Magnetic Resonance Imaging and Clinical Findings Associated with Choroid Plexus Spinal Cord “Drop” Metastases
ABSTRACT
A 5 yr old castrated male whippet presented with a unique presentation of ambulatory paraparesis and subsequent diagnosis of primary intracranial choroid plexus carcinoma, with metastases to the cervical, thoracic, lumbar, and sacral spinal cord segments. Magnetic resonance imaging was performed initially of the thoracolumbar vertebral column and was followed by MRI sequences of the brain for confirmation of the presence of a primary intracranial tumor. The dog was euthanized immediately following diagnostic imaging due to the severity of clinical signs and poor prognosis. Postmortem examination confirmed the presence of a primary choroid plexus carcinoma, with diffuse metastatic lesions to the spinal cord. To the authors’ knowledge, this is the first description of MRI findings of spinal cord metastasis of a choroid plexus carcinoma in a dog. Additionally, this case demonstrates primary clinical signs arising from metastatic lesions. This confirms the importance of extensive neuroimaging investigations when suspecting choroid plexus tumors and evaluating response to treatment regimens.
Introduction
Choroid plexus tumors are considered uncommon in canine patients, representing only 5 to 10% of all intracranial neoplasms.1,2 Previous studies indicate that the median age at time of diagnosis of these patients is 6 yr.1,3–6 A predilection for males and large-breed dogs has also been identified, with reports indicating that golden retrievers, Dalmatians, and English springer spaniels are overrepresented.1,3–6 These tumors are most commonly found arising within the third and fourth ventricles, often causing clinical signs secondary to associated hydrocephalus, such as behavioral changes, seizures, hemiparesis, tetraparesis, and, in some cases, cerebellar or vestibular dysfunction.2–5,7
Choroid plexus papillomas are relatively rare, benign neoplasms of young human children.8 They commonly affect those less than 2 yr of age, and they comprise approximately 2–5% of brain tumors in children and less than 1% of all brain tumors in people.8–10 Choroid plexus carcinomas make up a subpopulation of 10% of choroid plexus tumors.10 These more aggressive tumors are most common in children under 10 yr of age.10 Clinical signs in humans and dogs are often attributed to the associated hydrocephalus.8 Extraneural spread of choroid plexus tumors is extremely rare in human medicine, and reports are limited to small cases studies, which describe spread to tissues, including the lungs, abdomen, and clavicle.10–12 There are no reports in veterinary medicine of extraneural spread at this time. Ventricular spread to the subarachnoid space of the spinal cord, also known as drop metastases, has been reported in humans and dogs with choroid plexus carcinomas and papillomas.3–5,7–9,13,14 Drop metastases are not unusual in canine choroid plexus tumors, with a previous study in dogs reporting a prevalence of 19%.3 In choroid plexus carcinomas, the rate has been identified as high as 35% at the time of necropsy.3 This finding is very consistent with studies in the human literature, in which 12% of patients will have metastasis at the time of diagnosis.10 To our knowledge, none of the previous case reports of canine choroid plexus tumors describe the features of spinal cord drop metastasis on MRI. Additionally, there are very limited reports of patients with choroid plexus tumors, which were diagnosed at necropsy, exhibiting clinical signs from extracrainial metastasis of choroid plexus tumors in the veterinary literature.15
Case Report
Clinical Presentation
A 5 yr old castrated male whippet was evaluated for a 3-wk history of progressive pelvic limb weakness, reluctance to posture to urinate, and difficulty walking. Complete blood cell count and serum biochemical analysis were within reference ranges. Based on the young age of the patient and his clinical examination, no thoracic radiographs were completed.
On physical examination, the patient's body condition score was four out of nine. Neurological examination demonstrated normal mentation and cranial nerve examination. Evaluation of posture and gait found the dog to have a wide-based stance of the pelvic limbs, moderate paraparesis, and general proprioceptive ataxia of the pelvic limbs. Both pelvic limbs showed reduced postural reactions and muscle tone with normal muscle mass. The patellar reflexes were normal bilaterally, and the tail also appeared limp with minimal movement. The perineal reflex was normal, and the owners reported no changes in urination or defecation. Pain was elicited on lumbar palpation. Neuroanatomical localization of L4–S3 spinal cord segments was based on pelvic limb weakness, ataxia, reduced withdrawal reflexes, and tail paralysis.
Diagnostic Imaging
Magnetic resonance imaging of the T1–S3 vertebral column was conducted using a 1.5 Tesla magnet with a phased array coil. Intradural multifocal extramedullary masses were noted to be hyperintense on T2-weighted images (T2WI), isointense on T1-weighted images (T1WI), and uniformly contrast enhancing on T1WI following intravenous contrast administration (Gadopentetate dimeglumine, 94 mg/kg). Multifocal masses ranging from pinpoint to 0.6 mm were noted within the thoracic, lumbar, and sacral vertebral canal. The majority of metastatic lesions were small and located over the L4–L7 vertebral bodies. The masses gave the spinal cord an irregular shape within the dura. The spinal cord showed abnormal ill-defined intramedullary T2WI signal hyperintensity. At least one focal mass was noted over each vertebral body throughout the lumbar vertebral column and were less frequent within the thoracic vertebral column. The largest mass was dorsal to the T10/11 intervertebral disc space. In conclusion, multifocal homogeneous contrast-enhancing intradural extramedullary masses with associated intramedullary changes were present within the thoracolumbar and lumbosacral spinal cord segments (Figure 1).
Citation: Journal of the American Animal Hospital Association 53, 5; 10.5326/JAAHA-MS-6479
Based on the large number, relatively small size, and extensive distribution along the spinal cord and cauda equina, these findings were most consistent with metastatic lesions causing tumor seeding via the ventricular system and into the subarachnoid space. The route of hematogenous spread could not be completely excluded based on these findings alone but was considered less likely. Differentials following review of the initial MRI sequences included the following: neoplasia, such as choroid plexus tumor with associated metastasis; meningiomatosis; multifocal or metastatic histiocytic sarcoma; and lymphoma; with lesser consideration given to inflammatory diseases, including infectious or immune-mediated etiologies. Magnetic resonance imaging of the brain was completed to rule out spread from the intracranial structures or diffuse metastatic disease.
Inspection of the brain showed multiple intracranial masses. The first extraparanchymal mass measured 0.77 cm in diameter, adjacent to the left olfactory peduncle. It was iso- to slightly hyperintense to the normal grey matter on T2WI and mildly hyperintense on T1WI. Following contrast administration, a prominent rim was visible with mild heterogeneous contrast enhancement within the mass. This lesion had a broad base adjacent to the presphenoid bone and caused a mild mass effect. T2-weighted images showed a moderate amount of ill-defined hyperintensity that appeared within the grey and white matter surrounding the mass, consistent with perilesional edema. A fluid-attenuated inversion recovery sequence was not completed to further confirm this finding. A second, 0.73 cm in diameter, extraparanchymal mass was seen rostral to the cerebellum within the subarachnoid space, dorsal to the right rostral colliculus, and extended just caudal to the caudal colliculus. The mass slightly to the right of midline caused a mass effect with evidence of an ill-defined T2WI hyperintensity surrounding the lesion, suggesting edema. The mass was uniformly contrast enhancing similar to the lesions in the spinal cord. The third and largest mass measured 1.28 cm by 0.92 cm and arose from the right lateral aspect of the fourth ventricle. The bulk of the mass was located in the cerebellopontinemedullary angle. This mass had a broad base adjacent to the temporal bone and identical imaging characteristics to the second intracranial mass and spinal cord masses. Dural tails or associated meningeal enhancements were not identified with any of the intracranial lesions. Based on the imaging characteristics and location of the largest mass, these lesions were consistent with neoplasia such as a choroid plexus carcinoma, lymphoma, or granulomatous disease, with lesser consideration given to other metastatic or multifocal neoplastic conditions. Based on the compelling evidence of multifocal central nervous disease, the owners declined further treatment and elected humane euthanasia.
Confirmation of diagnosis was made at postmortem examination. Grossly, the extent of disease was wider spread than appreciated on MRI. Additional lesions were found in the cervical vertebral column, and an increased number of masses were reported throughout the lumbar spinal cord. The largest mass within the vertebral canal was located at the level of the C6 vertebra. This mass measured 1 cm in diameter, was intramedullary, oval-shaped, slightly raised, and grey in color. Extending from L5 and along the spinal nerve roots of the cauda equine, were grey, firm, slightly nodular, coalescing masses within the subarachnoid space.
On gross sectioning of the brain, two obvious discrete extraparanchymal masses were noted. The first most-rostral mass described on MRI was not identified on gross sectioning of the brain. The second mass began at the right margin of the third ventricle at the level of the pituitary gland (Figure 2A, B, C) and extended caudally to the right dorsolateral surface of the rostral brainstem. This mass varied in size from a pinpoint in diameter rostrally, to 0.2 cm caudally. The third mass was 0.5 cm in diameter, located along the right ventrolateral cerebellum, and appeared to be arising from the choroid plexus of the fourth ventricle. The mass previously noted within the C6 vertebral canal extended 0.3 cm into the spinal cord (Figure 2D) and effaced the normal architecture. At the level of L2, a 0.2 cm in diameter, intramedullary, grey mass obscured the normal spinal cord architecture (Figure 2E). The mass extended caudally from the lumbar intumescence circumferentially expanding the subarachnoid space.
Citation: Journal of the American Animal Hospital Association 53, 5; 10.5326/JAAHA-MS-6479
Histopathology was used to confirm the tumor type as a choroid plexus tumor, specifically a papillary carcinoma with meningeal carcinomatosis. All the masses were of similar architecture, intimately associated with the subarachnoid space, and invading into the parenchyma. The neoplastic cells were arranged in cords, and papilliferous projections were supported by a moderate amount of fibrous stroma. Multifocally, the neoplastic cells formed nests and tubules surrounded by concentric whorls of fibrous tissue (scirrhous response). The neoplastic cells were cuboidal to columnar in shape, with a moderate amount of eosinophilic cytoplasm, variably distinct cell borders, and contained apical structures suspicious for cilia. The nuclei were round with reticular chromatin and indistinct nucleoli. There was mild anisocytosis, anisokaryosis, and one mitotic figure per ten high-powered fields. Small numbers of lymphocytes and plasma cells were scattered along the periphery of the neoplastic nodules (Figure 2).
Discussion
The tumor morphology described was very characteristic of choroid plexus tumors across species and consistent with previous descriptions in the human and veterinary literature.2–6,8,9,13,14 This case has a unique clinical presentation secondary to metastatic lesions of a choroid plexus tumor and includes the first description of choroid plexus drop metastasis on MRI within the spinal cord of a dog.
As differential diagnoses in this case following advanced imaging focused primarily on metastatic neoplastic causes, MRI sequences through the brain were completed to identify a primary mass within the central nervous system. Following the acquisition of these images, we were able to eliminate the majority of other potential etiologies based on the finding of a considerable primary mass arising from the choroid plexus. The primary differential became that of a choroid plexus tumor with intracranial and lumbar spinal cord metastases. On T2WI, the metastatic lesions were hyperintense, showing intensities similar to cerebrospinal fluid (CSF). On T1WI, following contrast administration, the masses exhibited moderate uniform enhancement. A previous report described small cystic choroid plexus metastasis, which appeared to not contrast enhance.14 This was thought to be due to volume averaging of small fluid-filled structures that would appear hypointense.14 The likelihood of the remaining tumor types or an inflammatory process to develop the number of lesions seen in the brain and spinal cord was considered unlikely, although not impossible for tumors such as lymphoma or histiocytic sarcoma. Furthermore, diffuse spread through the neuroaxis by other tumors or inflammatory causes is generally via a hematogenous route, and the location and distribution of the masses seen here did not fit with this pattern. All lesions in this patient were adjacent to the ventricular system or in contact with the subarachnoid space, suggesting metastatic spread through CSF flow. Although the other differentials could not be completely discarded at the time of imaging, they were considered much less likely.
From the findings in this case, and in conjunction with the previous retrospective studies, the frequency of metastasis at the time of diagnosis may be more common than initially thought.3,6 With postmortem findings of 19% of choroid plexus tumors already showing metastasis, and, more specifically, 35% of choroid plexus carcinomas having metastasis, more thorough staging for patients with choroid plexus tumors should be performed when considering treatment options including radiotherapy or surgery.3 Additionally, it demonstrates a further need for studies better characterizing the biologic behavior of choroid plexus tumors. The large frequency of metastasis seen in veterinary patients at the time of postmortem examination may be due largely to the delay in diagnosis based on quite subtle and often overlooked clinical signs. This case report demonstrates that in rare cases of choroid plexus tumors we can see primary clinical signs arising from distant metastatic lesions before clinical signs associated with the primary tumor are evident. This type of finding would likely be more common in choroid plexus carcinomas, which have higher incidences of exfoliation into the CSF and therefore an increased rate of spread.3,14
Conclusion
Choroid plexus carcinomas make up a small group of unique masses affecting the brain. These rare tumors have a relatively high rate of distant metastasis within the neuroaxis, particularly in veterinary patients, as evaluated on postmortem examination.3 The relatively high rate of metastasis may be secondary to delayed diagnosis, as the initial clinical signs are often very subtle. Therefore, MRI of the entire neuroaxis is recommended prior to treatment of primary choroid plexus tumors to aid in prognosis and treatment recommendations. Here, we have provided a unique case report attributing presenting clinical signs secondary to metastasis of a primary intracranial choroid plexus tumor and the first description of the MRI characteristics of drop metastasis in dogs.
Depicted here are the sagittal (A–C) and transverse (D–F) images of the lumbar spinal cord, accompanied by transverse images (G–J) of the primary brain masses. The sequences are T2-weighted images (T2WI), T1-weighted images (T1WI), and T1WI with contrast from right to left in each row. The masses appear to be intradural extramedullary and of heterogeneous intensity ventral to the spinal cord on T2WI within the subarachnoid space over the L5 to L7 vertebral bodies (C, F). The masses appear isointense on T1WI (B, E) and show homogenous contrast enhancement (A, D). In images (G–J), transverse sections of the brain show a mass associated with the right brainstem in the region of the cerebellomedullarypontine angle. The mass has a broad base against the temporal bone. T1WI, T1-weighted images; T2WI, T2-weighted images.
Transverse sections of the mass at multiple levels. (A) shows a gross transverse section of the brain at the level of the hippocampus and rostral medulla oblongata at the level of the mesencephalic aqueduct. The arrows highlight the mass located within the subarachnoid space rostral to the cerebellum. (B) shows the mass at the same level in hemotoxylin and eosin stain. The papilliform mass can be seen arising adjacent to pons and within the subarachnoid space. At higher magnification, (C) shows the characteristic cuboidal epithelium creating a papillary appearance common to choroid plexus tumors. Images (D) and (E) demonstrate the drop metastasis at the level of C6 and L2, respectively. Basophilic staining masses are seen adjacent to the spinal cord within the meninges due to their seeding through the cerebrospinal fluid.
Contributor Notes
