Glossopharyngeal neuralgia, or vagoglossopharyngeal neuralgia as some would prefer,1 is a rare condition, occurring with a frequency of about 1% of that of trigeminal neuralgia. Medical treatment, particularly with carbamazepine, is usually effective. A significant number of patients do, however, become refractory and go on to surgical treatment. The best established surgical treatment is rhizotomy of the glossopharyngeal and upper vagal nerve roots, which seems to be invariably effective if the diagnosis is correct although it is not without morbidity and even mortality.2 Late recurrence after such treatment, as described below, has not previously been reported and raises interesting issues of mechanism and method of treatment which are considered in this brief report.

The patient initially presented in 1988 as a 23 year old woman with typical glossopharyngeal neuralgia, experiencing severe intermittent pain in the left side of the throat, the back of the tongue, and the ear. The pain was aggravated by talking and swallowing and relieved, to some degree, by pressure on the left side of the neck. At first there was a good response to carbamazepine. When medication was stopped after several months the pain returned and was less well controlled with a further course of the drug. Neurological examination, CT, and MRI were normal. In 1989 she underwent posterior fossa craniectomy and exploration of the IXth and Xth cranial nerve roots. No lesion, in particular no vascular compression, was identified. The left IXth nerve root and the two uppermost Xth nerve rootlets were divided adjacent to the brain stem. Her postoperative course was uncomplicated and she remained entirely symptom free for over 9 years.

In 1998, now aged 33 years, she developed recurrence of her original pain which she described as essentially identical to that at the initial presentation. Again the pain responded to carbamazepine but required a high dose (1200 mg daily) which was accompanied by troubling side effects (drowsiness and dizziness). In addition she was not completely pain free. Neurological examination and further MRI were normal. In October 1998 a further posterior fossa exploration was carried out. The previously divided nerve roots were identified and the completeness of the initial section confirmed. There was now, however, a large, ectatic vertebral artery to which the proximal ends of the previously sectioned roots were adherent and which was distorting the remaining Xth nerve rootlets and the XIth nerve. A microvascular decompression was carried out with a Teflon patch being placed between the ectatic artery and the normal and previously sectioned nerve roots. The procedure was without complication and the patient has remained well and entirely pain free since that time (18 months).

The first description of glossopharyngeal neuralgia is credited to Weisenberg in 1910, in a patient in whom the pain was secondary to a cerebellopontine angle tumour. The pain is characteristic although two variants have been described; an otitic form with pain predominantly deep in the ear, in the external acoustic meatus, and the mastoid region and an oropharyngeal form in which the pain is experienced in the pharynx, the tonsillar area, the soft palate, and the posterior third of the tongue. For patients refractory to medical treatment several surgical options are available including extracranial avulsion, intracranial preganglionic root section, trigeminal tractotomy, either open3 or percutaneous CT guided, and microvascular decompression. As mentioned at the outset, intracranial root section has been the most often employed and is generally regarded as curative. It was, however, realised early that section of the upper vagal rootlets is important in that some cases without the additional section were either not relieved or experienced early recurrence.2More recently microvascular decompression has been employed, particularly by Jannetta et al, with complete relief of pain in 76% and substantial improvement in a further 16% in the largest series, with a mean follow up of 48 months.4 As with trigeminal neuralgia, the actual incidence of presumed causative “neurovascular conflict” and, indeed, the exact mechanism of causation are as yet unresolved questions.

The particular dilemma posed by the present case had both a diagnostic and a therapeutic arm, referable in each case ultimately to mechanism. The only reported cases of recurrence after preganglionic section are the small group, referred to above, in whom only the IXth nerve root had been cut and who subsequently responded to section of the upper vagal rootlets and one patient who had had both a IXth and partial Xth rhizotomy and who later responded to a trigeminal nerve procedure. In these cases failure was typically either immediate or not long delayed. In the largest series reporting the results of treatment2and in a smaller series with long follow up5 there were no recurrences after preganglionic section of the IXth and upper Xth roots. Likewise, after total sensory root section via the posterior fossa (Dandy procedure) trigeminal neuralgia does not recur.

In our case, assuming completeness of the initial section, there seemed to be, essentially, three possible explanations for the recurrent pain. Firstly, that the pain was due to involvement of the remaining non-trigeminal somatic afferent components of the spinal trigeminal nucleus (in the VIIth and the remainder of the Xth cranial nerves); secondly, that it was some form of postdenervation pain akin to the anaesthesia dolorosa described after Vth nerve section1and in one instance after IXth nerve section; and, thirdly, it was a form of trigeminal neuralgia, there being a reported coincidence of the two forms of neuralgia in a few cases.2 The close similarity of the recurrent to the initial pain, both in nature and site, the long period since the initial pain, and the response to carbamazepine all favoured the first possibility. On the basis of this diagnosis, coupled with the patient's relative youth and, it must be said, her strong insistence, re-exploration was undertaken as described above. Whatever one's position on the vascular compression theory of the cause of cranial nerve neuralgias the findings were impressive and, in conjunction with the undesirability of further nerve section, encouraged treatment by microvascular decompression alone. The immediacy of pain relief, sustained now for 18 months, supports this decision. It might be argued that the presumed vascular compression was overlooked at the first procedure but there are several points against this. Firstly, both procedures were performed by the same surgeon, experienced in posterior fossa surgery, and the area of compression was in the same place as the initial section;. Secondly, the young age at first presentation is against a vascular pathology, particularly where the vertebral artery is causative; and thirdly, there is a reported incidence of new vascular compression in re-exploration for recurrent trigeminal neuralgia.

In conclusion, the salient points to emerge from this brief report are that vagoglossopharyngeal neuralgia can recur after IXth and partial Xth nerve section and that this patient provides evidence for a pure vagal neuralgia. This supposition is supported by the finding that even the most caudal vagal rootlets may carry general somatic afferent fibres to the spinal trigeminal tract.1 In addition, the two separate episodes with differing pathologies raise the interesting question of whether there is a particular propensity for neuralgia which may or may not require a vascular trigger. This bears on the point raised by Adams et al as to why there are so many possibly causative vessels and so few neuralgias.6