Cluster headache (CH) is a debilitating, severe form of headache. A novel non-systemic therapy has been developed that produces therapeutic electrical stimulation to the sphenopalatine ganglion (SPG). A transoral surgical technique for inserting the Pulsante SPG Microstimulator into the pterygopalatine fossa (PPF) is presented herein. Technical aspects include detailed descriptions of the preoperative planning using computed tomography or cone beam computed tomography scans for presurgical digital microstimulator insertion into the patient-specific anatomy and intraoperative verification of microstimulator placement. Surgical aspects include techniques to insert the microstimulator into the proper midface location atraumatically. During the Pathway CH-1 and Pathway R-1 studies, 99 CH patients received an SPG microstimulator. Ninety-six had a microstimulator placed within the PPF during their initial procedure. Perioperative surgical sequelae included sensory disturbances, pain, and swelling. Follow-up procedures included placement of a second microstimulator on the opposite side ( n = 2), adjustment of the microstimulator lead location ( n = 13), re-placement after initial unsuccessful placement ( n = 1), and removal ( n = 5). This SPG microstimulator insertion procedure has sequelae comparable to other oral cavity procedures including tooth extractions, sinus surgery, and dental implant placement. Twenty-five of 29 subjects (86%) completing a self-assessment questionnaire indicated that the surgical effects were tolerable and 90% would make the same decision again.
Cluster headache is a primary headache disorder marked by unilateral peri-orbital pain lasting between 15 min and 3 h, which follows a circadian and/or circannual pattern. Cluster headache is known as the ‘suicide headache’ due to the severity of the pain and the repetitive nature of the headache, and includes ipsilateral autonomic symptoms. Cluster headache exists in two clinical forms, episodic and chronic, with an overall annual prevalence of approximately 0.12%. Episodic cluster headache occurs in periods lasting 7 days to 1 year, separated by pain-free periods of 1 month or longer. Approximately 15% of patients suffer from chronic cluster headache, in which headaches occur without remission or with remission lasting less than 1 month during a year. Given the relentless nature of their disorder, cluster headache sufferers continue to search for therapies to resolve their headaches.
Abortive medications including sumatriptan injection and oxygen inhalation are often effective, however not all patients have relief, and many experience significant side effects from these therapies. The brief duration of attacks makes abortive therapy a challenge, and preventive medications including verapamil, lithium carbonate, divalproex sodium (valproate), corticosteroids, methysergide, melatonin, and topiramate are almost always provided. Side effects of preventive medications can be significant, ranging from nausea and fatigue to hypotension, bradycardia, AV block, and myocardial infarction. A new treatment option comprising sphenopalatine ganglion (SPG) stimulation via an inserted miniaturized microstimulator (Pulsante Microstimulator System, previously referred to as the ATI Neurostimulation System) has been made available as both an acute and preventive therapeutic option.
The SPG, also called the pterygopalatine ganglion, is located within the midface in the pterygopalatine fossa (PPF), which also contains the maxillary nerve emerging from the foramen rotundum and the terminal branches of the maxillary artery. The PPF is a defined anatomical bilateral space bounded anteriorly by the maxilla, posteriorly by the medial plate of the pterygoid process and greater wing of the sphenoid process, medially by the palatine bone, and superiorly by the body of the sphenoid process. Its lateral border is the pterygomaxillary fissure (PMF), which opens to the infratemporal fossa.
The SPG is a complex neural extracranial ganglion with multiple neural roots, including autonomic, sensory, and motor. The maxillary branch of the trigeminal nerve and the nerve of the pterygoid canal, also known as the vidian nerve, which is formed by the greater and deep petrosal nerves, send neural projections to the SPG. The fine branches from the maxillary nerve (called the pterygopalatine nerves) form the sensory component of the SPG. These sensory fibres pass through the SPG, without synapsing, to be distributed with the nerves that arise from the SPG.
The autonomic root of the SPG comprises both parasympathetic and sympathetic neurons and fibres. The parasympathetic root is conveyed to the SPG via the greater petrosal nerve, and the sympathetic root via the deep petrosal nerve. The greater petrosal nerve carries the preganglionic parasympathetic axons from the superior salivary nucleus, located in the pons, to the SPG. These fibres synapse onto the postganglionic parasympathetic neurons within the SPG. The deep petrosal nerve contains postganglionic sympathetic fibres which arise from the superior cervical sympathetic ganglion and merge with the greater petrosal nerve to form the nerve of the pterygoid canal (vidian nerve). The deep petrosal nerve and the vidian nerve carry sympathetic fibres to the SPG, and these fibres pass through the SPG without synapsing.
Postganglionic parasympathetic fibres that arise in the SPG are distributed through the ophthalmic and maxillary divisions of the trigeminal nerve to the lacrimal gland, nasal glands, palatine glands, and pharyngeal glands. In addition, numerous postganglionic parasympathetic branches have been shown to course superomedially from the SPG into the orbital cavity, which provide parasympathetic innervations to the meningeal and cerebral blood vessels.
The SPG is believed to play a role in the generation of headache pain and cranial autonomic symptoms associated with cluster headaches. The pain experienced during a cluster headache is thought to be a result of activation of the trigeminal-autonomic reflex. Cluster headache is a neurovascular disorder in which neural elements cause vessel dilation and/or activation of trigeminal nociceptive fibres, which is perceived as referred pain. These inputs also trigger a reflex connection between neurons in the pons, in the superior salivary nucleus, which results in an increase in cranial parasympathetic activity that is mediated through the SPG. Postganglionic parasympathetic fibres from the SPG are known to innervate the cerebral and meningeal blood vessels, and when activated, release neuropeptides that are known to be strong vasodilators that can directly or indirectly activate sensory trigeminal fibres, causing further activation of the pain pathway and parasympathetic outflow. The reflex then acts as positive feedback, causing further activation of the trigeminal-autonomic reflex, which can both initiate and sustain pain.
Due to the role of the SPG in the manifestation of the cranial autonomic symptoms and in initiating and sustaining headache pain (trigeminal-autonomic reflex), the SPG has been a target for clinical treatment and interventions for headache pain for over 100 years. Since Sluder first described the application of cocaine or alcohol to the SPG for the treatment of headaches, the SPG has been the target of a variety of surgical and non-surgical interventions for the treatment of headaches, including percutaneous alcohol injection, lidocaine or corticosteroid application, radiofrequency lesioning, and neuromodulation.
More recently, Schoenen et al. reported on 32 patients implanted with the Pulsante Microstimulator System, which includes a miniaturized implantable device with an integral lead containing six electrodes. Twenty-eight patients completed the randomized experimental period and 67.1% of the attacks treated with full stimulation resulted in pain relief, compared to 7.4% of sham-treated attacks ( P < 0.0001). In addition, 36% of patients experienced a ≥50% decrease in attack frequency versus baseline. Overall, 68% of the patients experienced a clinically significant improvement during the study by achieving pain relief in at least 50% of their full-stimulation treated attacks and/or experiencing at least a 50% reduction in cluster attack frequency compared to baseline.
The Pulsante Microstimulator System has received a CE mark in Europe and is labelled for the acute treatment of cluster headache; in some patients it has been associated with a reduction in the number of cluster headaches.
A detailed description of the surgical insertion procedure for the Pulsante SPG Microstimulator is presented herein, as well as the recommended preoperative surgical planning techniques, preoperative patient assessments, surgical techniques used during the insertion procedure, and postoperative instructions. Furthermore, a review of the expected surgical sequelae associated with this procedure is provided. While previous publications have described an overview of the insertion technique, this work presents a highly detailed description, as well as surgical outcomes from the largest cohort of patients analyzed as a whole to date. Clinical outcomes related to the effect on cluster headache are presented elsewhere. All data analyzed in this review were collected in the course of clinical studies which were approved by the local ethics committees (ClinicalTrials.gov: NCT01255813 , NCT01616511 , and NCT01677026 ). Each patient gave informed consent prior to inclusion. While aspects regarding patient selection and management are not provided here, an expert consensus has been published previously.
Materials and methods
Ninety-nine patients diagnosed with cluster headache (International Classification of Headache Disorders 2nd edition criteria (ICHD-2)) by a physician specializing in headache care (headache neurologist, pain specialist, and general neurologist), who presented for insertion of an SPG microstimulator through May 2014, were analyzed. These 99 patients included 32 Pathway CH-1 study patients who were part of a previously reported cohort, 11 continued access Pathway CH-1 patients, and 56 patients who participated in the Pathway R-1 registry. To date, no patient has been excluded from surgery due to anatomical constraints. A total of 18 surgeons across Europe (Belgium, Denmark, France, Germany, and Spain) performed these insertion procedures.
The Pulsante Microstimulator System (Autonomic Technologies Inc., Redwood City, CA, USA) provides a novel, non-systemic therapy designed to deliver patient controlled, on-demand stimulation of the SPG. The microstimulator is a miniaturized implantable device including the device body, an integral lead with six stimulating electrodes, and an integral fixation plate ( Fig. 1 ). The microstimulator is inserted such that the stimulating electrodes are positioned within the PPF proximate to the SPG on the side of the patient’s most prevalent headaches, with the body positioned on the lateral-posterior maxilla medial to the zygomatic arch, and anchored to the zygomatic process of the maxilla using the integral fixation plate. The microstimulator is inserted transorally using a minimally invasive gingival buccal approach, with the aid of custom surgical tools and intraoperative imaging (e.g., fluoroscopy). It is powered and controlled by a handheld, rechargeable remote controller held to the patient’s cheek ( Fig. 1 ).
The midface anatomy, in particular the PPF and PMF, is variable both intra- and inter-patient. Due to these variations, it is recommended that each patient receive a preoperative head computed tomography (CT) or cone beam computed tomography (CBCT) scan with slice thickness between 0.5 and 1.0 mm. To exclude possible dental pathology (e.g., apical periodontitis, etc.), a preoperative panoramic and/or dental X-ray of the maxillary premolars and molars on the side to be operated is also recommended. Additional preoperative anatomical evaluation should include the following: accessibility of the PPF, signs of regional infection, signs of osteodestructive disease, and pre-existing osseous defects.
PPF accessibility is determined by the shape, width, and height of the PMF, which is the lateral opening of the PPF. The PMF has a well-defined posterior margin formed superiorly by the lateral margin of the anterior surface of the base of the pterygoid process and inferiorly by the fused pterygoid plate with the maxilla. The anterior margin is formed by the curved contour of the posterior wall of the maxillary sinus. Typically, the PMF has an inverted triangle shape, but in rare cases it may be closed or fused to the posterior wall of the maxillary sinus. Understanding the PMF geometry is fundamental to the neurostimulator insertion procedure, since the integral lead must enter the PPF through the PMF. A minimum PMF width of ≥1.2 mm is recommended for surgical eligibility, as the integral lead diameter of the neurostimulator is 1 mm.
Additional scrutiny and care should be considered for any patient with signs of regional infection (e.g., active sinusitis or a dental pathology such as pericoronitis from a semi-impacted wisdom tooth). If any active infection is identified ipsilateral to the planned insertion site, the infection should be treated prior to attempting the insertion procedure. Patients with severe osteoporosis, especially those with iatrogenic osteoporosis caused by long-term steroid treatment, a common treatment for cluster headache, or those with extreme maxillary atrophy, should also be considered with caution due to an increased risk of posterior maxillary wall fracture or microstimulator misplacement. In addition, patients with previous anterior skull base surgeries or other osseous defects should be evaluated for eligibility based on surgical anatomy and accessibility of the PPF.
Once surgical eligibility is established, additional custom surgical planning is performed by the manufacturer and a recommendation for microstimulator length and target positioning is provided. Four lengths are available: short (3.6 cm), medium (4.4 cm), long (5.2 cm), and extra long (6.0 cm); the design is flexible to various anatomical ranges measuring ±4 mm from the nominal length. The distance between two fiducial landmarks is used to determine the appropriate microstimulator length: the superior aspect of the sphenopalatine foramen, which is the opening that connects the nasal cavity with the PPF, and the lateral-posterior aspect of the maxilla medial to the zygomatic bone. These landmarks are obtained from a three-dimensional (3D) rendering of the patient-specific anatomy (created from the preoperative CT or CBCT scan using Mimics Software (Materialise, Leuven, Belgium)). Additional considerations for sizing recommendations include positioning within the anatomy and microstimulator body depth relative to the surface of the patient’s cheek. Since the microstimulator is a radiofrequency device with power and control provided by an external remote controller through induction, the depth of the microstimulator body is critical for proper functionality.
In conjunction with providing a recommended length, the manufacturer provides a recommendation of the target placement using the patient-specific anatomy. Using Mimics, an exact model (stereolithography (STL) model) of the microstimulator is placed digitally within the patient-specific anatomy ( Fig. 2 ). The stimulating electrode target location is the putative location of the SPG, typically located posterior to the middle nasal turbinate, between the vidian canal and the foramen rotundum. The vidian canal is located in the posterior superomedial aspect of the PPF and the foramen rotundum location is superolateral in the PPF. The target location for the digitally placed microstimulator lead is the superomedial aspect of the PPF, more specifically at the superior aspect of the vidian canal and as far medial in the PPF as possible, against the lateral nasal wall, without entering the nasal cavity ( Fig. 3 ). The target position for the microstimulator body is also provided, taking into consideration the various curvatures of the posterior-lateral maxilla. Using these techniques, the surgeon gains a clear understanding of the patient-specific anatomy and can pre-identify anatomical challenges.
Finally, a two-dimensional (2D) representation of the 3D patient-specific anatomical rendering (digitally reconstructed radiograph or DRR) with the digitally placed microstimulator is created. During the procedure, fluoroscopic imaging allows for visualization of the microstimulator and surgical instruments. The 2D representation can be used to directly compare the live images taken during the procedure to those of the preoperative planning images. The DRR is created using custom add-in software for an advanced volume visualization software package (VolView; Kitware Inc., New York, USA), can be viewed at any orientation to directly match the intraoperative fluoroscopy images (specifically the anterior–posterior (AP) and lateral views) ( Fig. 4 ), and serves as a map to guide the surgeon towards the targeted microstimulator placement.
Intraoperative and perioperative management
The insertion procedure involves an oral mucosa incision with limited sub-periosteal tissue dissection from the lateral and posterior maxilla to place the stimulating electrodes into the PPF, and fixation of the microstimulator body to the zygomatic process of the maxilla. To date, all procedures have been performed under general anaesthesia in order to achieve working conditions for an accurate placement. With general anaesthesia, oral intubation is recommended with the intubation tube placed on the opposite side of the oral cavity from the planned microstimulator insertion. Nasal intubation may also be done; however resulting fluoroscopic images may be more difficult to assess. The patient should be placed supine on a fluoroscopy compatible table with the chin slightly elevated to allow for easier access to the oral cavity.
A single preoperative injection of an appropriate antibiotic with extended spectrum to anaerobes (such as aminopenicillins with beta-lactamase inhibitors, clindamycin, or newer macrolides) per local guidelines or standards is recommended such that sufficient levels are achieved pre-procedure. Oral decontamination prior to surgery is highly recommended and should include both a mouth rinse, e.g., chlorhexidine solution (0.12% solution), and a scrub of the oral mucosa with povidone–iodine swabs. These practices, along with proper preoperative dental care, constitute the best chance of avoiding infections in this sterile contaminate procedure.
In addition to general anaesthesia, it is recommended that 2–5 ml of local anaesthesia (e.g., xylocaine 1–2% or bupivacaine 0.5%) with 1:100,000 epinephrine is applied to the buccal gingiva prior to the mucosal incision to reduce intraoperative bleeding and postoperative pain. The minimally invasive microstimulator placement is done using a transoral approach with a 1.5- to 2-cm incision in the gingival crevice over the posterior maxillary buttress or paramarginal along the molars, leaving an inferior cuff of mucosa to allow ease of suture closure. Using a standard periosteal elevator (e.g., Freer rougine), the mucosal tissue is elevated from the underlying bone superiorly and laterally, creating an area in which to anchor the microstimulator fixation plate and an area for insertion of the microstimulator body. Care should be taken to maintain a sub-periosteal plane to avoid exposure of the buccal fat pad which, if exposed, can decrease the surgical field of view.
The procedure is assisted with the use of the ATI Surgical Introducers ( Fig. 5 ) provided by the manufacturer. Recommended surgical techniques may vary based on per-patient anatomy; variations are highlighted below. Following posterior lateral maxilla preparation, a sub-periosteal dissection from the posterior lateral maxilla to the PMF is done using the SI-100, a curved sub-periosteal elevator. While the SI-110 can be used, it is recommended that the initial dissection be performed with the straight tip of the SI-100. The SI-100 and SI-110 surgical introducers allow for blunt atraumatic sub-periosteal dissection while maintaining close contact to the posterior wall of the maxillary tuberosity to avoid trauma to the surrounding tissues. Both are malleable instruments that can be optimized for the contour of the posterior maxillary sinus curvature to ensure sub-periosteal dissection and avoid the soft tissue of the infratemporal fossa to minimize intraoperative bleeding. During SI-100 insertion, periodic use of intraoperative imaging, e.g. fluoroscopy, is recommended to verify surgical introducer location. Both instruments will meet resistance once placed at the PMF. In some cases, the width of the PMF may be larger than the thickness of the instruments and as such, care should be taken to not over-advance instruments into the PPF. Once the SI-100 is placed at the entrance to the PPF, it is important to aim the tip towards the superomedial aspect of the PPF (identified with the aid of the preoperative DRR images). The microstimulator is then inserted with the aid of the SI-120, Shielded Surgical Introducer.
Neurostimulator functionality should be tested prior to insertion by confirming communication and powering through impedance testing in a non-metallic bowl with sterile saline. Following testing, the fixation plate must be bent appropriately to the anatomy of the posterior lateral and zygomatic processes of the maxilla, with slight over-bending to avoid movement of the stimulating electrodes during anchoring. The microstimulator is then loaded into the SI-120 which is inserted using the existing surgical plane created by the SI-100. Periodic images are recommended to visualize the SI-120 position and to maintain the trajectory towards the superomedial aspect of the PPF. Near the PPF entrance, additional resistance may be encountered as the SI-120 enters the soft tissue space, which is required to achieve a posterior superomedial placement of the first stimulating electrode. The additional rigidity of the SI-120 is of particular usefulness in achieving proper placement by facilitating more control during insertion. Once target placement is achieved and verified with the DRR images, the microstimulator is pushed slightly forward to remove it from the SI-120 anchoring hub. The body is held against the posterior lateral aspect of the maxilla, the SI-120 retracted as the sheath opens around the lead, and the microstimulator is left behind in the anatomy.
In some anatomical variations, SI-120 use may be precluded and the SI-110 and the LB-100 are used. The SI-100 and SI-110 differ in the design of the distal tip ( Fig. 5 ); the SI-110 may be placed at the entrance to the PPF and the Lead Blank, LB-100, is used to create an implant path within the PPF. Subsequent placement is achieved by sliding the microstimulator along the insertion groove and split tip of the SI-110. Periodic images are recommended to verify microstimulator location and implant trajectory during insertion.
The microstimulator is anchored using the most distal fixation plate hole (the hole closest to the microstimulator body) using a standard midface craniofacial screw (1.5–1.8 mm in diameter, 4.0–6.0 mm in length at a level superior to the apices of the dental roots). Placement is confirmed by comparing fluoroscopic (AP and lateral) and DRR images. In addition, an intraoperative 3D-CBCT scan can be performed, providing more accurate verification of intraoperative SPG neurostimulator placement. In some cases, anchoring may cause the integral lead to move due to the rocking of the microstimulator on the zygomatic process of the maxilla. Thus, careful pre-bending of the fixation plate and anchoring of the first screw is critical. Also, the entire surgical procedure must be kept as atraumatic as possible with thoughtful attention to all resistance met and gentle probing for a trajectory to allow the microstimulator placement according to plan.
After anchoring, but prior to incision closure, electrode impedance testing should be performed to ensure proper functionality. Wound closure is then performed using 4–0 resorbable or non-resorbable sutures.
Postoperative patient instructions should include recommendations for avoiding undue pressure to the surgical area, soft diet, no smoking, no gum chewing, and proper oral hygiene with a chlorhexidine 0.12% mouth rinse twice a day for the first week. Analgesics may be prescribed as deemed appropriate. Within 7–10 days of the procedure, intraoral wound control and suture removal should be completed. Any postoperative wound infections should be treated by surgical wound debridement and oral or intravenous antibiotics as appropriate.
On the first postoperative day, CT or CBCT scans should be taken to confirm proper placement of the microstimulator. If location revision is advised, the microstimulator should be removed and replaced within the first days after the initial procedure to achieve ideal placement.
Explant procedures should be performed under local anaesthesia using standard protocols to open the incision and remove the bone screws from the fixation plate.
Ninety-nine cluster headache patients who presented for insertion of an SPG microstimulator as a part of either the Pathway CH-1 study or the Pathway R-1 registry were analyzed. Baseline patient characteristics are provided in Table 1 .