Isolated dysphagia as initial sign of anti-IgLON5 syndrome

Investigations.

Fiber optic endoscopic evaluation of swallowing (FEES) revealed severe oropharyngeal dysphagia with poor oral bolus control, premature spillage leading to predeglutitive aspiration, a severely delayed swallowing reflex, and residues in the epiglottic valleculae and in the piriform sinus, with solid and half-solid consistencies tested, leading to postdeglutitive penetration and aspiration (figure, C). The FEES tensilon test showed no improvement of dysphagia.¹

Figure Investigations

(A) T1 gadolinium-enhanced axial MRI shows no evidence of underlying brainstem or cranial nerve pathology. (B) No intrathecal immunoglobulin synthesis was observed in CSF analysis using the Reiber scheme. (C) Fiber optic endoscopic evaluation of swallowing reveals postdeglutitive residue in the valleculae epiglotticae (3), on top of the epiglottis (2), and in the piriform sinuses (1).

Videofluoroscopic swallowing study showed poor oral bolus control and, as subsequently confirmed by high-resolution manometry, inadequate opening of the upper esophageal sphincter and ineffective esophageal peristalsis due to disharmonic and partially retrograde contractions. Cerebral MRI (figure, A) including contrast-enhanced high-resolution imaging of brainstem and cranial nerves was unremarkable. EEG was unremarkable. Nerve conduction studies had normal findings. EMG including repetitive nerve stimulation was normal without indication of a myasthenic reaction or myopathy.

Serum acetylcholine receptor antibodies, muscle-specific tyrosine kinase antibodies, and titin antibodies were negative (Euroimmun; Lübeck, Germany). Screening for antiganglioside antibodies was negative. CT examination of the chest and abdomen with contrast agent showed no evidence of malignancy. Moreover, gynecologic examination showed no evidence of an underlying tumor. ECG was normal. Screening for systemic autoimmune diseases and hematologic malignancy was negative. Except for a probably irrelevant decrease in the serum/CSF glucose quotient (0.4, reference value 0.6–0.9), routine CSF analysis was unremarkable. No intrathecal immunoglobulin (Ig) synthesis (figure, B) or isolated oligoclonal IgG bands were observed and accordingly no antibody-producing plasma cells were found in the CSF as revealed by flow cytometry. Autoantibody screening was positive for IgLON5 IgG (+++, 1:1,000) serum levels but no other known neural autoantibody. CSF was not analyzed for the presence of neural autoantibodies. However, as measured by the total IgG ratio between CSF and serum (0.003), relevant CSF titers of IgLON5 IgG can be expected even by passive diffusion.

Sixteen days after admission, right-sided ptosis and peripheral facial palsy occurred, followed by respiratory deterioration necessitating orotracheal intubation and subsequent tracheotomy. Subsequent FEES revealed bilateral vocal cord palsy with complete glottis closure.

Ambulatory polysomnography (PSG) (Somnoscreen; Somnomedics GmbH, Randersacker, Germany) without videography and audiography was conducted in the intensive care unit (ICU). It was applied according to American Academy of Sleep Medicine (AASM) standard criteria (AASM Scoring Manual 2.0)² including EEG (F4, C4, C3, O2, M1, and M2 electrodes), electro-oculography, EMG (mental, both anterior tibialis muscles), cardiorespiratory recording (single-channel ECG, thoracic and abdominal respiratory movements [piezo], transcutaneous oxygen saturation), and body position. Nasal airflow (thermistor) and nasal pressure cannula were not applied.

Sleep stages and associated events were scored according to suggested definition by Sabater et al.³ At the time of PSG, the patient was tracheotomized, but weaned from mechanical ventilation. For that reason, potential stridor or obstructive sleep apnea syndrome were not assessable.

Sleep efficacy was reduced to 55.1%. Undifferentiated non-REM (NREM) sleep with frequent movement artifacts occurred during the first and second third of the night associated with recurring arousals with and without association with partly highly frequent periodic limb movements (92.2/h). Periodic limb movements were also present in wakefulness (95.2/h). Normal slow wave sleep, predominantly occurring in the last third of the night, was associated with frequent spindles. It was regularly fragmented by leg movements. REM sleep was reduced (1.5%) and showed bursts of muscle activity in mentalis and tibial anterior muscles, compatible with REM sleep without atonia. Respiratory parameters were normal (apnea hypopnea index 0.9/h, oxygen desaturation index 0.2/h, mean oxygen saturation 98%, minimal oxygen saturation 91%, breathing frequency 17/min).

Due to the absence of video recordings, REM sleep behavior disorder or purposeful behaviors were not assessable but recurrent motoric artifacts especially during undifferentiated NREM might be suspicious for possible behavior during REM and NREM sleep.

Due to the detection of IgLON5 IgG, an immune-mediated neuropathy of the cranial nerves or brainstem pathology was suspected and a total of 5 cycles of plasmapheresis (treatment of 1.5 plasma volumes every other day) combined with IV methylprednisolone pulse therapy (1 g/d for 5 consecutive days) was initiated. This treatment led to mild improvement of vocal cord paresis and marked remission of ptosis and facial nerve palsy.

Hence, immunosuppressive therapy was continued with oral glucocorticoids (80 mg prednisolone). After the patient was weaned from the respirator, FEES revealed that although vocal cord paresis showed further improvement, severe pharyngeal dysphagia persisted. The patient was transferred to a rehabilitation unit with a tracheal cannula in place without any oral intake. Nutrition was given via percutaneous endoscopic gastrostomy tube.

At a follow-up visit 6 weeks later, dysphonia had substantially improved. Speaking was loud and clear when the cuff was deflated and the cannula was manually closed. FEES revealed that vocal cord paralysis had resolved and the patient was able to swallow her saliva without any problems. When given small amounts of liquids, the swallowing reflex was triggered; however, the whiteout was shortened and weak, corresponding to a reduced laryngeal elevation and weak pharyngeal contraction. Following our standard decannulation protocol, the tracheal cannula was removed.⁴ Swallowing training started with small amounts of pudding textures, while nutrition was supplied by tube feeding. Treatment with prednisolone was tapered below Cushing threshold. No relapses of ptosis, facial nerve palsy, or other focal neurologic deficit were observed. The patient died unexpectedly 5 weeks later. A few days earlier, the patient had again developed an aggravated stridor.