Clinical Radiology (1992) 46, 139 141
Case Report: Paraganglioma of the Cauda Equina C. M Y L O N A S
Department of Neurological Surgery, Walsgrave Hospital, Coventry The clinical, radiological and pathological features of a paraganglioma of the cauda equina are described, including magnetic resonance imaging features. The literature is reviewed and discussed. Mylonas, C. (1992). ClinicalRadiology 46, 139-141. Case Report: Paraganglioma of the Cauda Equina
Paragangliomas of the cauda equina are rare. The first published, although unrecognized case was reported in 1970 [1] and since then 67 cases have been published. A further patient with this lesion is presented and discussed with particular reference to the difficulties in radiological and neuropathological diagnosis and management. CASE REPORT A 62-year-old retired cleaner presented with a 7 year history of back pain precipitated by a fall from a ladder in which he sustained a crush fracture of TI0. He subsequently returned with a worsening of symptoms over 1 year and particularly the last 3 months. On this occasion he complained of pain radiating to both legs aggravated by coughing and weight bearing. Examination revealed loss of the normal lordotic curve with local tenderness at L4 and L5 and a spastic paraparesis with a sensory level at TII. Plain radiology revealed only marked degenerative changes in the posterior joinl~ of the lower l u m b a r spine. Magnetic r~sonance imaging (MRI) using a Picker 2055HP 0.5 T system with surface coils and 6 m m thick sagittal Tl-weighted (spin-echo 500/20) images (Fig. 1) showed an intradural, extramedullary t u m o u r isointense with the cord and occupying the entire canal at the level of the
conus. Six millimetre thick sagittal T2-weighted images (SE 2000/100) showed the t u m o u r hypointense compared to CSF (Fig. 3). It showed moderate homogeneous enhancement on the sagittal Tx-weighted images after intravenous gadopentetate (Magnevist, Schering Health Care Ltd) (Fig. 2). At surgery a purplish t u m o u r was found amongst the nerve roots with its upper pole attached to the conus and gaining a blood supply from it. Once this was divided the residual excision was straightforward and complete• The eonus had been pushed posteriorly and deformed by the tumour. H & E staining showed a vascular neoplasm with uniform cells and associated areas of perivascular rosetting. Mitoses were extremely rare. Staining for glial fibrillary acidic protein was uniformly negative• Neurone specific enolase was positive in all t u m o u r cells. Electron microscopy revealed m a n y angular cells possessing round or oval nuclei with little hetero-chromatin. Higher magnification (Fig. 4) showed many small spherical dense cored neurosecretory granules with an average diameter of 100 nm. These possessed electron dense cores with a dark limiting membrane beneath which was an electron lucent 'halo'• Fine filaments (Fig. 5) were abundant. N o convincing cilia or microvilli were s e e n . The patient made a good post-operative recovery and was treated with a course of radiotherapy on the strength of the preliminary histological report of the H & E stained sections, interpreted as suggesting a diagnosis of ependymoma. The final histological diagnosis based on the electron microscopic appearances was a paraganglioma of the cauda equina. At 6 m o n t h follow-up, he was neurologically intact and asymptomatic.
Fig. 1 - Sagittal T~-weighted M R I (spin-echo 500/20) demonstrating a well defined intradural extramedullary mass of homogeneous intensity occupying the entire width of the canal at the Ll level. Correspondence to: C. Mylonas, Department of Neurological Surgery, Walsgrave Hospital, Clifford Bridge Road, Walsgrave on Sowe, Coventry CV2 2JB.
Fig. 2 Sagittal Ti-weighted M R I (spin-echo 500/20) after i.v. gadopentetate demonstrating uniform enhancement and displacement of the conus posteriorly.
140
CLINICAL RADIOLOGY
DISCUSSION
Fig. 3 - Sagittal T2-weighted MRI (spin-echo 2000/100) showing the tumour and demonstrating minimal hyperintensity compared to the conus but hypointensity compared to CSF.
Fig. 4 - H i g h power (x 15000) electron micrograph demonstrating many neurosecretory granules (arrow).
Fig. 5 High power (x20000) electron micrograph showing the abundant fine filaments, often arranged in whorls (arrow).
The pre-operative diagnosis of paraganglioma of the cauda equina is difficult with very few clues from the clinical history that differentiate this from other causes of the cauda equina syndrome. In a recent review from the Mayo clinic [2] there was a slight male preponderance and a mean age of 51, although it has been described in a 13-year-old boy [3]. Virtually all patients present with a history of lumbar back pain with or without sciatic radiation which may often be bilateral. There is no doubt that diagnosis was delayed in this patient due to the coincidental but otherwise unrelated fall. It is impossible to say whether this patient's long history of back pain was as a result of the fall or the tumour but the radicular pain was undoubtedly due to the latter. The duration of symptoms in this case is typical of most described in the literature, varying between a few months and several years. Neurological signs are often absent (seen in only 11 of 31 cases described by Sonneland et al. [2]). Only one case has been reported to be endocrinologically functional at this site [4] and one patient is thought to have bled into the subarachnoid space [5]. Bony erosion has been described [6,7] and may give a clue on plain radiographs but until recently myelography has been the mainstay of pre-operative anatomical definition. This, however, has often proved difficult due to the presence of the tumour resulting in a 'dry' or 'traumatic tap' at myelography. MRI has significant advantages over myelography but has been described in only 3 previous cases [8-10]. The common differential diagnoses that must be considered on the basis of the MRI findings include meningioma, neurofibroma and ependymoma. The degree of paramagnetic enhancement in our case was less than might be expected with any of these differential diagnoses and the relatively hypointense signal on T2weighted images is atypical of a neurofibroma. Ependymorea often do not have such a uniform rounded appearance as did this case. Although the imaging findings were not 'classical' of any of the differential diagnoses described above, the possibility of a paraganglioma was not considered. This reflects the general lack of awareness of the condition. However, it is clear that with the current clinical techniques and imaging available it is still not possible to make a definitive diagnosis pre-operatively. Histopathological diagnosis can also be difficult. Our case like many previously published, was reported as an ependymoma on light microscopy. The tumour has also been mistaken for metastatic adenocarcinoma [11]. The main confusion with ependymoma at the light microscopic level stems from a radial perivascular pattern of cells which may look like the perivascular pseudorosettes of ependymoma and occasional papillary epithelial structures which resemble ependymal tubules. The ultrastructural and immunohistochemical appearance of paragangliomas is unique and easily distinguishes such lesions, their characteristic feature being the presence of dense core granules on electron microscopy. These have been shown to contain predominantly dopamine, adrenaline and noradrenaline [12] although others have found 5H T at higher levels [10]. Despite this, only one case has been shown to be endocrinologically functional [4]. The true significance of the correct diagnosis of a paraganglioma of the cauda equina lies in its apparent
PARAGANGLIOMAOF THE CAUDA EQUINA benign biological b e h a v i o u r c o m p a r e d with other sites. A l t h o u g h they are i n v a r i a b l y intradural, six cases [2,4, l 1,13,14] have been associated with extradural extension a n d because of their e n c a p s u l a t i o n they can usually be excised totally. D i s t a n t metastases have n o t been described b u t local recurrence can be a p r o b l e m if subtotal r e m o v a l is performed. O f the 67 cases thus far reported there have been no instances o f recurrence where complete excision was possible. There have been 10 cases where s u b t o t a l excision was performed. I n three of these no r a d i o t h e r a p y was given, one p a t i e n t died of a c o m p l i c a t i o n o f his t u m o u r [2]. I n another, a recurrence occurred 12 years later [4] a n d one p a t i e n t was lost to follow-up [2]. The r e m a i n i n g seven patients received r a d i o t h e r a p y a n d of these two have recurred at 1 year [2] and 9 years [ l l ] respectively. At present there are n o long-term follow-up d a t a on these patients and whereas total excision would a p p e a r to be a n adequate t r e a t m e n t offering cure, the a u t h o r agrees with others [7] that l o n g - t e r m o u t l o o k must be g u a r d e d with the limited follow-up reported in the literature. This is particularly i m p o r t a n t for those t u m o u r s resected subtotally with or w i t h o u t a d j u v a n t radiotherapy. In conclusion, p a r a g a n g l i o m a s in extrathecal locations have a p o o r prognosis, hence there is a need for further d o c u m e n t a t i o n of the c a u d a e q u i n a t u m o u r s in order to establish whether or n o t their b e h a v i o u r is u n i f o r m l y benign. With the present imaging available, a definitive pre-operative diagnosis is unlikely b u t this diagnosis needs to be b o r n e in m i n d in the atypical case. Acknowledgements.The author wishes to thank Mr A. E. Booth for kind permission to discusshis patient and to Dr D. J. Bealeand Dr H. L. Whitwell for advice on the radiology and histopathology respectively.
141
REFERENCES
1 Miller CA, Torack RM. Secretory ependymoma oi" the filum terminale. Acta Neuropathologica 1970;15:240 250. 2 Sonneland PRL, Scheithauer BW, LeChago J, Crawford BG, Onofrio BM. Paraganglioma of the cauda equina region. Cancer 1986;58:1720 1735. 3 Softer D, Pittaluga S, Caine Y, Feinsod M. Paraganglioma of the cauda equina. Cancer 1983;51:1907 1910. 4 Boker D-K, Wassman H, SolymosisL. Paragangliomas of the spinal canal. Surgical Neurology 1983;19:461 486. 5 Houroupian D, Kerson LA, Saiontz H, VatsamisM. Paraganglioma ofcauda equina. Cancer 1974;33:1337-1348. 6 Ironside JW, Royds JA, Taylor CB, TimperleyWR. Paraganglioma of the cauda equina: A histological, ultrastructural and immunocytochemical study of two cases with a reviewof the literature. Journal of Pathology 1985;145:195 201. 7 O'Sullivan MG, Keohane C, Buckley TF. Paraganglioma of the cauda equina: Case report and review of the literature. British Journal of Neurosurgery 1990;4:63 67. 8 Nagata K, Kakizoe M, Takagi H, Ohashi T, Yamamoto S, Inoue A, et al. Paraganglioma of the cauda equina. Kurume Medical Journal 1988;35:95 99. 9 Hayes E, Lippa C, Davidson R. Paragangliomas of the cauda equina. American Journal of Neuroradiology 1989;10:$45 $47. 10 Piggot TJD, Lowe JS, Morrel K, Kerslake RW. Paraganglioma of the cauda equina. Report of three cases. Journal of Neurosurgery 1990;73:455 458. 11 Taxy JB. Paraganglioma of the cauda equina. Cancer 1983;51:19041906. I2 Llena JF, Wisoff HS, Hirano A. Gangliocytic paraganglioma in cauda equina region with biochemical and neuropathological studies. Journal of Neurosurgery 1982;56:280-282. I3 Lagace R, Delage C, Gagne F. Paraganglioma of the filum terminale. Canadian Journal of Neurological Science 1978;5:257 260. 14 Gaffney EF, Doorly T, Dinn JJ. Aggressive oncocytic neuroendocrine tumour ('oncocytic paraganglioma') of the cauda equina. Histopathology 1986;10:311 319.
Case Report: Paraganglioma of the Cauda Equina C. M Y L O N A S
Department of Neurological Surgery, Walsgrave Hospital, Coventry The clinical, radiological and pathological features of a paraganglioma of the cauda equina are described, including magnetic resonance imaging features. The literature is reviewed and discussed. Mylonas, C. (1992). ClinicalRadiology 46, 139-141. Case Report: Paraganglioma of the Cauda Equina
Paragangliomas of the cauda equina are rare. The first published, although unrecognized case was reported in 1970 [1] and since then 67 cases have been published. A further patient with this lesion is presented and discussed with particular reference to the difficulties in radiological and neuropathological diagnosis and management. CASE REPORT A 62-year-old retired cleaner presented with a 7 year history of back pain precipitated by a fall from a ladder in which he sustained a crush fracture of TI0. He subsequently returned with a worsening of symptoms over 1 year and particularly the last 3 months. On this occasion he complained of pain radiating to both legs aggravated by coughing and weight bearing. Examination revealed loss of the normal lordotic curve with local tenderness at L4 and L5 and a spastic paraparesis with a sensory level at TII. Plain radiology revealed only marked degenerative changes in the posterior joinl~ of the lower l u m b a r spine. Magnetic r~sonance imaging (MRI) using a Picker 2055HP 0.5 T system with surface coils and 6 m m thick sagittal Tl-weighted (spin-echo 500/20) images (Fig. 1) showed an intradural, extramedullary t u m o u r isointense with the cord and occupying the entire canal at the level of the
conus. Six millimetre thick sagittal T2-weighted images (SE 2000/100) showed the t u m o u r hypointense compared to CSF (Fig. 3). It showed moderate homogeneous enhancement on the sagittal Tx-weighted images after intravenous gadopentetate (Magnevist, Schering Health Care Ltd) (Fig. 2). At surgery a purplish t u m o u r was found amongst the nerve roots with its upper pole attached to the conus and gaining a blood supply from it. Once this was divided the residual excision was straightforward and complete• The eonus had been pushed posteriorly and deformed by the tumour. H & E staining showed a vascular neoplasm with uniform cells and associated areas of perivascular rosetting. Mitoses were extremely rare. Staining for glial fibrillary acidic protein was uniformly negative• Neurone specific enolase was positive in all t u m o u r cells. Electron microscopy revealed m a n y angular cells possessing round or oval nuclei with little hetero-chromatin. Higher magnification (Fig. 4) showed many small spherical dense cored neurosecretory granules with an average diameter of 100 nm. These possessed electron dense cores with a dark limiting membrane beneath which was an electron lucent 'halo'• Fine filaments (Fig. 5) were abundant. N o convincing cilia or microvilli were s e e n . The patient made a good post-operative recovery and was treated with a course of radiotherapy on the strength of the preliminary histological report of the H & E stained sections, interpreted as suggesting a diagnosis of ependymoma. The final histological diagnosis based on the electron microscopic appearances was a paraganglioma of the cauda equina. At 6 m o n t h follow-up, he was neurologically intact and asymptomatic.
Fig. 1 - Sagittal T~-weighted M R I (spin-echo 500/20) demonstrating a well defined intradural extramedullary mass of homogeneous intensity occupying the entire width of the canal at the Ll level. Correspondence to: C. Mylonas, Department of Neurological Surgery, Walsgrave Hospital, Clifford Bridge Road, Walsgrave on Sowe, Coventry CV2 2JB.
Fig. 2 Sagittal Ti-weighted M R I (spin-echo 500/20) after i.v. gadopentetate demonstrating uniform enhancement and displacement of the conus posteriorly.
140
CLINICAL RADIOLOGY
DISCUSSION
Fig. 3 - Sagittal T2-weighted MRI (spin-echo 2000/100) showing the tumour and demonstrating minimal hyperintensity compared to the conus but hypointensity compared to CSF.
Fig. 4 - H i g h power (x 15000) electron micrograph demonstrating many neurosecretory granules (arrow).
Fig. 5 High power (x20000) electron micrograph showing the abundant fine filaments, often arranged in whorls (arrow).
The pre-operative diagnosis of paraganglioma of the cauda equina is difficult with very few clues from the clinical history that differentiate this from other causes of the cauda equina syndrome. In a recent review from the Mayo clinic [2] there was a slight male preponderance and a mean age of 51, although it has been described in a 13-year-old boy [3]. Virtually all patients present with a history of lumbar back pain with or without sciatic radiation which may often be bilateral. There is no doubt that diagnosis was delayed in this patient due to the coincidental but otherwise unrelated fall. It is impossible to say whether this patient's long history of back pain was as a result of the fall or the tumour but the radicular pain was undoubtedly due to the latter. The duration of symptoms in this case is typical of most described in the literature, varying between a few months and several years. Neurological signs are often absent (seen in only 11 of 31 cases described by Sonneland et al. [2]). Only one case has been reported to be endocrinologically functional at this site [4] and one patient is thought to have bled into the subarachnoid space [5]. Bony erosion has been described [6,7] and may give a clue on plain radiographs but until recently myelography has been the mainstay of pre-operative anatomical definition. This, however, has often proved difficult due to the presence of the tumour resulting in a 'dry' or 'traumatic tap' at myelography. MRI has significant advantages over myelography but has been described in only 3 previous cases [8-10]. The common differential diagnoses that must be considered on the basis of the MRI findings include meningioma, neurofibroma and ependymoma. The degree of paramagnetic enhancement in our case was less than might be expected with any of these differential diagnoses and the relatively hypointense signal on T2weighted images is atypical of a neurofibroma. Ependymorea often do not have such a uniform rounded appearance as did this case. Although the imaging findings were not 'classical' of any of the differential diagnoses described above, the possibility of a paraganglioma was not considered. This reflects the general lack of awareness of the condition. However, it is clear that with the current clinical techniques and imaging available it is still not possible to make a definitive diagnosis pre-operatively. Histopathological diagnosis can also be difficult. Our case like many previously published, was reported as an ependymoma on light microscopy. The tumour has also been mistaken for metastatic adenocarcinoma [11]. The main confusion with ependymoma at the light microscopic level stems from a radial perivascular pattern of cells which may look like the perivascular pseudorosettes of ependymoma and occasional papillary epithelial structures which resemble ependymal tubules. The ultrastructural and immunohistochemical appearance of paragangliomas is unique and easily distinguishes such lesions, their characteristic feature being the presence of dense core granules on electron microscopy. These have been shown to contain predominantly dopamine, adrenaline and noradrenaline [12] although others have found 5H T at higher levels [10]. Despite this, only one case has been shown to be endocrinologically functional [4]. The true significance of the correct diagnosis of a paraganglioma of the cauda equina lies in its apparent
PARAGANGLIOMAOF THE CAUDA EQUINA benign biological b e h a v i o u r c o m p a r e d with other sites. A l t h o u g h they are i n v a r i a b l y intradural, six cases [2,4, l 1,13,14] have been associated with extradural extension a n d because of their e n c a p s u l a t i o n they can usually be excised totally. D i s t a n t metastases have n o t been described b u t local recurrence can be a p r o b l e m if subtotal r e m o v a l is performed. O f the 67 cases thus far reported there have been no instances o f recurrence where complete excision was possible. There have been 10 cases where s u b t o t a l excision was performed. I n three of these no r a d i o t h e r a p y was given, one p a t i e n t died of a c o m p l i c a t i o n o f his t u m o u r [2]. I n another, a recurrence occurred 12 years later [4] a n d one p a t i e n t was lost to follow-up [2]. The r e m a i n i n g seven patients received r a d i o t h e r a p y a n d of these two have recurred at 1 year [2] and 9 years [ l l ] respectively. At present there are n o long-term follow-up d a t a on these patients and whereas total excision would a p p e a r to be a n adequate t r e a t m e n t offering cure, the a u t h o r agrees with others [7] that l o n g - t e r m o u t l o o k must be g u a r d e d with the limited follow-up reported in the literature. This is particularly i m p o r t a n t for those t u m o u r s resected subtotally with or w i t h o u t a d j u v a n t radiotherapy. In conclusion, p a r a g a n g l i o m a s in extrathecal locations have a p o o r prognosis, hence there is a need for further d o c u m e n t a t i o n of the c a u d a e q u i n a t u m o u r s in order to establish whether or n o t their b e h a v i o u r is u n i f o r m l y benign. With the present imaging available, a definitive pre-operative diagnosis is unlikely b u t this diagnosis needs to be b o r n e in m i n d in the atypical case. Acknowledgements.The author wishes to thank Mr A. E. Booth for kind permission to discusshis patient and to Dr D. J. Bealeand Dr H. L. Whitwell for advice on the radiology and histopathology respectively.
141
REFERENCES
1 Miller CA, Torack RM. Secretory ependymoma oi" the filum terminale. Acta Neuropathologica 1970;15:240 250. 2 Sonneland PRL, Scheithauer BW, LeChago J, Crawford BG, Onofrio BM. Paraganglioma of the cauda equina region. Cancer 1986;58:1720 1735. 3 Softer D, Pittaluga S, Caine Y, Feinsod M. Paraganglioma of the cauda equina. Cancer 1983;51:1907 1910. 4 Boker D-K, Wassman H, SolymosisL. Paragangliomas of the spinal canal. Surgical Neurology 1983;19:461 486. 5 Houroupian D, Kerson LA, Saiontz H, VatsamisM. Paraganglioma ofcauda equina. Cancer 1974;33:1337-1348. 6 Ironside JW, Royds JA, Taylor CB, TimperleyWR. Paraganglioma of the cauda equina: A histological, ultrastructural and immunocytochemical study of two cases with a reviewof the literature. Journal of Pathology 1985;145:195 201. 7 O'Sullivan MG, Keohane C, Buckley TF. Paraganglioma of the cauda equina: Case report and review of the literature. British Journal of Neurosurgery 1990;4:63 67. 8 Nagata K, Kakizoe M, Takagi H, Ohashi T, Yamamoto S, Inoue A, et al. Paraganglioma of the cauda equina. Kurume Medical Journal 1988;35:95 99. 9 Hayes E, Lippa C, Davidson R. Paragangliomas of the cauda equina. American Journal of Neuroradiology 1989;10:$45 $47. 10 Piggot TJD, Lowe JS, Morrel K, Kerslake RW. Paraganglioma of the cauda equina. Report of three cases. Journal of Neurosurgery 1990;73:455 458. 11 Taxy JB. Paraganglioma of the cauda equina. Cancer 1983;51:19041906. I2 Llena JF, Wisoff HS, Hirano A. Gangliocytic paraganglioma in cauda equina region with biochemical and neuropathological studies. Journal of Neurosurgery 1982;56:280-282. I3 Lagace R, Delage C, Gagne F. Paraganglioma of the filum terminale. Canadian Journal of Neurological Science 1978;5:257 260. 14 Gaffney EF, Doorly T, Dinn JJ. Aggressive oncocytic neuroendocrine tumour ('oncocytic paraganglioma') of the cauda equina. Histopathology 1986;10:311 319.
