Cranial Polyneuropathy Secondary to Remote Iophendylate Myelography

ABSTRACT

A 68-year-old woman with controlled hypertension, and degenerative joint disease of the spine for which she had undergone several myelograms and three surgeries 30–32 years earlier, presented with a 2 year history of painless, oblique, binocular diplopia. Her prior ophthalmic evaluations were consistent with an isolated left trochlear nerve paresis. She had magnetic resonance imaging (MRI) showing multiple foci of T1-weighted hyperintensities around the midbrain and brainstem thought to represent subarachnoid fat from a ruptured dermoid cyst. An extensive evaluation revealed a left trochlear nerve paresis as well as diminished sensation in the distributions of the first and second divisions of the left trigeminal nerve. Review of her MRI and history of myelograms raised the possibility of focal inflammation from intrathecal iophendylate (Pantopaque®). Repeat MRI was obtained that showed T1-weighted hyperintensities similar to her previous MRI, but in this study, T1-weighted fat suppression imaging also was performed and revealed these foci to be of low signal intensity, consistent with retained iophendylate.

Introduction

Radiocontrast agents are invaluable substances that not only enhance the visibility of internal structures but also offer diagnostic direction. Their use in neuroradiology can be found as early as 1918 with the development of ventriculography and pneumoencephalography by Dr Walter Dandy, cerebral angiography in 1930 by Dr Egas Moniz, and myelography with iophendylate in 1944.^1,2 Today, contrast media are used most often with magnetic resonance imaging (MRI), computed tomography (CT), and catheter angiography, with well-known adverse reactions including nephrogenic systemic fibrosis, anaphylactoid reactions, and contrast nephropathy. Although rare, side effects may manifest years after use, especially with poorly absorbed contrast agents. I report a case of iophendylate-induced inflammation causing a cranial polyneuropathy, including a left trochlear nerve paresis and left trigeminal sensory neuropathy, about 30 years after myelography.

Case report

A 66-year-old woman presented to the Wilmer Emergency Room with painless, oblique, binocular diplopia. Her medical history included controlled primary hypertension, benign colonic polyps with recurrent colitis, chronic back pain, and degenerative joint disease of the spine for which she had undergone several myelograms and three surgeries 30–32 years earlier. She had no prior ocular history, had no history of smoking or illicit drug use, and consumed alcohol socially. She was on propranolol, losartan, alendronate, and verapamil, and had no known drug allergies.

Eleven months earlier she had fallen and hit the back of her head. Three months later, she developed intermittent binocular, oblique double vision and eventually was found to have a left hypertropia of 9 prism dioptres (PD), increasing to 14 PD on right gaze. The hypertropia worsened on head tilt towards the left shoulder and disappeared on head tilt towards the right shoulder. There was 7–10 degrees of left excyclotorsion. A diagnosis of traumatic left trochlear nerve paresis was made, and magnetic resonance imaging (MRI) of the brain and orbits was obtained. The MRI was interpreted by a radiologist as follows:

On the sagittal T1-weighted images multiple small foci of increased signal intensity are identified along the anterior aspect of the midbrain as well as the borders of the suprasellar cistern. These also are seen on the optic chiasm and pituitary stalk. These findings are more significant on the left side. A slightly larger focus is identified on the left side between the fifth nerve and the seventh/eighth nerve complex, close to the internal auditory canal. Some of these foci are seen accumulating in the interpeduncular cistern. None of the foci enhance after administration of contrast. Impression: Fat droplets consistent with a ruptured dermoid cyst.

Assessment by a neurologist revealed the patient to have a left trigeminal sensory neuropathy and possibly a peripheral vestibulopathy, in addition to the left trochlear nerve paresis. Accordingly, she underwent a more extensive assessment including an electroencephalogram, gallium scan, audiometry, and two lumbar punctures, none of which revealed any specific abnormality. She was treated with an 8-PD Fresnel prism placed base up over her right spectacle lens that fully corrected her diplopia.

The following year she began to experience episodes of difficulties with balance and was found to have mild bilateral vestibular dysfunction. She was treated with vestibular rehabilitation therapy with significant improvement. One year after this she began to experience episodes of slurred speech, falls without dizziness, and progressively worsening balance requiring a cane. These symptoms continued to worsen over the next month. She underwent repeat MRI that showed no change from the prior study and MR angiography showed no abnormalities. It was at this time that she was referred to the Neuro-Ophthalmology Division of the Wilmer Eye Institute where she was found to have normal visual acuity, colour vision, and visual fields, as well as isocoric pupils that reacted normally to light and near stimulation. Sensorimotor examination revealed an 8-PD left hypertropia, increasing slightly on right gaze and stable on left gaze, associated with worsening of the hypertropia on head tilt towards to the left shoulder and disappearance of the hypertropia on head tilt towards the right shoulder. Lancaster red-green testing demonstrated a small amount of excyclotorsion of the left eye. With her prism glasses she had no significant shift at distance or near, with normal stereopsis (9/9 circles) on the Titmus test. Corneal sensation was equal and normal bilaterally, but facial sensation was slightly diminished in the first and second distributions of the left trigeminal nerve. Facial movement was normal bilaterally; there was no ptosis or proptosis. Intraocular pressures were normal and slit-lamp exam revealed only mild cataracts. Dilated fundus exam was normal in both eyes except for mild left fundus extorsion.

Given the patient’s history of previous myelography, consideration was given to the possibility of focal intracranial inflammation caused by retained iophendylate. Review of her prior MRI showed multiple, small, subarachnoid lesions that were hyperintense on T1-weighted images and correspondingly hypointense on T2-weighted images and were located around the suprasellar cistern, optic nerves, left trigeminal nerve, tuber cinereum, and midbrain. Repeat MRI was obtained showing foci with T1- and T2-weighted signals similar to her previous MRI, but in this study T1-weighted fat suppression imaging also was requested, and this revealed these foci to be of high signal intensity (Figures 1–3), thus inconsistent with fat. It was therefore concluded that these lesions were most likely retained iophendylate from the patient’s previous myelography. Because of the inflammatory effects of this contrast agent, the patient was offered a course of steroid therapy, but she declined. She remains stable to date.

Figure 1.

T1-weighted sagittal magnetic resonance imaging without contrast shows hyperintense lesions within the lateral ventricle and at the skull base (arrows). Note that the clivus, which contains fat, also is hyperintense (asterisk).

Figure 2.

T2-weighted sagittal magnetic resonance imaging without contrast shows that the lesions are hypointense (arrows).

Figure 3.

T1-weighted fat-suppressed sagittal magnetic resonance imaging without contrast shows that the lesions remain hyperintense (arrows), but the clivus is no longer hyperintense (asterisk). Thus, the lesions are not droplets of fat.

Discussion

This patient initially presented with diplopia caused by a trochlear nerve palsy that became symptomatic several months following mild blunt trauma to the back of the head. The palsy was attributed to the trauma and, indeed, this may have been the case; however, her trigeminal sensory neuropathy as well as the subsequent development of progressive difficulties with balance associated with evidence of bilateral vestibular dysfunction would seem unlikely to have been caused by the injury and more likely were related to the retention of iophendylate.

Iophendylate (Pantopaque® in the USA; Myodil in the UK), was an oil-based contrast medium widely used for myelography and ventriculo-cisternography beginning in the 1940s and continuing until the era of CT scanning, when water-soluble iodinated contrast began to be used. It was popular for many years because it was considered safer than other agents at the time. The major problem with this agent, however, was its poor absorption, clearing at a rate of 0.5 to 3 ml per year.³ In general, great care was taken to aspirate completely the iophendylate after myelography, but this usually was not possible. Acute adverse events could occur as early as an hour or two after myelography and included fever, neck stiffness, cranial nerve palsies, optic neuritis, cortical blindness, dizziness, and myelopathy.^1,4–6 Chronic retention usually manifested as an adhesive arachnoiditis that often was asymptomatic; however, in 1–2% of individuals, it caused non-specific lumbosacral back pain, urinary retention, progressive myelopathy, hydrocephalus, brainstem dysfunction, and/or seizures, years after initial exposure.^7–10 For example, the patient reported by Pascuzzi et al.⁸ experienced seizures attributed to iophendylate 10–15 years after myelography, whereas the patient reported by Huang et al.⁹ developed iophendylate-induced arachnoiditis mimicking an intramedullary spinal cord tumour 30 years after myelography. In all cases, residual iophendylate was present in the subarachnoid space adjacent to the area of dysfunction.

Pathologically, the arachnoiditis caused by iophendylate is a fibrovascular proliferation of the leptomeninges with granuloma formation and infiltrates of lymphocytes, plasma cells, and foreign-body giant cells adjacent to encysted iophendylate.⁴ The subarachnoid space may be virtually obliterated by this process in severe cases. The pathogenesis is not known; however, proposed mechanisms include a hyperosmolar effect, a chemotoxic effect, or an immunological reaction.⁴

The MRI appearance of iophendylate is that of a signal intensity similar to fat on routine sequences; i.e., lesions that are hyperintense on T1-weighted images and hypointense on T2-weighted images. The radiographic differential diagnosis of these lesions thus includes metastatic disease such as melanoma, haemorrhage, vascular malformations, and ruptured dermoid cyst; however, in contrast to fat, iophendylate remains hyperintense on fat suppressed T1-weighted images.^11,12 Therefore, requesting appropriate imaging sequences, specifically T1-weighted fat suppression imaging, is key to the diagnosis of iophendylate toxicity, thus avoiding unnecessary testing.

Corticosteroids have been reported to improve symptoms in select cases of iophendylate toxicity, whether administered orally, intravenously, and/or intrathecally.¹³ Surgical options include spinal decompression, removal of encapsulated iophendylate, and adhesiolysis of arachnoid membranes.¹⁴ Close follow-up with serial imaging should be pursued to monitor for progression or recurrence of iophendylate toxicity.

Although iophendylate-based radiography has not been used for over 4 decades, its pathologic effects still may develop in older individuals who underwent myelography 40–50 years ago. Awareness of its side effects, appearance on MRI, and differentiating features from radiographically similar lesions can greatly aid in its diagnosis and avoid unnecessary testing. In addition, this case emphasises three important points. Firstly, there is no substitute for a complete medical history. Secondly, all contrast agents, regardless of the route of administration, may have potential short-term and long-term toxicity. Finally, in cases where neuroimaging is a key factor, the clinician and the radiologist must work together as a team.

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).