Abstract
Objective
The objective of this research is to describe novel procedural treatments for hemicrania continua that allow patients to remain off indomethacin.
Methods
Case reports are presented.
Results
We describe four distinct patients with indomethacin-responsive hemicrania continua who were unable to discontinue the use of indomethacin without headache recurrence. No other medications were effective for their syndrome. Secondary causes of headache were ruled out in each case. Each patient underwent diagnostic blockade of either the atlanto-axial joint, C2 dorsal root ganglion or sphenopalantine ganglion depending on their clinical examination and presence of cranial autonomic symptoms. A positive response led to a radiofrequency ablation of the C2 ventral ramus, C2 dorsal root ganglion or sphenopalantine ganglion, which provided headache relief in all case patients as complete as indomethacin. Long-term follow-up of these patients has shown that all have remained essentially headache free without the need for indomethacin. One patient has needed repeat radiofrequency procedures with consistent response.
Conclusion
Hemicrania continua is defined by its sensitivity to indomethacin but very few patients are able to discontinue the medication without headache recurrence. As the risks of chronic indomethacin use are substantial, alternative treatments are necessary to protect patient health. We are now able to suggest several radiofrequency ablation procedures as effective as indomethacin with long-term follow-up.
Keywords
Introduction
Hemicrania continua patients and response to treatment.
HC: hemicrania continua; RF: radiofrequency; SPG: sphenopalatine ganglion; DRG: dorsal root ganglion.
Case 1
A 30-year-old woman with history of episodic migraine since age 22 presented with a new right-sided occipito-nuchal headache with radiation to the frontal region. From the onset it occurred on a daily basis and at the time of presentation had been present for six weeks. There was no triggering event. The baseline pain was of low intensity with exacerbations of severe head pain, which would last for several days. The exacerbations had associated ipsilateral conjunctival injection. On exam the patient exhibited right upper cervical facet irritation on neck extension and rotation plus greater occipital nerve (GON) tenderness to palpation. Her pain, however, was not lessened by a large-volume suboccipital nerve blockade (9 cc of 1% lidocaine with 40 mg triamcinolone). Neuroimaging (magnetic resonance imaging (MRI) brain, magnetic resonance angiography (MRA) brain/neck) was negative for secondary causes of headache. An indomethacin trial was given for a presumed diagnosis of HC (she met International Classification of Headache Disorders, third edition beta (ICHD-3 beta) criteria except for duration of symptoms) (6), and on a dose of 50 mg three times daily she achieved complete headache freedom. She was unable to discontinue the indomethacin after six months of treatment, as her headaches quickly returned (thus now meeting ICHD-3 beta criteria for HC). No other medications, including melatonin, were effective. She then underwent a diagnostic right atlanto-axial injection that provided complete pain relief. This was followed by a radiofrequency ablation of the C2 ventral ramus. This led to complete relief of head pain and cranial autonomic symptoms and allowed her to discontinue the indomethacin. After one year of pain relief, the patient’s headache and associated cranial autonomic symptoms returned. This necessitated the use of indomethacin (with complete pain freedom) until a repeat radiofrequency procedure was performed. The latter provided an additional 21 months of headache and autonomic symptom freedom, which required no additional indomethacin. A third radiofrequency was needed after an additional headache recurrence and she is currently 14 months headache free, without the need for indomethacin. Thus overall the procedures have lasted for 47 plus months in duration. She has had no adverse effects from any of the radiofrequency procedures. She notes that the pain relief is as complete as with indomethacin.
Case 2
A 47-year-old man presented with a 27-year history of daily right-sided fronto-temporal headache that had initially been intermittent but progressed to daily within a few months. He suffered with persistent low-grade pain with exacerbations of severe headache lasting for several days’ duration. During his severe pain attacks, there was associated ipsilateral ptosis, conjunctival injection, lacrimation, rhinorrhea and agitation. His examination demonstrated right GON and upper cervical facet tenderness. His clinical history suggested a diagnosis of HC and as such, an indomethacin trial was completed and he became pain free at a dose of 75 mg three times daily, thus satisfying ICHD-3 beta criteria (6). Unfortunately, he developed significant lower extremity edema requiring discontinuation of the medication. Once off, his persistent daily headache returned and no other medications including melatonin and celecoxib, were able to mitigate his pain. Neuroimaging (MRI brain, MRA brain/neck) was negative for secondary causes of headache. He was referred for anesthesiologic procedures. Based on the presence of significant cranial autonomic symptoms, he underwent a diagnostic right SPG block. This provided complete anesthetic phase pain relief with subsequent radiofrequency ablation of the SPG. At his 2.5-year follow-up, he continues to be pain free in the absence of medication use. He experienced ipsilateral palatal numbness following the procedure that resolved over several months.
Case 3
A 45-year-old woman presented for evaluation of a new-onset daily headache that had been present for seven months. She had a remote history of episodic migraine but had been headache free for 20 years prior to onset of this current headache. Her head pain was strictly right sided with a baseline of mild persistent pain and exacerbations of severe head pain that could last for several days. During her exacerbations, she experienced ipsilateral conjunctival injection and rhinorrhea. On examination she had right GON and upper cervical facet tenderness. She completed an indomethacin trial for presumed HC and achieved pain freedom on 75 mg three times daily. If she tried to reduce her dose by only 25 mg, her headaches returned. She was diagnosed with HC by ICHD-3 beta criteria (6). Melatonin was unsuccessful at treating her head pain. She was referred for anesthesiologic procedures. A C2 DRG block led to complete mitigation of pre-block GON tenderness and headache. She then underwent a C2 DRG rhizotomy that led to complete relief both of head pain and cranial autonomic symptoms. At her 17-month follow-up, she remains pain free. She had no adverse events from the procedures. Neuroimaging (MRI brain, MRA brain/neck) to rule out secondary causes were negative.
Case 4
A 49-year-old woman presented for evaluation of a pure left-sided headache that had been daily from its onset seven years prior. She had no previous headache history. Her pain was located in the left periorbital and frontotemporal region, which she described as a pressure sensation. She reported exacerbations of severe pain that would last for several days with associated ipsilateral ptosis, lacrimation, conjunctival injection and rhinorrhea. Examination revealed left GON tenderness, left-sided scalp allodynia, and left-sided ptosis. She had complete headache relief on indomethacin 50 mg three times daily, and a diagnosis of HC was made using ICHD-3 beta criteria (6). Over time she was able to decrease her daily indomethacin dose to 125 mg but pain recurrence prevented any further decrease. No other medications including melatonin and acetazolamide were effective and as such she was referred for anesthesiologic procedures. A diagnostic C2 DRG block produced headache relief for two months, with only occasional need for indomethacin. The diagnostic procedure was later followed by radiofrequency ablation of the C2 DRG, allowing her significant pain freedom and the ability to reduce her indomethacin use to only three or four days per month (25 to 50 mg in a day), as compared to the pre-procedure dose of 125 mg per day. She is currently 14 months post-procedure and continues to have minimal to virtually no headache. There were no associated procedural complications.
Methods/procedures
The procedures for radiofrequency ablation of the C2 ventral ramus, C2 DRG and SPG have been published elsewhere and are beyond the scope of this manuscript (7–10). Images of the procedures are included as Figures 1–3.
Radiofrequency ablation of the C2 ventral ramus. Radiofrequency ablation of the C2 dorsal root ganglion (DRG). Radiofrequency ablation of the sphenopalatine ganglion (SPG).
Adverse events
Possible adverse events associated with radiofrequency ablation and diagnostic blockade of the SPG and high cervical structures in the neck include: post-operative pain, bleeding and infection; thus most patients are given a single pre-operative dose of clindamycin. Multiple fluoroscopy views (anterior to posterior (AP), lateral, oblique) are used to lessen the potential for complications (especially hematoma from vessel contact during the course of the C2 ventral ramus and SPG procedures). In addition, the SPG procedure carries a 20% chance of producing ipsilateral palatal numbness, which most often resolves within three to six months. In all instances our four case patients had no significant post procedure side-effects (one patient had palatal numbness post-SPG). All stated they would be willing to repeat their procedures again as they had significant headache response to them (case patient 1 already had on several occasions).
Discussion
HC can be a life-long disorder and little is known about its natural history (4,5). It is defined by its responsiveness to indomethacin but the mechanism by which indomethacin eradicates this pain syndrome is unknown (1–4). Several rare cases have been reported in which indomethacin was felt to cure the disorder, but in most, chronic unremitting treatment is necessary and thus the risk of GI and renal complications is extremely high (4,5,11). Aside from indomethacin, there are very few medicinal alternative treatments for HC, including high-dose cyclooxygenase-2 (COX-2) inhibitors, which have the same GI risks as indomethacin upon reaching these doses, melatonin (which has a similar chemical structure to indomethacin), verapamil, Boswellia serrata, and several antiepileptic medications (3,12–14). No treatment has been shown to be as effective as indomethacin. Previously, in patients who could not tolerate indomethacin or who could not take it because of other medical contraindications, options were very limited. Recently, however, several interventional procedures have shown some efficacy for HC. In the realm of anesthesiologic procedures, Guerrero et al. (15) reported on the effects of GON, supraorbital nerve (SON), trochlear nerve or a combination of (SON + GON) blocks in nine patients with indomethacin-responsive HC who could not tolerate the medication long term. Each patient showed some tenderness over the represented peripheral nerves. Five of nine individuals had complete response to the procedures, while the remainder had a partial response, lasting from two to 10 months. The GON and SON blocks used anesthetics (bupivacaine, mepivacaine) while the trochlear block contained a corticosteroid. Weyker et al. (16) used radiofrequency ablation of the SON in three patients with reported HC (two of three showed indomethacin responsiveness, while one never tried indomethacin thus not achieving full diagnostic criteria). After positive response to diagnostic blockade with bupivacaine and triamcinolone, each patient had radiofrequency ablation and showed complete relief of head pain at seven to 12 months’ follow-up (5).
In the realm of neurosurgical interventional procedures, percutaneous peripheral nerve stimulation is another newer option for HC. Schwedt et al. (17) treated two patients with occipital nerve stimulation (ONS) and while neither patient became pain free at three months’ follow-up, there was a significant reduction in headache frequency as well as reduction in pain severity. Each patient suffered complications, however, including stimulator lead migration and one developed an infection. A newer wireless stimulator device called Bion was placed over the GON in six patients with HC in a study reported by Burns et al. (18). Patients were followed for a medium of 13.5 months and while none had pain freedom, four of the study individuals improved substantially (80%–95%), one had 30% improvement, while one worsened by 20%. Adverse events were very minimal with the main issue being prominent paresthesias from overstimulation. The mechanism of action by which pain relief is achieved with GON stimulation in HC is not fully understood (17,18). It has been proposed that GON stimulation in cluster headache must work by central neuromodulation and not via an effect at the peripheral GON site because it can take weeks to months to see any pain relief after stimulator placement (19). This may also be the case for HC, as in the Bion patients there was also a delayed response of weeks before any pain relief was noted (18,19). With all percutaneous stimulatory devices there is a risk of infection, device failure, and even mortality, thus it is our opinion the latter should be considered only “a last-ditch effort” after all medicinal and anesthesiologic procedures have failed.
Our goal was to establish another alternative interventional procedure that is “safer” and has less potential morbidity than the procedure for neurostimulator implantation. The representative cases illustrate that we have a non-indomethacin treatment regimen for HC that achieves essentially a pain-free state with the same consistency of response as indomethacin. Our cases have attained a long duration effect, and this can be replicated if necessary if there is headache recurrence, with minimal to no undue side effects and none of the risks of long-term indomethacin usage. The exact nature of how these procedures are turning off the pain of HC can only be hypothesized because at present we do not know what causes the primary disorder itself. We do know from functional imaging that two areas activate in primary HC, including the hypothalamus and the brainstem (20). Both of these areas can possibly be modulated using the anesthesiologic procedures we are performing on our patients. Hypothetically when a patient undergoes a radiofrequency lesion of the C2 DRG or the C2 ventral ramus, we are modulating input to the trigemino-cervical complex and then through the accessory trigemino-hypothalamic tract we are modulating the hypothalamus itself (21). The trigeminal sensory nucleus in the brainstem can also be modulated via the upper cervical afferents through the trigemino-cervical complex (22). Thus in addition to eliminating head pain, the radiofrequency procedures could potentially alter activation of the trigemino-autonomic reflex and secondarily alleviate cranial autonomic symptoms (which was shown in our case patients).
The SPG has both indirect and direct connections to the hypothalamus and the sensory trigeminal nucleus and is part of the trigemo-autonomic complex. Thus inhibition at the SPG via blockade could suppress the production of head pain and cranial autonomic symptoms (23). SPG modulation (radiofrequency and stimulator placement) has recently been shown to help alleviate cluster headache, which is another hypothalmic-modulated trigemino-autonomic cephalagia like HC, so there is past precedence of modulation of this autonomic ganglia leading to headache and cranial autonomic symptom relief (24,25).
Our suggestions on how physicians should approach the HC patient after making a diagnosis and ruling out secondary causes would be to first treat with typical indomethacin dosing for three to six months (50 mg–150 mg daily of short-acting formulation) to achieve pain freedom and then try to taper off to see if the patient has the curative form of the syndrome or the episodic variant (7). If headaches return then an indomethacin to melatonin switch protocol can be initiated, which has been suggested by one of the authors previously to see if the patient has melatonin-responsive HC (12). If the patient cannot achieve pain freedom on melatonin or get to minimal daily dosing of indomethacin with co-administration of melatonin, then the presented anesthesiologic procedures are recommended. If the patient’s exam demonstrates significant GON or upper cervical facet irritation or if he or she had some response to GON nerve blocks, then we suggest diagnostic C2 DRG or atlanto-axial injections. Corresponding radiofrequency procedures would then include lesions of the C2 DRG or C2 ventral ramus. In addition, if the patient has had cervical post-traumatic HC, then cervical-directed procedures would be recommended as first line. Examination is also key to which cervical procedure is chosen. If, however, the headaches have significant cranial autonomic symptoms and/or there is a lack of cervical symptoms, or minimal to no cervical spine/GON tenderness on examination, then we suggest SPG blockade as the initial intervention. If there is no response to the cervical or SPG procedures, then the other maneuver(s) should be tried. If only partial response is achieved with either procedure, then a combination of procedures is suggested. If these procedures fail, then neurostimulation could be warranted as a next more-invasive approach or the procedures suggested by Guerrero et al. (15) and Weyker et al. (16).
In conclusion, based on our current study and from recent previous investigations, there clearly appears to be promising non-indomethacin, non-neurostimulator, anesthesiologic approaches for the HC patient who requires lifelong indomethacin treatment and who cannot be transitioned to melatonin or other medications (12–14). Our study regimen has long-term follow-up data and has shown consistency of response over time.
Clinical implications
There are few alternative treatments for hemicrania continua outside of indomethacin, which is associated with significant patient morbidity with chronic use. Our suggested pain anesthesiologic procedures, which include radiofrequency ablation of the C2 ventral ramus, C2 dorsal root ganglion and sphenopalatine ganglion, have shown substantial long-term pain relief without the need for indomethacin.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest
None declared.