Clinical Reasoning: A 63-Year-Old Man With Progressive Multicranial Neuropathy and Leptomeningeal Enhancement

Abstract

A 63-year-old man, with a history of melanoma and basal cell carcinoma, presented with progressive right-sided facial numbness, vertical diplopia, and headache. Brain MRI revealed leptomeningeal enhancement of multiple cranial nerves and an enhancing mass-like lesion along the anterolateral surface of the pons and midbrain. Subsequent brain biopsy demonstrated the final diagnosis. This case highlights the broad differential diagnosis of leptomeningeal disease, emphasizing the role of specific clinical, laboratory, and imaging cues in guiding clinical reasoning.

Section 1

A 63-year-old man with diabetes, prior melanoma, and prior basal cell carcinoma presented with an 8-week history of progressive right-sided facial numbness, vertical diplopia, and headache. The numbness initially involved the right periorbital region and gradually extended to the entire right face, followed by vertical diplopia and a mild holocranial headache. His back and chest melanoma and his left eyebrow basal cell carcinoma were excised 2 and 3 years earlier, respectively. He had not received chemotherapy and was under regular dermatologic follow-up. The patient had no history of occupational exposures or familial tumor predisposition syndrome.

Neurologic examination revealed anisocoria with right eye mydriasis, mild right ptosis, and limited up-gaze and abduction in right eye. Decreased facial sensation involved all right-sided trigeminal distributions. There was no pain during eye movements. The funduscopic exam showed no papilledema. The remainder of the exam was normal.

Questions for Consideration:

1. Where does the patient's clinical examination localize?

2. How does the patient's oncologic history influence the initial assessment?

3. What is the differential diagnosis?

4. What diagnostic workup is recommended?

GO TO SECTION 2

Section 2

The clinical presentation suggests multiple right-sided cranial nerve (CN) deficits, including partial oculomotor (CNIII), trigeminal (CNV), and abducens (CNVI) palsies. The absence of upper motor neuron signs or limb sensory changes suggests likely sparing of the brainstem parenchyma but does not rule out noneloquent CNS involvement. The unilateral deficits suggest a focal disease process. The normal funduscopic examination argues against increased intracranial pressure (ICP). Headache likely indicates meningeal irritation. Leptomeningeal or pachymeningeal disease in the posterior fossa is the most likely localization.

Possible etiologies include neoplastic, vascular, infectious, and inflammatory. The patient's history of melanoma raises concern for leptomeningeal carcinomatosis through hematogenous or perineural spread. The latter seems unlikely given the distant location of his prior chest and back melanoma. Basal cell carcinoma could spread perineurally, but its indolent nature and lack of recent recurrence make this unlikely. Leptomeningeal dissemination (LMD) from undiagnosed malignancies, systemic or CNS, needs consideration. Vascular etiologies involving the right cavernous sinus, such as cavernous sinus thrombosis, carotid-cavernous fistula, and Tolosa-Hunt syndrome, are less likely because of the absence of typical ocular findings (pulsatile exophthalmos, orbital bruit, chemosis, and orbital pain) and the subacute temporal profile. Chronic meningitis, specifically tuberculous and fungal meningitis, should be considered given possible immunocompromise from diabetes, although the patient had not received chemotherapy for his prior melanoma and basal cell carcinoma. Inflammatory etiologies, including neurosarcoidosis, other granulomatous disease, immunoglobulin G4–related disease, or vasculitis, are relevant differential diagnoses.

The diagnostic workup with brain and skull base MRI showed leptomeningeal enhancement of multiple cranial nerves and an enhancing mass-like lesion along the brainstem surface (Figure 1, A–H). No cavernous sinus thrombosis was seen. Extensive laboratory testing was unremarkable, as shown in eTable 1. Notably, serology for syphilis, Lyme disease, HIV, and T-spot test for tuberculosis were negative. The rheumatologic workup was also negative. Three CSF studies, performed to enhance cytopathology yield, showed mildly elevated protein, but were otherwise unremarkable, including negative cytopathology, flow cytometry, and microbiological studies. Dermatologic evaluation showed no cancer recurrence. Formal ophthalmologic examination ruled out signs of intraocular lymphoma (anterior chamber cells, keratic precipitates, vitreous or subretinal lesions). CT scan of the chest, abdomen, and pelvis and a whole-body ¹⁸F-FDG-PET scan showed non-FDG–avid lung nodules, too small for biopsy. Brain ¹⁸F-FDG-PET showed increased uptake in the enhancing mass-like lesion (Figure 1, I and J). A complete spine MRI with and without contrast was normal. Targeted or untargeted leptomeningeal biopsies were deemed high risk with low yield and were deferred.

Figure 1. Imaging.

Postgadolinium T1-weighted brain MRI reveals thickened patchy and nodular contrast enhancement most pronounced along the right anterolateral pons (arrow) (A), extending into the interpeduncular cistern (arrow) (B), without diffusion restriction on diffusion-weighted imaging (DWI) (C), and apparent diffusion coefficient map (D). DWI signal is isointense to the gray matter. Postcontrast thin sections on skull base MRI demonstrate (arrows) leptomeningeal enhancement in the right internal auditory canal along the right cranial nerve (CN) VII and VIII complex (E), bilateral CN V (F), and bilateral CN III (G). The right CN V is inseparable from the mass-like lesion at the root entry zone, as seen on constructive interference in steady state (CISS) sequence (H). Brain ¹⁸F-FDG-PET shows (arrows) increased uptake along the right anterolateral surface of the pons and the interpeduncular cistern (I, J). Maximal standard uptake value (SUVmax) was 7.9 (normal <2.5). Follow-up skull base MRI demonstrates disease extension (arrows) with new leptomeningeal enhancement along the right cerebellar folia (K), which represented the biopsy target, and increased enhancement of the right CN VII and VIII complex (L).

The diagnostic workup thus far has excluded systemic malignancy, infection, vascular pathology, and inflammatory etiologies other than neurosarcoidosis. LMD from CNS malignancy remains possible, given the lack of biopsy.

Questions for Consideration:

1. What treatments should be considered now?

2. What is the next step for the diagnostic workup?

GO TO SECTION 3

Section 3

Steroid therapy could be considered for possible neurosarcoidosis, which can be isolated in 5%–10% of cases. Sarcoid granulomas can mimic mass-like lesions on imaging, and a favorable response to steroids would support this diagnosis. The negative infectious workup indicates safety of steroids. A limitation is that CNS lymphoma might transiently respond to steroids, interfering with biopsy and delaying diagnosis. However, the absence of restricted diffusion on MRI argues against lymphoma.

An empiric 5-day course of intravenous methylprednisolone was initiated, followed by a slow taper of 1 mg/kg oral prednisone. Close clinical and radiologic follow-up was arranged.

Over a 6-week follow-up, the patient's clinical status continued to deteriorate. He developed right-sided upper and lower facial weakness and right hypoacusis, concerning for involvement of the right CNVII and CNVIII. New right-sided limb dysmetria suggested ipsilateral cerebellar hemisphere involvement. A follow-up brain MRI confirmed disease extension (Figure 1, K and L). The patient underwent a biopsy through right suboccipital craniectomy, with histologic findings displayed in Figure 2.

Figure 2. Histopathology.

Hematoxylin and eosin stain show cellular glial neoplasm (A), with frequent (arrows) mitotic activity (B, C). Tumor cells are immunoreactive for GFAP (D) and OLIG2 (E), indicative of relatively preserved glial differentiation, while Ki-67 proliferation index is moderate (F). Tumor cells are negative for IDH1-R132H (G) and H3K27M protein (H). SOX-10 (I) labels scattered tumor cells. Panels A, D–I are ×10 magnification; panels B and C are ×100 magnification.

Question for Consideration:

1. What is the final diagnosis based on pathology?

Discussion

The patient was diagnosed with glioblastoma (GBM), isocitrate dehydrogenase (IDH) 1/2-wildtype, presenting with LMD. A next-generation sequencing panel revealed pathogenic variants in PTEN, TERT promoter, and NF1. Whole-brain radiation was implemented, but the patient unfortunately died 4.4 months after symptom onset.

LMD is an underdiagnosed GBM manifestation, due to low diagnostic sensitivity and occurrence of asymptomatic cases.^1-3 Postmortem studies indicate a prevalence of up to 25%.^2,3 LMD at initial GBM diagnosis, particularly without clear clinical or imaging parenchymal involvement, is very uncommon.⁴ This presented a unique challenge in our case.

LMD pathogenesis involves cell migration from the initial tumor site along brain vessels to subpial, subarachnoid, and subependymal spaces. Primary LMD indicates a pattern of early CSF seeding, with limited tumor expansion, as in this case. Secondary LMD indicates delayed dissemination, typically during GBM treatment or recurrence.²

The clinical presentation of LMD is heterogeneous, typically subacute and progressive, including cranial nerve deficits, increased ICP, hydrocephalus, meningism, focal deficits, or diffuse encephalopathy.^2,5,6 The latter is the most common presenting symptom, particularly in the elderly.^2,7 Cranial nerve deficits involving CNII, III, IV, V, VI, and/or VII occur in 6% of cases, with CNIV and VII most frequently involved.^2,5,6 Hydrocephalus or ICP elevation occur in 25% of cases.^2,5,6 Drop metastases to the spinal cord are rare but warrant spine MRI when LMD is suspected.⁶

Tumor sites, specifically infratentorial, periventricular, pineal, and spinal, represent critical risk factors of symptomatic LMD. This reflects proximity to CSF spaces or probability of mass effect.^2,3 In this case, the lesion's location on the surface of the pons and midbrain suggested a possible brainstem origin. Despite the posterior fossa location, the mass-like lesion did not compress the fourth ventricle.

LMD may be favored by specific histologic GBM features, including astrocytic phenotype, as in the current case.^2,3 High proliferation rate and poor glial differentiation are possible additional features, which were not confirmed here (Figure 2).^2,3 Several genetic and epigenetic alterations in the tumoral cells have been reported, though with unclear specificity or sensitivity.^2,3 Genetic profiling in this case confirmed PTEN mutation as a risk factor of LMD. The identified mutations in the TERT promoter and NF1 are associated with a worse prognosis in IDH1/2-wildtype GBM, independent of LMD.^2,3

Brain MRI with contrast has a sensitivity >90% for diagnosing symptomatic LMD.⁸ Typical findings include subarachnoid, ventricular, or pial enhancement, either nodular or diffuse.^8,9 Cranial nerve enhancement is seen in 11%–19%, more frequently at the root zone and in brainstem GBM, reflecting contiguous tumor spread.^8,9 A markedly decreased apparent diffusion coefficient (ADC), indicative of hypercellularity, is supportive of CNS lymphoma, while ADC values are variable in GBM.¹⁰ Diffusion-weighted imaging (DWI) signal isointense to the gray matter, as in this case, may suggest increased cellularity, supporting neoplastic etiology.

The elevated FDG-PET signal we identified indicates a hypercellular lesion, but its ability to differentiate neoplastic from inflammatory conditions is unclear because of variable cutoffs and low sensitivity.¹¹

Repeated lumbar puncture (LP) for cytopathology is needed given low sensitivity, but the yield remains low—up to 45% after 1 LP, 86% after 3, and 93% after 4 or more.^2,3,6 Lack of pleocytosis with high protein or lactate, or mild pleocytosis with predominant macrophages,¹² are typical CSF findings, but not sensitive or specific. Liquid biopsies, involving the analysis of tumor components in the CSF (cells, DNA, RNA, and exosomes), may increase the yield of CSF studies.^2,13 Limited access, high costs, and lack of validation are barriers to their clinical use.

Survival in LMD from GBM is significantly lower than in nondisseminated GBM, with a median similar to our case of 4.7 months.^2,3 Treatment options are limited, involving a multimodal approach with whole-brain radiation therapy, chemotherapy, and surgery.¹⁴ Notably, the extent of LMD on imaging does not correlate with survival.³

In conclusion, our case highlights the diagnostic challenges of GBM presenting with leptomeningeal disease, particularly when a parenchymal source is not clear. The differential diagnosis is especially broad in oncology patients. Key clinical and laboratory findings—such as steroid unresponsiveness, rapid disease progression, and elevated CSF protein without pleocytosis—can suggest GBM. Distinctive MRI features include mass-like enhancement along the brainstem surface, lack of marked diffusion restriction, DWI isointense to the cortex, and cranial nerve enhancement at the root zone. A multidisciplinary approach is crucial for accurate diagnosis, including evaluation for systemic disease and need for biopsy. The limited sensitivity of CSF cytology, even after repeated tests, and the challenges posed by posterior fossa surgery further complicate LMD diagnosis. Advances in imaging, liquid biopsies, and genetic analysis may improve diagnostic and prognostic accuracy. Limited treatment options for LMD, leading to poor outcomes, remain a critical area for future research.

Appendix. Authors

Name	Location	Contribution
Arens Taga, MD	Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, MD	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data
Carlos G. Romo, MD	Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, MD	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data
Sharika Rajan, MD	Division of Neuropathology, Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, MD	Major role in the acquisition of data
Doris D.M. Lin, MD, PhD	Division of Neuroradiology, Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, MD	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data
Grace Tobin, MD	Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, MD	Analysis or interpretation of data
Arun Venkatesan, MD, PhD	Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, MD	Drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data
Carlos A. Pardo, MD	Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, MD	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data

Study Funding

No targeted funding reported.

Disclosure

A. Taga is supported by the NIH (National Institute of Neurological Disorders and Stroke) through an administrative supplement to award R25NS065729. C.A. Pardo is supported by NIH R01 NS123712. The content is solely the author's responsibility and does not necessarily represent the views of the NIH. Go to Neurology.org/N for full disclosures.