Ecchordosis physaliphora (EP) is an uncommon notochordal remnant at the posterior clivus seen in approximately 2% of autopsies. Most of the patients are asymptomatic. It can be confused with its malignant counterpart, chordoma.Reference Mehnert, Beschorner and Kuker 1 We report a symptomatic middle-aged female with EP who presented with meningitis following a several-year history of intermittent cerebrospinal fluid (CSF) leak.
Case Report
A middle-aged patient presented to the emergency department with new onset of severe headache and confusion. The physical examination was remarkable for nuchal rigidity; the neurological examination showed no other focal findings. Laboratory testing revealed an elevated white blood count of 19.8×109/L; a diagnosis of meningitis was confirmed after lumbar puncture revealed cloudy CSF with elevated white blood cells, high protein, and low glucose. The unenhanced computed tomography (CT) scan was unremarkable except for opacification of the sphenoid sinus with fluid. The patient was diagnosed with bacterial meningitis (Streptococcus salivarius grew from CSF samples) and treated with antibiotics. A subsequent CT scan of the paranasal sinuses performed 3 days later demonstrated a small clival defect associated with a small osseous stalk protruding into the prepontine cistern (Figure 1A). Magnetic resonance imaging done on the following day demonstrated a nonenhancing small T2-hyperintense and T1-hypointense retroclival lesion associated with a direct connection between the prepontine cistern and the sphenoid sinus suggestive of a CSF leak on the axial fast imaging employing steady-state acquisition sequence. A diagnosis of EP was suggested (Figure 1B-D).
Figure 1 (A) Sinus computed tomography (CT; sagittal 3 mm). There is a small retroclival defect associated with a small osseous stalk protruding into the prepontine cistern. Opacification of the sphenoid sinus with attenuation consistent with fluid suspicious of cerebrospinal fluid (CSF) leak. (B) Magnetic resonance imaging (MRI; 1.5 Tesla). Axial fast imaging using steady-state acquisition (0.8 mm thickness). There is a small intermediate signal intensity lesion (straight arrow) attaching to the retroclivus via a short pedicle. There is a CSF leak (curve arrow) on the left side of the lesion from the prepontine cistern to the left sphenoid sinus. Dark signal intensity in the center of the lesion is the bony protrusion component as seen on prior CT. The findings are consistent of an ecchordosis physaliphora (EP) causing a CSF leak. (C) MRI (1.5 Tesla). Sagittal T2 (3-mm thickness). There is a small, mildly T2-hyperintense lesion attaching to the retroclivus via a short pedicle (arrow). There is an associated tiny bone defect in the midline clivus causing a CSF leak between the prepontine cistern and the left sphenoid sinus. The findings are consistent with an EP causing CSF leak. (D) MRI (1.5 Tesla). Postgadolinium axial T1 (2-mm thickness). There is no significant enhancement of the T1-hypointense retroclivus lesion, which favors a benign etiology such as EP rather than chordoma or metastasis.7 , 23
The patient was treated for meningitis with a full recovery and continued to have CSF leakage. The patient was treated with an endoscopic endonasal approach for repair of sphenoid bony defect with fat graft after resection of prepontine lesion.
Microscopically, it consisted of sheets and clusters of large cells with central nucleus and vacuolated cytoplasm in myxomatous stroma, associated with fibrovascular tissue (Figure 2A, B). There was no significant mitotic activity (Ki67 labeling was less than 1%) and no cellular pleomorphism. The characteristic physaliferous cells were immunoreactive to S100 and monokeratin (Figure 2C, D, respectively), in keeping with ecchordosis physaliphora.
Figure 2 Biopsy from the mass shows a cluster of large physaliferous cells with a characteristic bubbly cytoplasm and a round to oval nuclei in the myxomatous stroma (A: hematoxylin and eosin, ×40; B: hematoxylin and eosin, ×200). The lesional cells were immunoreactive to S100 (C: ×100) and monokeratin (D: ×100).
Discussion
Most cases of EP are seen incidentally or on autopsy. Very few symptomatic cases of EP have been reported.Reference Cha, Jarrahy, Yong, Eby and Shahinian 2 Typically, symptoms are directly related to the mass effect of EP on the surrounding structures, perilesional hemorrhage, or CSF leak.Reference Mehnert, Beschorner and Kuker 1 , Reference Cha, Jarrahy, Yong, Eby and Shahinian 2 There have been only four cases of EP associated with CSF leak reported in the literature.Reference Dias, Nakanishi, Mangussi-Gomes, Canuto, Takano and Oliveira 3 - Reference Alli, Clark and Mansell 6 We report the fifth case of EP presenting with CSF leak to remind readers that this can be a rare cause of meningitis.
EP is a benign, gelatinous hamartomatous lesion that arises from an ectopic notochordal remnant occurring along the midline craniospinal axis anywhere from the dorsum sella to the coccyx.Reference Mehnert, Beschorner and Kuker 1 It is found incidentally in approximately 0.5% to 2% of autopsies.Reference Cha, Jarrahy, Yong, Eby and Shahinian 2 Most of the intracranial EPs are found as retroclival masses located intradurally in the prepontine cistern. It is difficult to accurately differentiate between EP and its malignant counterpart chordoma (2%-4%) clinically, radiologically, and even histopathologically.Reference Mehnert, Beschorner and Kuker 1 , Reference Cha, Jarrahy, Yong, Eby and Shahinian 2 Distinction between EP and chordoma is important for treatment planning and prognosis, and there are certain characteristics that can be used to differentiate them.
CT scan has some role in diagnosing EP; it can typically demonstrate a tiny osseous stalk.Reference Mehnert, Beschorner and Kuker 1 Magnetic resonance imaging is the modality of choice in detecting, localizing, and characterizing EP. EP is typically T1-hypointense and T2-hyperintense. An important differentiating imaging feature of EP is the lack of gadolinium enhancement, whereas chordoma would show significant enhancement.Reference Mehnert, Beschorner and Kuker 1 - Reference Dias, Nakanishi, Mangussi-Gomes, Canuto, Takano and Oliveira 3 Heavily T2-weighted image used for the CSF space is very useful in diagnosis of smaller lesions, such as that of our case. This sequence is also useful to delineate the CSF leakage. EP and chordoma share a common embryonic notochordal origin, and therefore share similar histopathological features. Hypocellularity, sparse pleomorphism, and very low mitotic index favor the diagnosis of EP over chordoma, but are not pathognomonic.Reference Mehnert, Beschorner and Kuker 1
Five-year survival for chordoma is only about 10%.Reference Cha, Jarrahy, Yong, Eby and Shahinian 2 Clinically, most of the EPs are asymptomatic, whereas most previously reported cases of intradural chordoma had cranial nerve palsy or brain stem symptoms.Reference Mehnert, Beschorner and Kuker 1 The symptoms are directly related to the mass effect of EP on the surrounding structures, perilesional hemorrhage, or CSF leak.Reference Mehnert, Beschorner and Kuker 1 , Reference Cha, Jarrahy, Yong, Eby and Shahinian 2 The cause of CSF leakage late in life in what is presumably a congenital lesion remains unknown; others have speculated that head injury, infection, or age-related bone density changes may be contributory.Reference Macdonald, Cusimano, Deck, Gullane and Dolan 5 One might speculate that pulsations of the basilar artery contribute to a CSF leak; however, in our case, the dural defect was not immediately underneath the basilar artery.
To conclude, this case highlights the importance of identifying EP as a cause of CSF leak and meningitis. Following treatment of meningitis, surgery can be offered for removal of this benign entity and repair of the CSF fistula.
Acknowledgements
The authors thank Dr. Alex Easton for his assistance with preparing the histopathology.
Consent was not required for this study. There is no additional unpublished data from this study that will need to be made available. No funding was received for this project.
Disclosures
CF, DBC, NS, and JJSS have nothing to disclose.
Statement of Authorship
JJSS and DC wrote the case report and searched the literature. This was JJSS’s and DC’s patient and they provided the leadership in starting this case. CF edited the report and provided an update on the literature search. NS provided the pathology for this case.
