Neurosyphilis Manifesting as a Longitudinally Extensive Transverse Myelitis: A Case Report

Background

In the present era, the initial approach to the management of longitudinally extensive transverse myelitis centers primarily on immune-mediated etiologies, but infections remain an important consideration in the differential diagnosis. Historically, neurosyphilis has established itself as a well-known cause of infectious myelitis with an ability to infect the spinal cord at any stage of treponemal infection. The most common forms are tabes dorsalis, a late complication of years-long treponemal infection of the dorsolateral columns and dorsal root ganglia presenting with gait ataxia and lancinating pains, and syphilitic meningomyelitis, produced by robust inflammation of the meninges, spinal cord parenchyma, and associated arteries.^1,2 As described in this unusual case of ventral-predominant syphilitic myelitis, it may also more rarely present in a subacute fashion that closely mimics the autoimmune myelitides, a highly challenging clinical dilemma given the differences in the treatment of infections and autoimmune disorders and the risk of future disability without the rapid administration of an effective treatment.

Case

A 29-year-old male developed ascending bilateral lower extremity numbness and paresthesias to the umbilicus, involving his genitalia and rectum, as well as urinary retention and constipation over a period of 1 week. Five days from onset of sensory symptoms, he developed lower extremity weakness progressing to an inability to ambulate over a 2-day period prior to presentation. Medical history was unremarkable. He was not on medications at the time of presentation. He was monogamous with 1 sexual partner for several years. Extended review of symptoms was otherwise negative. Vital signs were remarkable only for tachycardia to 109 beats per minute. The neurologic exam revealed normal mental status and cranial nerves. Bulk and tone were normal throughout. All muscle groups of the upper extremities were 5/5 in strength bilaterally. In the lower extremities, the right leg muscles were diffusely 2/5 in strength, and the proximal left leg muscles were 2/5 proximally and 3/5 distally. He had decreased sensation to pinprick and temperature in a patchy distribution from the legs to the umbilicus, with preserved vibration and proprioception. Biceps, triceps, brachioradialis, patellar and Achilles reflexes were 3+ throughout. Hoffman’s reflexes were absent, but pectoralis reflexes and crossed adductor responses were present. Plantar reflexes were upgoing bilaterally. He had no dysmetria or ataxia on finger to nose testing. He was unable to ambulate due to weakness.

Lower extremity weakness, sensory deficits to pinprick and temperature from the level of the umbilicus inferiorly, and lower extremity hyperreflexia localize to the bilateral corticospinal and spinothalamic tracts within the anterior thoracic spinal cord, with a sensory level suggesting localization to the T10 level. The presence of hyperreflexia combined with extensor plantar responses argued against localization to peripheral nerves or cauda equina. Isolated involvement of the conus medullaris was considered unlikely given the sensory level in the thoracic region and impairment in lower extremity strength. Multifocal brain lesions were considered unlikely in the setting of relatively symmetric bilateral lower extremity sensorimotor deficits without localizing findings exclusively referrable to the brain on exam. A 1.5T MRI of the total spine performed with and without contrast showed a T2 hyperintensity along the ventral cord at the level of C6-C7, and a long-segment T2 FLAIR/STIR hyperintensity in the central portion of T1-T11 with associated heterogeneous enhancement and cord expansion. No leptomeningeal or dural enhancement or thickening was noted. MRI of the thoracic spine with diffusion-weighted imaging did not show restricted diffusion. Brain MRI with and without contrast was unremarkable. CSF examination showed an elevated white blood cell count (311, 85% lymphocytes, normal ≤4), elevated total protein (134 mg/dL, normal range 20–40 mg/dL), and a glucose of 46 mg/dL. CSF bacterial cultures and viral PCRs (HSV, VZV, CMV, EBV and enterovirus) were negative. Two unique oligoclonal bands were present in the CSF, and IgG index was elevated at 1.09 (normal <.7). Blood cell count, liver enzymes, serum creatinine, and thyroid function studies resulted within normal range. ESR and CRP were elevated at 64 mm/h and 12.4 mg/dL, respectively. HIV antibodies, hepatitis B/C panel, and quantiferon gold were negative. SSA/SSB antibodies, rheumatoid factor, ANA, ANCA, and ACE levels were all within normal limits. Aquaporin-4 antibodies and MOG antibodies were negative.

Clinical presentation demonstrated a subacute longitudinally extensive transverse myelitis (LETM) with no brain or optic nerve lesions occurring in the setting of a lymphocytic pleocytosis and intrathecal production of antibodies. Inflammatory and infectious etiologies were high on the differential diagnosis. The most likely inflammatory causes included neuromyelitis optica spectrum disorder and MOG antibody disease given the longitudinally extensive lesions in the spinal cord. Multiple sclerosis was less likely given the longitudinally extensive nature of the thoracic spinal cord lesion, the absence of brain lesions, and the robust inflammatory CSF response. Infections were initially deemed less likely in the absence of any systemic symptoms and signs. Anterior spinal cord infarction was thought less likely in the absence of vascular risk factors, the progressive evolution of symptoms, and negative diffusion-weighted imaging. High dose intravenous steroids were started empirically given concern for autoimmune myelitis. On day 3, a serum RPR returned positive with a high titer of 1:64, and CSF VDRL was reactive with a titer of 1:8, confirming a diagnosis of neurosyphilis. The steroids were stopped and treatment for neurosyphilis was started with a 14-day course of intravenous penicillin G 12 million units every 12 hours.

Given the rarity of reports of LETM attributable to neurosyphilis, it remained unclear whether this clinical syndrome was caused by neurosyphilis or if an autoimmune disorder was superimposed. The decision was made to treat as neurosyphilis with a possible concomitant immune-mediated myelitis with 5 sessions of plasma exchange in addition to standard treatment for neurosyphilis. After finishing treatment, the patient was discharged to rehabilitation. At 6 week follow-up, he improved and was able to independently ambulate albeit with a spastic gait. Exam showed lower extremity hypertonia and spasticity, hip flexor strength 4/5 on the right and 5-/5 on the left, knee extensors 5/5 bilaterally, knee flexors and dorsiflexors 5-/5 on the right, 5/5 on the left, and plantar flexors 5/5 bilaterally. He had resolution of urge incontinence. Aquaporin-4 and MOG antibodies were repeated and remained negative on follow up 8 weeks after admission. Given the monophasic presentation and absence of aquaporin 4 and MOG antibodies on 2 separate assessments, he was not started on immunosuppressive therapy. Repeat RPR was 1:16 at two months’ follow up, which constituted a clinically meaningful treatment response. He followed up again at 7 months with sustained improvement in leg strength, gait function, and bladder control.

Discussion

Our case involved LETM in the setting of a new neurosyphilis diagnosis. An extensive workup of other infectious etiologies was unremarkable. Aquaporin-4 and MOG antibodies were negative at initial presentation as well as at 2-month follow up, as it is recognized that these antibodies may initially result as negative with subsequent positive results in the setting of NMOSD (10%-20%) or MOG antibody disease.^2-4 The presence of neurosyphilis in this patient presented diagnostic ambiguity regarding whether it may have been the direct etiology for the myelitis, whether neurosyphilis may have provoked an immune-mediated response, or whether neurosyphilis was incidentally discovered. This patient had a pattern of anterior spinal cord involvement evolving subacutely with early loss of strength and pain/temperature sensation and preservation of proprioception and vibration sense, inconsistent with the chronic course and involvement of the dorsal roots and columns more commonly seen in tabes dorsalis. Meningovascular syphilis with occlusion of the anterior spinal artery was considered. However, diffusion-weighted imaging of the thoracic spinal cord did not show definite restricted diffusion.^1,5,6 CTA of the chest and abdomen did not show signs concerning for aortitis or vasculitis. Finally, imaging did not show evidence of pachymeningitis or gumma formation resulting in compressive myelopathy (Figures 1 and 2).

OPEN IN VIEWER

Figure 1.

Panels A-D. A-B: Sagittal T1-weighted post-contrast images of the thoracic spine demonstrating heterogenous enhancement from T1-T11. C-D: Sagittal T2-weighted images of the thoracic spine demonstrating a confluent cord hyperintensity along the thoracic spinal cord from T1-T11.

OPEN IN VIEWER

Figure 2.

Panels A-F. A-C: Axial T2-weighted images of the cervical/thoracic spine demonstrating hyperintense signal at the ventral aspects of C6, C7, and T1. D-F: Axial T1-weighted post-contrast images of the cervical/thoracic spine demonstrating patchy enhancement at C6, C7 and T1.

The patient was diagnosed with syphilitic myelitis, which is rarely described but has been associated with longitudinally extensive lesions.^5-10 Most cases of LETM were reported prior to the discovery of the various antibodies associated with LETM, leading to diagnostic uncertainty, but repeated negative aquaporin-4 and MOG antibodies in this case argues against either of these as the cause for this presentation. Some of the few reported cases of syphilitic myelitis have demonstrated radiographic evidence of meningeal involvement, leading to a “candle guttering appearance” of subpial enhancement on MRI. ⁷ Our patient demonstrated a pattern of enhancement affecting the central and anterior spinal cord parenchyma with no clear radiographic evidence of meningeal involvement. In the era of growing emphasis on antibody-mediated causes of myelitis, this clinical case of syphilitic meningomyelitis serves as a reminder that infections must be included in the differential diagnosis for LETM, as immunomodulatory treatments may worsen outcomes in infectious myelopathies if the underlying infection has not been treated. Prior STI testing in this patient had not included screening for syphilis, and thus the duration of infection was unknown. We recommend that patients presenting with LETM should be screened for syphilis, which represents a treatable cause of infectious myelitis. Our case emphasizes the importance of maintaining a broad differential diagnosis early in the case of inflammatory myelopathies.