Multiple myeloma presenting as an intramedullary spinal cord tumor: a case report and review of the literature

Diagnosing MM often requires the evaluation of clinical, radiographic, histopathologic, and laboratory findings [6‐8]. In symptomatic patients, metastasis to the axial skeleton most often presents with back pain, vertebral fractures, or paresthesia and paresis due to spinal cord compression [9]. Lytic bone lesions may result in hypercalcemia, and renal dysfunction may present as anemia and proteinuria [10]. The differential diagnosis of MM may include monoclonal gammopathy of uncertain significance, Waldenstrom macroglobulinemia, and other plasma cell dyscrasias. The CRAB acronym (calcium elevation, renal dysfunction, anemia, and bone disease) was established by the International Myeloma Working Group to summarize the aforementioned clinical manifestations of MM as well as to differentiate between MM and similar plasma cell dyscrasias such as solitary plasmacytoma [11]. Differentiating MM from solitary plasmacytoma involves review of radiographic imaging and bone marrow biopsy [1]. Both dyscrasias require biopsy of bone lesions with evidence of clonal plasma cells. However, in solitary plasmacytoma, the CRAB symptoms are absent; there is no evidence of clonal plasma cells on bone marrow biopsy; and radiographic imaging reveals no other abnormalities aside from the primary lesion.

EMH is relatively uncommon at diagnosis but may occur later in the disease progression or at the time of relapse. Incidence of EMH in newly diagnosed MM is approximately 7–18% [12‐14]. This proportion increases to 6–20% late in the disease course [13‐16]. Several studies have shown that patients with EMH at presentation have significantly shorter survival with conventional chemotherapy [2, 12, 13]. In addition, Pour et al. described significantly worse outcomes in patients with soft tissue–related EMH than in bone-related EMH, with median survival rates of 5 months and 12 months, respectively (P = 0.022) [2]. Median survival of patients with CNS EMH shows similarly poor outcomes, with most studies reporting median survival of 2–8 months [3, 4, 17‐22]. However, several of these studies included only intracranial metastases, and the others did not specify location further than detailing CNS involvement. Due to the extreme rarity of EMH involving the spinal cord, it is difficult to determine whether patient survival may differ from intracranial involvement. In consideration of the small sample size, we report an overall survival of 1–11 months on the basis of our literature review.

Involvement of the CNS is relatively uncommon, with approximately 1–2% of patients with MM exhibiting a secondary CNS malignancy [23, 24]. Most often, these occur as either an intraparenchymal or meningeal lesion. MM metastasis to the spinal cord is exceedingly rare. Our review returned a total of six reported patients with metastatic, intramedullary spinal cord MM or plasmacytoma in the past 15 years, including our patient’s case. Five patients were male (83.3%), and one patient was female (16.7%). Two cases had a final diagnosis of plasmacytoma (33.3%), and four had a diagnosis of MM (66.7%). Bence-Jones proteinuria was reported in two cases (33.3%). The level of spinal cord metastasis varied from the cervical cord to the cauda equina.

Diagnosis of MM in the CNS often involves contrast-enhanced MRI of the head and/or spine. It should be noted that there remains concern for the use of iodine-based contrast agents in MM and monoclonal gammopathies [25]. Many radiologists consider the use of iodinated contrast material to be contraindicated in the setting of MM due to impaired renal function. Following administration of gadolinium contrast, MM may present as diffuse leptomeningeal enhancement or punctate, intraparenchymal lesions [26]. MM of the spinal cord shares similar characteristics. In all cases in which radiologic findings were reported, contrast-enhanced T1 weighted image (T1-WI) MRI of the spine showed a contrast-enhancing lesion; these lesions can present diffusely, with multiple enhancing lesions spread across spinal levels.

Several theories have been posited to suggest how CNS involvement in MM may arise. Local paraskeletal seeding of meninges and subsequent invasion into neural tissue has been suggested. Gozzetti et al. found neuroimaging evidence of contiguous spread arising directly from adjacent bone lesions [4]. Another hypothesis suggests hematogenous spread from traversal of the arachnoid veins by myeloma cells [5]. Spillage of cells into the cerebrospinal fluid would thus then be evident on cytologic assessment [20]. In this way, the neural invasion of MM closely reflects that seen in acute lymphoblastic leukemia, in which CNS involvement initially involves the adventitia bordering the arachnoid veins [27]. Few studies have investigated the molecular basis of CNS myeloma, but the acquisition of p53 gene mutations seems to be associated with advanced forms of disease [28, 29]. In a sample of nine patients, Chang et al. identified an 89% rate of p53 deletions, which stands in stark contrast to 10–15% of patients with MM who harbor p53 deletions but do not have CNS involvement [29]. In our patient’s case, immunostaining revealed only scattered p53 overexpression, suggestive of physiological upregulation rather than a gene mutation. Multiple lytic and lucent osseous lesions throughout the spinal axis were identified, notably in the thoracic spine. These were attributed to either multiple osseous degenerative changes or multifocal myelomatous disease. Thus, it is possible that the intramedullary disease in our patient arose from contiguous spread from nearby bone lesions. However, the presence of the second T6–T7 lesion located dorsally along the cord favors hematogenous spread from penetrating branches of the posterior spinal arteries, because contiguous spread throughout the width of the spinal cord from a vertebral lesion seems less likely.

Treatment of CNS MM is not well defined; however, some studies have suggested systemic therapy, occasionally with adjuvant radiation [20]. Systemic strategies present challenges because the blood–brain barrier (BBB) precludes treatment with traditional chemotherapeutics, such as high-dose melphalan, cyclophosphamide, proteasome inhibitors, or monoclonal antibodies. High-dose methotrexate or cytarabine is effective in penetrating the CNS but is ineffective against MM [30]. Thus, effective therapy necessitates good BBB permeability as well as action against MM. Thalidomide, lenalidomide, and pomalidomide have been shown to traverse the BBB in primates [31], and combination thalidomide and bendamustine has been shown to achieve a robust effect [32]. Of course, its use has yet to be validated in large-sample studies. More recently, selinexor was approved for patients with advanced MM, and this agent seems to cross the BBB. Combination BRAF and mitogen-activated protein kinase kinase 1/2 inhibitors dabrafenib/trametinib have also been employed to good response; however, recent evidence suggests potential mechanisms of drug resistance [33]. Given the reported radiosensitivity of plasma cell dyscrasias, concurrent radiotherapy may be considered [4, 34, 35]. Ultimately, intrathecal chemotherapy with a systemic anti-MM immunomodulatory regimen and cranial and/or spinal irradiation seems an ideal approach to management [21, 30]. Of six patients in our review, three patients (50%) were treated with chemoradiation therapy (chemo-RT); two patients (33.3%), including our own, underwent resection with adjuvant chemo-RT; and one patient (16.7%) underwent biopsy with chemo-RT.

All patients treated with chemo-RT saw progression in paresis and other neurological deficits over the course of treatment. Reported overall survival ranges from 1.4 to 11 months among patients treated with chemo-RT only. The role of surgery in intracranial EMH is unclear, and resection is not frequently performed [20]. However, in spinal cord metastasis, resection may assume a more prominent role because mass effect from intramedullary lesions may perturb motor and sensory tracts and compress nerve roots, contributing to radiculopathies. We found no data for patients treated with resection and chemo-RT. Hans et al. noted that surgical excision should be considered whenever possible. However, it is difficult to interpret the effect of surgery on patient morbidity and mortality, given the paucity of data. Our patient did not experience any perioperative complications, but the long-term effect of surgery on disease progression has yet to be observed.