Neurostimulation techniques for treating intractable headache range from invasive deep brain stimulation to less invasive peripheral implantation and stimulation. In 1999, Weiner and Reed described the beneficial effects of subcutaneous occipital nerve stimulation (ONS) in patients with medically intractable, highly disabling chronic headache disorders.⁶ Open-label studies have suggested that this treatment modality is effective and safe.⁷ A 52-week open-label extension study has reported 30% responder rate of 59.5% and 50% responder of 47.8% with headache days being significantly reduced by 6.7 days/month at the end of 52 weeks.⁸ Although the exact mechanism of action is poorly understood, it is thought to act by stimulating the distal branches of C1, C2 and C3 in turn influencing the TCC favourably by inhibiting the nociceptive process resulting in an improvement of the symptoms. This hypothesis is supported by a fluorodeoxyglucose–positron emission tomography (PET) study on 10 patients with medically refractory chronic cluster headache which found that ONS exerts its effect via slow neuromodulatory processes in the central pain matrix. Hypermetabolism was found in several areas of the pain matrix prior to stimulation and returned to normal after 3–6 months stimulation.⁹

The primary aim of this study was to assess the reversibility of a range of psychophysical markers of pain processing in patients with intractable migraine before and after ONS insertion.

Methods

A total of 13 patients (M:F, 10:3) undergoing routine occipital nerve stimulator insertion for intractable primary headache were enrolled to undergo QST measurements. All patients were screened by a multidisciplinary team including a pain physician, neurologist with a special interest in headache, psychologist and a neurosurgeon prior to undergoing ONS. Baseline QST was measured prior to ONS intervention and then further QST measurements were taken at 2 weeks, 1 month, 3 months, 6 months and 1 year. All QST measurements were taken by a single, trained operator to ensure consistency and reproducibility of the results. The following QST measurements were recorded:

• PPTs. PPT through algometry provides a quantitative assessment of sensory perception to increasing mechanical stimuli. A hand-held pressure algometer (Algometer type II, Somedic Production AB, Sweden, diameter contact tip 10 mm; cover 2 mm thick rubber; and standardised and constant speed of pressure increase of 0.3 kg/s) was used to measure the PPTs at five standardised points, temporal bone, maxilla, cranium and two points along the mandibular. A set of measurements were also taken from the medial aspect of the arm as a control. The probe was placed perpendicular to the skin in the predefined points and pressure was applied until the subject perceived the pressure as pain and the procedure was immediately terminated. Measurements were taken three times at each site and an average PPT was used for statistical analysis. This average PPT formed the baseline and comparison for the following test.

• CPM. Ischaemic compression of the left arm was used as heterotopic noxious conditioning stimulation for evoking CPM. The pressure cuff was inflated above systolic pressure (200 mm Hg) for 10 minutes, or until a visual analogue scale rating of 6/10 cm was achieved (where 0 cm represented ‘no pain’ and 10 cm represented ‘maximal pain’), repeated PPTs were measured on the cheekbone while the cuff was in situ. The arm cuff was released once PPT and cuff pain assessments were finished (maximum of 10 minutes).

• ONS Procedure. Percutaneous lead placement was performed under general anaesthesia with St Jude Medical’s Octrode leads (8 evenly spaced 4 mm contacts) placed bilaterally parallel to the greater occipital nerve. All implants were done by VM (single operator) and the implantable pulse generator (IPG) (Genesis/Prodigy St Jude Medical, Plano, TX, USA) were placed in left upper buttock region.

Statistical analysis

The data obtained from the QST will be analysed using Student t-test or Mann–Whitney test, and will be presented as means with standard deviations.

Results

Patient demographics

A total of 13 patients with medically intractable primary headache disorder were recruited for this study. The underlying headache diagnosis were chronic migraine (n = 8), hemicrania continua (n = 2) and cluster headache (n = 3). The average age of the patients was 44.9 years and the gender was male 10: female 3 (Table 1).

Table 1.

Patient baseline characteristics.

Age (years, mean ± SD)	44.9 ± 10.1
Gender ratio (M/F)	10:3
Duration of migraine (years migraine experienced prior to study entrance, mean ± SD)	12.2 ± 3.3 years
Phenotyping	Migraine: 8 Cluster: 3 Hemicrania continua: 2

SD: standard deviation.

Quantitative sensory testing results

All patients prior to ONS had abnormally lowered PPTs. There was no significant change in the PPT measurements in the patients following the insertion of the ONS and this remained at subsequent follow-up visits (Figure 1).

Figure 1.

Pressure pain thresholds in subjects’ pre- and post-ONS for intractable migraine.

All patients included in the study had an abnormal (reduced) baseline CPM response, that is, PPT pre-cuff inflation higher than PPT post-cuff inflation (61.9 kPa vs 48.0 kPa).

Following the insertion of the ONS, a ‘normal’ CPM response, that is, PPT pre-cuff inflation lower than PPT post-cuff inflation (57.6 kPa vs 91.2 kPa cuff inflated) was seen in all but three patients, as early as 2 weeks following the procedure. At 1-month post-procedure, this was maintained in all but one patient, and at 3 months this improvement was demonstrated in all subjects and subsequent follow-up (Figure 2).

Figure 2.

The effect of a noxious stimulus (cuff) on PPT’s (maxilla) pre- and post-visits following an occipital nerve stimulator for migraine.

Discussion

This is the first reported study looking at the effect of peripheral and central sensitization following ONS.

QST is a semi-qualitative psychophysical measure of large and small afferent fibre function used to assess the functional changes and reorganisation of the nervous system, especially the nociceptive system. It is generally used to characterise the peripheral and central sensitization models in neuropathic pain. These tests range from simple bedside tests that can be conducted in outpatient clinics to expensive laboratory research-based tools. The tests used in QST can be broadly classified as either static or dynamic measurements. Static measurements depict a single point on the pain experience continuum, for example, the threshold determination of pressure when it becomes ‘painful’ to the participant (PPT response). The static measures have the advantage of using an easily defined endpoint, which is stable and reproducible in practice but represents only one part of the pain process. In contrast, dynamic QST measurements may be used to capture the endogenous pain modulatory process. CPM investigates activity within the spinal dorsal horn arising from peripheral nociceptive inputs and can be modulated by powerful descending inhibitory and facilitatory mechanisms. An example is provided by the phenomenon of CPM, also known as diffuse noxious inhibitory control (DNIC) or heterotrophic noxious conditioning stimulation.¹⁰ This refers to an altered response to a painful stimulus following the administration of a second conditioning stimulus. CPM provides one of the main supraspinal pain inhibitory pathways and is impaired chronic migraine.

Currently, there is no evidence to support the effect of general anaesthesia on sensory characteristics measured by QST; however, there is some suggestion that long-term opioid use may influence QST response. A recent study demonstrated that chronic opioid intake reduced only temperature sensitivity but not pain sensitivity measured by QST.¹¹,¹²

The use of QST in migraine remains scant with only a handful of studies describing the use of QST in identifying possible pain pathways¹³ or as an objective biological measurement of headache pain¹⁴ but not used as in this study to characterise an effect from the intervention of ONS.

The patients undergoing ONS had low PPT’s in the affected area. After ONS, there was no statistical change in the thresholds. However, all patients demonstrated an impaired CPM and this improved after ONS as early as 2 weeks. We know that impaired CPM response is associated with the development of a chronic pain state irrespective of the underlying aetiology and has a high pain morbidity and hence a poor outcome.¹⁵ Although this study suggests that with the insertion of ONS, CPM improves, however, the mechanism by which ONS works remains to be elucidated. This result may possibly support the central action of ONS by modulating the TCC and affecting the central sensitization. The reversal of the CPM was seen in all but three patients at 2 weeks, in all but one patient at 1 month and all patients at 3 months which was consistently maintained to 12-month follow-up demonstrating a consistent and maintained reversal in CPM and central sensitisation. In contrast to the effect on CPM, ONS seems to have no beneficial effect on the PPTs. Traditional ‘static’ QST like PPTs seem to depict only one part of the pain pathway and, therefore, unable to characterise different pain processes, this may account for why there was no improvement within PPTs. In contrast, ‘dynamic’ QST captures the endogenous modulation response and may be more relevant in the ‘central’ action of ONS. Also, current evidence suggests that dynamic tests may be more suitable in predicting outcomes for pain interventions.¹⁶ Our current work suggests that ONS may alter central pain processing mechanisms. Although prediction of phenotyping on the basis of sensitization was beyond the scope of this preliminary work, we feel that targeting ONS in patients with central sensitisation (impaired CPM) maybe beneficial because these patients may have the greater capacity to change.

Interest in ONS for intractable headache is growing fast due to a large proportion of patients not responding to preventive drugs; however, to date most studies have been ‘open’ without a sham arm and therefore a placebo effect cannot be excluded. Previous findings, however, have shown that ONS usually requires weeks or months to become effective and headache attacks usually return shortly after the stimulator has been switched off, becomes displaced or the battery becomes depleted.¹⁷ This would suggest that a placebo effect was not the cause, however, the placebo effect cannot be definitively ruled out.

ONS has been associated with various complications like lead migration and device infection;^10,19 in this study, there was no lead migration or device infection in the 12-month study period demonstrating that in this study ONS is effective and well-tolerated therapeutic option in the management of intractable headache.

The use of metabolic neuroimaging provides a new insight into the pathophysiology of intractable headache. After ONS neuroimaging exhibited metabolic normalisation and the return of balance within dysfunctioning pain control centres in patients with drug resistant chronic cluster headache,¹⁸ it might be useful to correlate the responses with QST.

ONS has been extensively investigated as a treatment for chronic migraine; however, there remains a need for well-designed, randomised controlled studies, with adequately timed double-blind comparison to demonstrate true difference between sham and stimulated patients. Further work needs to correlate the neurophysiological outcomes with clinical outcomes to allow phenotyping of predictors of response to ONS. It would also be worthwhile to compare and correlate the response to neurophysiological tests with metabolic brain imaging in patients undergoing ONS.

Conclusion

We have demonstrated for the first time a sustained reversal of the loss of CPM in patients undergoing ONS for intractable headache as measured by quantitative sensory testing (QST). In contrast, there was no change in peripheral pain thresholds. The mode of action of ONS has not been fully elucidated but may point to the central action of ONS by modulating the TCC that affects central sensitisation.