Objective:

To identify patterns in gadolinium enhancement on MRI in patients with biopsy-confirmed leptomeningeal disease. 

Background:

The leptomeninges, comprised of pia and arachnoid mater, are the innermost layer covering the brain and spinal cord. Leptomeningeal enhancement is commonly encountered in clinical practice and often presents a diagnostic challenge; the differential diagnosis is broad, and data can be nonspecific. With minimal literature on leptomeningeal enhancement patterns, we aimed to determine if there are consistent clinical, imaging, or laboratory clues suggesting a particular diagnosis. 

Design/Methods:

We identified adult patients with leptomeningeal gadolinium contrast enhancement on MRI who underwent meningeal biopsy for diagnostic purposes between January 2010 and September 2022. Pathology reports in the setting of stereotactic electroencephalography electrode placement and resection of known brain tumor or other lesions were excluded. Patient demographics and clinical/diagnostic data were recorded, including MRI findings, cerebrospinal fluid analysis, presenting symptoms, neurologic examination, and biopsy report with final diagnosis. 

Results:

As expected, a wide range of diagnoses were represented in 67 patients who met inclusion criteria, including inflammatory conditions such as neurosarcoidosis, cancers including lymphoma, infections, and amyloid-related conditions. A subset of 11 patients had non-specific inflammation on pathology and without definitive diagnosis despite extensive additional testing. Patterns of leptomeningeal enhancement emerged when comparing subgroups. These included a speckled or nodular pattern without local edema in sarcoidosis, a brainstem outline pattern associated with infectious entities, and asymmetric, cortical, posterior-predominant enhancement in amyloid cases. All lymphoma cases had concurrent parenchymal enhancement and were not purely leptomeningeal. The vasculitis cases typically had teardrop shaped enhancement. Interestingly, the patterns were specific and did not meaningfully overlap between the entities. 

Conclusions:

Patterns in leptomeningeal enhancement may be helpful in recognizing commonly cryptogenic CNS disease. Knowledge of these patterns could aid in selecting the optimal order of testing and opting for empiric treatments in nondiagnostic workups.