Combined Occipital Nerve and Sphenopalatine Ganglion Neuromodulation for Refractory Craniofacial Pain with Trigeminal Autonomic Features: A 12-Year Follow-Up (Case Report)

Background

Craniofacial complex regional pain syndrome (CRPS) is a rare but clinically significant condition in which features of CRPS – including allodynia, hyperalgesia, and sympathetically-mediated autonomic disturbances – occur in the trigeminal distribution following craniofacial trauma or surgical procedures. It does not correspond to a distinct ICHD-3 category; the closest equivalents are Painful Post-Traumatic Trigeminal Neuropathy (PPTTN, ICHD-3 13.1.2) for post-injury presentations, Persistent Idiopathic Facial Pain (PIFP, ICHD-3 13.12) when the mechanism is less clearly defined, or Chronic Paroxysmal Hemicrania (CPH, ICHD-3 3.2.2), which belongs to the TAC group. Its clinical phenotype may closely mimic TACs, making differential diagnosis challenging. The Budapest criteria, developed for limb CRPS, require adaptation for craniofacial cases, and very few reports in the literature describe cases meeting these criteria in the orofacial or craniofacial region.^1–4

Trigeminal autonomic cephalalgias (TACs) are characterized by severe facial pain and headache accompanied by autonomic symptoms, such as rhinorrhea, nasal congestion, conjunctival injection, lacrimation, and changes in facial color and temperature.^5,6 Most TACs are unilateral and include various pain syndromes ranging from paroxysmal hemicranias to chronic cluster headaches.^5–7 Treatment strategies for chronic cluster headaches include both acute and preventative approaches. Acute treatments comprise oxygen, triptans, and non-invasive neuromodulation.^8,9 Preventive strategies involve various medications (verapamil, lithium, topiramate, gabapentin), interventional procedures, and neuromodulation techniques targeting occipital nerves, SPG, and the vagus nerve.^8–11

Pharmacological management of neuropathic pain follows a stepwise approach. First-line agents include α2δ-ligand inhibitors, tricyclic antidepressants, and serotonin-norepinephrine reuptake inhibitors. Second-line options include tramadol, topical lidocaine, and 8% capsaicin patches. For refractory cases, NMDA receptor antagonists, cannabinoids, and botulinum toxin A represent third-line interventions. When pharmacological measures fail, interventional and neuromodulation therapies become essential components of management.¹²

Despite this wide spectrum of therapeutic measures, many patients continue to experience debilitating symptoms, highlighting the need for novel combinatorial approaches.¹³

Case Report

A 41-year-old male presented to our pain clinic with severe right-sided hemicranial and facial pain. His medical history included COPD, diabetes mellitus, nephrolithiasis, and two episodes of transient CRPS syndrome following orthopedic surgeries (left upper and lower extremity CRPS-I following knee and shoulder surgeries, each resolving over approximately 8–12 months with conservative management and sympathetic blocks, predating the current craniofacial pain by several years). This prior history of CRPS suggests a constitutional predisposition to complex regional pain responses.

The patient reported severe persistent headache and facial pain with superimposed paroxysmal attacks that began following occipital plastic surgery for excision of a nevus. The pain appeared as right hemicrania, most severe in temporal area and right eye. Pain attacks were accompanied by severe autonomic symptoms including facial edema, redness, ptosis, lacrimation, and facial allodynia predominantly on the right side. The pain was accompanied by some photophobia, without photophobia, nausea and/or vomiting. The patient was not agitated during the attack. Attacks typically last 2–6 hours, occurring 1–2 times daily, with pain severity of 9–10/10 on the Numerical Rating Scale (NRS) during attacks and 5–6/10 for background pain.

Diagnostic workup included triphasic bone scan, brain MRI (contrast-enhanced 1.5T; no trigeminal nerve enhancement, no structural lesion, no white matter changes), and lumbar puncture, all without significant findings. After psychiatric and psychological evaluations, a diagnosis of refractory craniofacial pain was established, with features consistent with craniofacial CRPS and a TAC-like phenotype (see Discussion) was established.

The patient had been treated with numerous medications including indomethacin, other NSAIDs, oxygen, calcium channel blockers, topiramate, pregabalin, gabapentin, carbomazepine, lamotrigine, antidepressants, neuroleptics, opiates, benzodiazepines, medical cannabis, and intravenous infusions of lidocaine and magnesium in appropriate dosages for at least two months each. He also underwent multiple interventional procedures including trigger point injections, occipital nerve blocks, suboccipital compartment blocks type1 and 2, superficial cervical plexus blocks, stellate ganglion blocks, SPG blocks, and pulsed radiofrequency treatments of the SPG and the Gasserian ganglion, C2 dorsal root ganglion pulsed radiofrequency. Unfortunately, none provided significant relief.

Given the refractory nature of his pain, the patient underwent ONS implantation. This provided 30–50% pain reduction but did not allow medication discontinuation. The patient used the system continuously, with specific stimulation parameters: proximal and distal contacts as anodes, internal contacts as cathodes, amplitude 5.5–5.7mA (daytime) and 2.2–4.7mA (nighttime), pulse width 700µs, and frequency 40Hz (right) and 50–80Hz (left).

Despite improvement in headache, significant facial pain persisted. Seven years after ONS implantation, the patient underwent an additional SPG neuromodulation on the right side, targeting complementary neural pathways (Figure 1a and b).

Figure 1 SPG and ONS neuromodulation leads and IPG’s position: AP (a) and Lateral (b) view, CT. 1 – Occipital lead, 2 – SPG lead, 3 – ONS IPG, 4 – SPGS IPG.

During the 12-year follow-up, several device-related complications occurred, including 6 ONS system revisions due to lead migration, fracture, infection, and pressure sores. The SPG system required one revision due to a cervical decubitus. No serious adverse events such as CNS infection, hemorrhage, or permanent neurological deficits occurred.

The SPG stimulation parameters included distal contact as anode, middle contacts as cathodes, 2.0mA amplitude, 530µs pulse width, and 40Hz frequency continuously.

Following all revisions and changes, the patient has two Montage© IPG systems of Boston Scientific with bilateral occipital octrodes and one right sphenopalatine fossa octrode.

Long-Term Follow-Up

Following SPG implantation, the patient experienced a marked reduction in pain intensity. At present, after 12 years of follow-up (including 4 years post-SPG implantation), he rates his pain as 1–2 out of 10 on the Numerical Rating Scale (NRS).

He reports occasional brief exacerbations during weather changes, but pain intensity rarely exceeds 3/10. The patient continues using medical cannabis but has discontinued all other pain medications. His quality of life improved significantly, enabling marriage and family life with two children, though he did not return to his previous occupation as a driver.

Discussion

Neuromodulation in the head and facial region presents technical challenges due to complex anatomy and pain mechanisms yet has emerged as a promising approach for refractory pain conditions.^14,15 Our patient’s clinical presentation is most consistent with craniofacial complex regional pain syndrome (CRPS), representing a rare but important diagnostic category. The clinical phenotype - post-traumatic onset following cranial surgery, persistent allodynia, skin color and temperature asymmetry, prominent autonomic features, and a prior constitutional history of limb CRPS - collectively fulfills the Budapest criteria adapted for the craniofacial region. In ICHD-3 terminology, this maps most closely to Painful Post-Traumatic Trigeminal Neuropathy (13.1.2), without actual trigeminal pathology (the surgical procedure was done in occipital area). The TAC-like phenotype (periorbital pain, lacrimation, ptosis, rhinorrhea) reflects the sympathetically mediated autonomic involvement that characterizes both conditions and does not exclude CRPS as the primary mechanism. The diagnostic complexity is further supported by the failure of indomethacin (which would be expected to respond in paroxysmal hemicrania and hemicrania continua) and triptans, and by the non-classical attack duration of 2–6 hours (atypical for cluster headache, which typically lasts 15–180 minutes according to ICHD-3 criteria).^1,2 The diagnostic complexity highlights the overlap between these conditions, both characterized by sensory, autonomic, and psychological disturbances with similar treatment approaches.

The pathophysiology underlying these conditions involves sophisticated neural pathways. TACs engage both the trigeminovascular system and autonomic nervous system, with the hypothalamus playing a crucial role.^5,6 The trigeminal cervical nucleus extends as far as the C4 level, explaining why ONS (targeting greater occipital nerve from C2) can benefit trigeminal neuralgias and primary headaches.^16,17

Our approach of targeting complementary neural structures has solid neuroanatomical foundations. ONS modulates the trigeminocervical complex through retrograde mechanisms, while SPG stimulation directly targets parasympathetic outflow responsible for autonomic features.^8–10,17 The SPG’s complex neural structure includes sensory, parasympathetic, and sympathetic components, making it particularly relevant for TAC autonomic symptoms. SPG stimulation also influences the V2 trigeminal branch, affecting the trigeminovascular system.¹⁸ This explains the differential response to each modality in our case, with ONS more effective for the occipital component and SPG stimulation addressing the facial pain component.

The sequential implantation approach used has distinct advantages over simultaneous implantation. It allowed for clear assessment of each modality’s specific effects and independent optimization of each system. This staged approach may be preferable in complex cases where the relative contribution of different pain generators is unclear, providing valuable information about which neural pathways predominate in an individual patient’s pain condition.

Our case aligns with emerging evidence supporting both ONS and SPG stimulation for refractory craniofacial pain. The ICON study demonstrated ONS efficacy in medically intractable chronic cluster headache,¹⁰ and the recent investigator-initiated, double-blind, randomized, placebo-controlled HortONS study by Fogh-Andersen et al confirmed the safety and efficacy of burst-mode occipital nerve stimulation in chronic cluster headache.¹⁹ Randomized trials further support SPG stimulation for both acute and preventive treatment.^8,9 However, literature on combined approaches remains limited, particularly with extended follow-up.

Recent reports have demonstrated encouraging results with combined neuromodulation approaches. Ma et al¹³ showed that dual neurostimulation (peripheral and trigeminal ganglion) in herpes zoster ophthalmicus pain syndrome provided better control compared to peripheral neurostimulation alone. William et al²⁰ reported combined trigeminal and sphenopalatine ganglion stimulation for intractable craniofacial pain, while Hayek et al²¹ described successful management of post-traumatic trigeminal neuropathic pain with combined peripheral nerve and SPG stimulation. Most recently, Andreani et al²² reported long-term effectiveness of combined unilateral SPG and occipital nerve stimulation in a cohort of seven patients with refractory chronic cluster headache, with a mean follow-up exceeding eight years – the largest series to date of this combined approach and a direct parallel to the strategy described in our case.

The 12-year follow-up in our case is particularly valuable, as it demonstrates sustained efficacy of combined neuromodulation, suggesting that tolerance does not necessarily develop with appropriate device management. For patients with sympathetically maintained orofacial pain refractory to pharmacological and interventional measures, cervical spinal cord stimulation represents an additional neuromodulation option, as demonstrated by Javed et al in a recent case series and narrative review,²³ further expanding the therapeutic armamentarium available for refractory craniofacial pain. From a research standpoint, the report suggests that further investigation into multimodal neuromodulation strategies may be warranted. However, the implications of this case should be interpreted cautiously. As a single-patient report with diagnostic uncertainty and limited standardized outcome measures, the findings cannot be generalized. Prospective studies and larger case series would be necessary to determine whether similar outcomes can be consistently achieved in broader patient populations. Also, given the 12-year follow-up, the lack of longitudinal outcome data is a major limitation. Our patient’s favorable response may be attributed to several predictive factors: predominantly unilateral pain distribution, distinct autonomic features, absence of significant structural pathology, and careful selection of complementary neuromodulation targets.

Conclusion

This case demonstrates that combined ONS and SPG stimulation can provide effective, sustained pain relief in carefully selected patients with refractory craniofacial pain, whether classified as craniofacial CRPS (ICHD-3 13.1.2 / 13.12) or secondary TAC with autonomic features. The remarkable 12-year follow-up confirms the long-term benefit and safety of this approach. While our experience is limited to a single case, it suggests that combined neuromodulation warrants further investigation through prospective studies with larger cohorts. For practitioners treating refractory craniofacial pain - including craniofacial CRPS with autonomic features - this case highlights the value of sequential or combined neuromodulation strategies when single-modality approaches provide incomplete relief. It also underscores the importance of long-term follow-up in assessing neuromodulation interventions and suggests potential predictive factors for favorable outcomes.