Прогрессивный миелит в 63-летней женщине: доклад о рассмотрении дела на Конференции Национального общества с множественным склерозом.

Case Presentation

A 63-year-old woman with triple-negative (estrogen, progesterone, and human epidermal growth factor receptor 2) stage IIA left invasive ductal breast adenocarcinoma with BRCA2 (PALB2) variation presented to another institution with 3–4 weeks of bilateral ascending paresthesias to her waist, paraparesis (documented as 3–4/5 on the Medical Research Council [MRC] muscle strength grading scale in knee flexion, ankle dorsiflexion, and ankle eversion), and bowel/bladder incontinence. Her neurologic symptoms started 3 months after breast cancer diagnosis and approximately 2 weeks after her third cycle of pembrolizumab (immune checkpoint inhibitor [ICI] through programmed cell death 1 [PD-1] blockade), paclitaxel, and carboplatin. MRI demonstrated multiple, short-segment T2-hyperintense lesions from the cervicomedullary junction to the conus without corresponding restriction diffusion and with peripheral gadolinium enhancement ( Figure 1 ). MRI of the brain showed mild burden of small, scattered T2 hyperintensities in the supratentorial white matter attributed to microvascular disease. She was started on oral dexamethasone 4 mg 3 times daily.

MRI of the Patient With MOG-IgG + and Myelitis After Immune Checkpoint Inhibitor

(A) Sagittal T2 image of C1 hyperintensity with additional focus of T2 hyperintensity in the upper thoracic cord. (B) Axial T2 image of upper thoracic cord T2 hyperintensity. (C) Sagittal T2 image of focused conus hyperintensity with (D) Sagittal image of T1 postcontrast showing associated enhancement.

Three weeks later, while on oral dexamethasone, she presented to our institution with worsening symptoms and inability to ambulate. Examination demonstrated 1/5 hip flexion/extension, 2/5 knee flexion/extension, and 5/5 distally on the MRC strength grading scale on the left and 4/5 hip flexion on the right; hyperreflexic patellar reflexes with crossed adduction; bilateral Babinski signs; and a T10 sensory level to pinprick. The Expanded Disability Status Scale (EDSS) score was 8.0 at admission. Repeat MRI showed interval evolution of the spinal cord hyperintense lesions with decreasing size and resolution of enhancement but addition of new foci of T2 hyperintensity and enhancement throughout the cervical and thoracic cord ( Figure 2 ). Normal/negative bloodwork included complete blood count, complete metabolic panel, HIV serology, varicella zoster IgG, vitamin B12, and folate. CT of the chest, abdomen, and pelvis was negative for metastases or hilar lymphadenopathy. Breast ultrasound showed resolution of breast mass. CSF demonstrated 14 nucleated cells/µL (normal 0–5; 95% lymphocytes), glucose of 98 mg/dL (normal 40–70; corresponding blood glucose 226 mg/dL; CSF/blood glucose ratio of 0.43), elevated protein of 85 mg/dL (normal 15–45), elevated IgG index of 0.71 (normal 0.00–0.61), absent oligoclonal bands, negative cytology, negative CSF bacterial and fungal cultures, negative cryptococcus antigen, and negative herpes simplex virus PCR.

Progression of Further Spinal Cord Lesions 3 Weeks After Initial MRI Imaging

(A) Sagittal T1 postcontrast image of the cervical spinal cord with multiple areas of enhancement. (B) Axial T1 postcontrast image of the cervical spinal cord showing cervical peripheral cord enhancement. (C) Sagittal T1 postcontrast image of the thoracic spinal cord with faint enhancement. (D) Axial T1 postcontrast image of the thoracic spinal cord with peripheral cord enhancement.

Differential Diagnostic Considerations

The differential diagnosis of a subacute myelitis includes infectious, neoplastic, metabolic/nutritional, and immune-mediated/neuroinflammatory conditions. A vascular etiology was considered unlikely in this clinical scenario based on subacute time course of symptoms. Clinical, radiographic, serum, and CSF studies helped to narrow the differential diagnosis.

Clinical Course and Demographics

The explanation for the patient's subacute onset of symptoms over days to weeks could represent demyelinating pathologies such as multiple sclerosis (MS), but her older age, absence of previous typical demyelinating events, lack of brain lesions, and negative oligoclonal bands argue against a diagnosis of MS. She had no systemic signs of infection, and CSF studies were negative for multiple infectious etiologies. Normal CSF/serum glucose ratio that is between 0.4 and 0.8 (rather than low), lack of systemic manifestations, and absence of mediastinal lymphadenopathy argue against a neuroinflammatory disorder such as sarcoidosis. 1 , 2 Her older age and cancer history raise the diagnostic concern for direct metastatic disease from breast cancer, a paraneoplastic neurologic syndrome, or myelitis related to an immune-related adverse event (irAE) from ICI use. A second unrelated malignancy, such as primary CNS lymphoma, was considered less likely. Negative CSF cytology and absence of residual tumor on breast ultrasound reduced the likelihood of metastatic involvement of the spinal cord.

MRI Features

The MRI features were typical of an inflammatory demyelinating myelitis with multiple short-segment, contrast-enhancing lesions. The radiographic imaging was atypical for a paraneoplastic condition, given lack of symmetric, dorsal-lateral tract-specific involvement, nor was it characteristic of myelin oligodendrocyte glycoprotein (MOG)-IgG myelitis, given the lack of a longitudinally extensive lesion and absence of involvement restricted to the gray matter. 3 , 4 However, the T2 hyperintensity and focal contrast enhancement of the conus could be consistent with myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) spinal cord pathology. 5 , 6 The lack of meningeal or leptomeningeal involvement argues against neurosarcoidosis and carcinomatous meningitis. Furthermore, the absence of diffusion restriction made CNS lymphoma less likely. 7

Final Diagnostic Conclusions

Serum and CSF autoimmune/paraneoplastic encephalopathy antibody panels were negative. Serum aquaporin-4 antibody tested by cell-based assay (CBA) was negative. Serum MOG-IgG tested by CBA was positive with a titer of 1:100 (normal, <1:20). The response to immunosuppressive treatments, lack of other features of metastatic disease, and temporal association of myelitis after the third cycle of ICI suggest that this was a myelitis related to pembrolizumab (irAE).

Treatment and Outcome

The patient was treated with IV methylprednisolone 1,000 mg daily for 5 days and 6 sessions of plasmapheresis, followed by a six-month course of oral prednisone taper starting at 60 mg daily. After the diagnosis of a CNS irAE and recent radiographic testing, including CT and breast ultrasound that showed complete radiographic response of tumor, further pembrolizumab infusions and chemotherapy were held. At time of discharge, the patient was nonambulatory and required a wheelchair for assistance. Five months later at follow-up, her lower limb strength had improved and she was ambulating with assistance of a walker (EDSS score of 6.5). Repeat MRI testing 6 months after symptom onset showed improvement in T2/STIR signal changes throughout the cord without enhancement. Repeat serum MOG-IgG at 6 months showed a titer of <1:20 without any further immune-directed therapy. She underwent left partial mastectomy with no residual cancer in the breast and 0 of 3 sentinel nodes involved on biopsy 6 months after hospitalization, followed by adjuvant radiation therapy at 9 months. At 1 year after admission, serum MOG-IgG remained negative (titer <1:20). MRI showed stable T2/STIR hyperintensities at the cervicomedullary junction and cervical cord without enhancement. Neurologic examination 2 years from symptom onset showed 5/5 strength in the legs with exception of 4/5 bilateral knee flexion. She could walk a short distance unassisted but required a walker for longer distances (EDSS score of 5.5). Bilateral salpingo-oophorectomy is planned in setting of germline PALB2 variation.

Discussion

In this case, the time interval between the treatment of breast cancer with ICI and the development of neurologic symptoms raises the question of whether her myelitis occurred because of augmentation of immune system activity in the setting of ICI (irAE) paraneoplastic immunity. ICIs target inhibitory regulatory steps in T-cell activation to enhance antitumor response and, by doing so, may promote activation of CD4 + and CD8 + T cells, which may then target the nervous system. 8 Pembrolizumab works by blocking the PD-1 receptor on T cells, thus enhancing T-cell activation and proliferation. Favoring the diagnosis of irAE is the temporal association (less than 2 months from ICI start) and the improvement with steroids and cessation of ICI. Neurologic complications occur in approximately 2%–4% of patients using ICI, with peripheral nervous system irAE involvement being 3 times more likely to occur than a CNS irAE. 9 , 10 Myelitis is uncommon and limited to case reports/series.

The possibility of an underlying paraneoplastic neurologic disorder that was preexisting cannot be ruled out because she presented with an intermediate risk phenotype of myelitis per most updated diagnostic criteria for paraneoplastic neurologic syndromes and concurrent cancer diagnosis. However, MOG-IgG is considered a low-risk antibody for paraneoplastic neurologic syndromes with weak association to malignancy. 11 The association of MOG-IgG with paraneoplastic syndromes is rare and limited to a few case reports. 12 - 14 Of these reports, 2 were in the setting of breast cancer and one presenting with myelitis. 15 , 16 Our patient did not have classical paraneoplastic antibodies associated with myelopathy (CRMP5/CV2, amphiphysin, ANNA1/Hu). Our patient did not have symmetric tract-specific T2 signal typical of paraneoplastic myelopathies. 17 MOG immunostaining was not performed as part of the histopathologic testing of this patient's breast tissue because it is not a standard evaluation. Ectopic MOG expression has been found in an ovarian teratoma in a patient with a suspected paraneoplastic neurologic disorder. 12

Although atypical, this patient could meet the proposed 2023 international diagnostic criteria for MOGAD. 6 The MOG-IgG CBA titer at 1:100 has a specificity of 97.8% and a positive predictive value of 82%. 18 In addition to the medium-high titer serum MOG-IgG, radiographic findings of myelitis with conus medullaris involvement and absence of oligoclonal bands supported the patient meeting diagnostic criteria for MOGAD. 6 However, in the setting of ICI, we consider that this patient should be classified and treated as an irAE myelitis, rather than MOGAD. MOG-IgG returned negative on subsequent testing, and no further inflammatory events have occurred with 2-year follow-up, thus far suggesting that this was a monophasic event. The mechanisms by which the tumor, the immune system, and ICI interplay to cause CNS pathology and the generation of MOG-IgG is unknown.

From a treatment standpoint, the patient improved clinically and radiographically after treatment with corticosteroids, plasma exchange, and discontinuation of pembrolizumab. Optimal treatment of CNS irAE is not yet known. Current American Society of Clinical Oncology guidelines for grade 3 (severe or medically significant symptoms but not immediately life threatening) demyelinating neurologic irAE recommend high-dose steroids with consideration of IV immunoglobulin or plasmapheresis. 19 Case reports have further described potential utility of tocilizumab. 20

This case highlights several clinical points including the emerging spectrum of CNS irAE including myelitis, outside of the typical radiologic and serologic descriptions of paraneoplastic myelopathy.

Acknowledgment

The authors thank the patient for granting permission to publish this information.

Glossary

cell-based assay

Expanded Disability Status Scale

immune checkpoint inhibitor

immune-related adverse event

myelin oligodendrocyte glycoprotein

myelin oligodendrocyte glycoprotein antibody-associated disease

Medical Research Council

multiple sclerosis

programmed cell death 1

Author Contributions

K.A. DiMauro: major role in the acquisition of data; study concept or design; analysis or interpretation of data. M.E. Heber: major role in the acquisition of data; study concept or design. J. Lee: analysis or interpretation of data. J.A. Cohen: drafting/revision of the manuscript for content, including medical writing for content. E.P. Flanagan: drafting/revision of the manuscript for content, including medical writing for content. C. Riley: drafting/revision of the manuscript for content, including medical writing for content. M.D. Goldman: drafting/revision of the manuscript for content, including medical writing for content. S.S. Zamvil: drafting/revision of the manuscript for content, including medical writing for content. A.C. Kunchok: major role in the acquisition of data; study concept or design; analysis or interpretation of data.

Study Funding

The authors report no targeted funding.

Disclosure

K.A. DiMauro, M.E. Heber, and J. Lee report no disclosures; J.A. Cohen has consulted for Astoria, Bristol-Myers Squibb, Convelo, EMD Serono, FiND Therapeutics, INMune, and Sandoz; and served as an Editor of Multiple Sclerosis Journal; A.C. Kunchok has consulted for Genentech, Horizon Therapeutics and EMD Serono and is an editor for the Neurology journal. Go to Neurology.org/NN for full disclosures.

References