81 M.A. Hayat (ed.), Tumors of the Central Nervous System, Volume 6: Spinal Tumors (Part 1),
Tumors of the Central Nervous System 6, DOI 10.1007/978-94-007-2866-0_11,
© Springer Science+Business Media B.V. 2012
11
Abstract
Chordomas are challenging tumours which usually occur in the axial skeleton, and are particularly diffi cult to manage when they are situated in the clivus and upper cervical spine due to the proximity of eloquent struc- tures. They also have a tendency for local recurrence since enbloc resec- tion is rarely possible in these locations. They have characteristic radiological appearances on MRI, and pathologically are characterised by the presence of Physallipherous cells. The mainstay of treatment is maxi- mal tumour excision at the time of fi rst presentation, followed by high dose radiation treatment such as proton beam irradiation.
Keywords
Chordoma • Cervical spine • Clivus • Tumor • Surgery
Introduction
Chordomas were probably fi rst described by Virchow in 1857 (Crockard 1985 ) . They are tumours which develop from rests of embryonal notochord remnants, and are usually found in the axial skeleton, most commonly in the clivus and sacrum, accounting for 1–8% of primary malig- nant bone tumours and 20% of spinal primary tumours (Crockard and Johnston 1993 ) . The inci- dence in Caucasians is 0.5 per million, and they typically occur in the 30–50 year old age group,
with a roughly equal sex distribution, whereas spinal chordomas tend to occur slightly later, and more commonly in males (James and Crockard 1991 ) . Fifty percent of chordomas occur in the sacrum, 35% in the skull base, and 15% in other parts of the spine.
The bones of the skull base are pre-formed in cartilage and then undergo enchondral ossifi ca- tion, although the orbital roof and greater wings of the sphenoid form from intramembraneous ossifi cation. Chordomas arise from notochordal remnants within the bones of the skull base and spine. In the human, the notochord is a longitudi- nal structure that develops from day 17 and degenerates by week 10, although remnants of it persist in the nucleus pulposus of the cartilagi- nous discs. Its role is to control segmentation of
D. Choi (*)
National Hospital for Neurology and Neurosurgery , Queen Square , London , UK
e-mail: [email protected]
Chordomas of the Clivus
the developing embryo and formation of the vertebral column, signalled via molecules such as the protein product of a gene called “Sonic Hedgehog”, which induces differentiation in the ventral and lateral neural tube, but by week 10 in utero this phenomenon is complete and the notochord is not longer needed.
Some clival chordomas are discovered inci- dentally after a routine MRI. They are slow grow- ing and locally invasive, resulting in symptoms and signs when neighbouring structures are affected, often presenting late. Most patients present with pain due to instability of the cranio- cervical junction, or vague pain due to infi ltration of the clivus and irritation of the dura. Neck pain, headache, occipital neuralgia and mechanical pain are typical of chordomas extending into the C1/2 level. If spinal cord compression occurs then patients may experience paraesthesia, numb- ness, weakness, L’Hermitte’s symptoms, and demonstrate hyperefl exia, clonus, and extensor plantar responses. Compression of the brainstem may produce dizziness, vertigo, syncope, nystag- mus, brainstem vascular syndromes, dysarthria, dysphonia, dysphagia, diplopia, facial numbness, sleep-apnoea and nightmares, as well as cranial nerve palsies with long tract signs and respiratory diffi culties depending on the degree of compres- sion caused by the tumour (Colli and Al-Mefty 2001 ; Tzortzidis et al. 2006 ) . Foramen magnum lesions may involve the lower cranial nerves (IX, X, XI, XII) by direct invasion of nerves, or neuropraxia due to compression.
Radiology
Chordomas of the clivus are usually midline tumours, conforming to the position of the degen- erated notochord in the embryo, seen as an isodense or hypodense soft tissue mass on CT.
CT scans demonstrate the degree of bone inva- sion, but do not clearly show the extent of neuro- logical involvement. MRI visualises the soft tissue tumour and nervous structures more clearly and forms an essential part of investigations:
T1-weighted MRI shows isointense regions of tumour, often heterogeneous due to cystic
and haemorrhagic areas, and interspersed areas of bone involved by tumour (Fig. 11.1a ).
T2-weighted images show hyperintensity of the tumour, and it is often easier to see the extent of the tumour on T2 imaging (Fig. 11.1b ). X-rays are generally less helpful in the diagnosis of chordomas, but are useful for post-operative assessment of fi xation (Fig. 11.1c ).
Pathology
Conventional chordomas are soft grey tumours, well demarcated from adjacent tissues, although often invading into adjacent bone and soft tissues.
Histological examination reveals typical neoplas- tic cells with bubbly cytoplasmic vacuoles, known as “physalipherous” cells (Fig. 11.2 ). A dedifferentiated type of chordoma, with a high proportion of mitotically active cells, is a rare form which resembles and behaves more like a high-grade sarcoma.
Chordomas can occasionally metastasise to other parts of the body, although it is more com- mon to see chordomas affecting neighbouring tis- sues, sometimes with “skips” across apparently normal intervening tissue. Most chordomas are slowly growing and locally infi ltrative, with a high rate of local recurrence and seeding after surgery.
Treatment
The best treatment for chordomas is surgical resection with clear margins where possible, however usually the eloquent neighbouring struc- tures of the skull base do not allow en-bloc resec- tion. Surgery often requires a midline transoral or related approach (Rock et al. 1993 ) .
Proton-beam irradiation is recommended after surgery, although other radiation modalities including intensity modulated radiotherapy, gamma knife radiosurgery, and cyberknife radia- tion are also useful, although outcomes data is awaited (Gay et al. 1995 ; Foweraker et al. 2007 ) . Standard fractionated radiotherapy is not as effec- tive, since chordomas generally do not respond to lower dose modalities of radiation treatment.
Protons are produced by a cyclotron and exhibit the Bragg peak effect: most of the energy is released after travelling a specifi c distance into the tissue, giving a greater therapeutic effect to the tumour compared to surrounding tissues.
Other isotopes such as heavy carbon beams may also be used, but these types of therapy are only available in a few research institutions around the world.
Transoral Surgical Technique
The modern day transoral operation was devel- oped by Greenberg et al. ( 1968 ) , Grison ( 1967 ) , Menezes et al. ( 1980 ) , and more recently, Crockard ( 1985 ) .
The transoral operation uses a modifi ed otolaryngology retractor to hold the mouth and
Fig. 11.1 ( a ) T1 weighted sagittal MRI scan of a clival chordoma ( b ) T2 weighted sagittal MRI scan of a clival chordoma, showing the extend of tumour and more clearly
than T1 weighted imaging or CT scans. ( c ) X-ray showing post-operative occipitocervical fi xation
tongue open, allowing visualisation of the poste- rior oropharynx and soft palate (Fig. 11.3 ). The soft palate may be divided in the midline if required to allow access superiorly to the lower clivus, whereas tongue retraction allows inferior visualisation down to the C2 vertebral body.
If greater superior exposure is required, then
transnasal endoscopy or maxillotomy can be performed, whereas inferior extension down to the C3/4 disc level may be achieved by splitting the mandible and tongue in the midline.
Transoral surgery should not be performed in the following circumstances: if there is active oral sepsis, mouth opening is limited to less than 25 mm, or if there is a fi xed-fl exion deformity of the neck. If tumours extend further than 1.5 cm from the midline, an additional or alternative lateral approach may be required. These operations require a skilled anaesthetic and the- atre team also, together with specialised aftercare on the intensive care unit.
In the operating room, the patient is positioned supine on the operating table, with head fi rmly fi xed in a 3-pin head rest and extended approxi- mately 15° (more extension allows a greater view of the clivus, but one should beware not to over- extend the neck in patients with pre-existing neurological compression). For a right-handed surgeon, the best position is to stand on the patient’s right side, with surgical assistant on the patient’s left side, and operating room nurse at the head end of the table, with the anaesthesiol- ogy team at the foot end of the operating table.
It is better to use a dedicated transoral retractor system such as the Crockard transoral set (Codman) which allows easier and safer access.
This comprises an oral and tongue retractor, retractors for the soft palate and endotracheal/
nasogastric tubes, a long monopolar diathermy blade, and appropriate bayoneted forceps and dissectors of suffi cient length. An operating microscope, high speed surgical drill, and frame- less stereotactic equipment are also useful.
The mouth is cleaned with an aqueous solu- tion of 0.5% chlorhexidine and local anaesthesia (1% lidocaine with 1:200,000 adrenaline) is injected submucosally in the midline at the back of the pharynx prior to making an incision in the pharyngeal mucosa. The mucosa is retracted with a pharyngeal self-retaining retractor which is also used to retract the deeper longus colli and capitis muscles after they are divided by a vertical mid- line incision. The tumour is removed with dissec- tion using various dissection instruments, suction
Fig. 11.2 Haematoxylin and Eosin stained histology of a chordoma, showing the typical physalipherous cells ( arrows )
Fig. 11.3 Schematic diagram of the transoral approach to midline tumours
and ultrasonic aspiration. The dura should never be breached unless there is already intradural spread of tumour, in order to minimize the chances of tumour seeding and recurrence.
Complications of transoral surgery include infection, haematoma, dysphagia and incompe- tence of the soft palate which produces nasal regurgitation. Occasionally neurological damage may occur to the cranial nerves and brainstem.
It is sometimes necessary to fuse the craniocervical junction by a posterior occipitocervical fusion and instrumentation, which can be done on the same day as the tumour resection, or as a staged procedure on a separate occasion.
Prognosis
Prognosis is infl uenced mainly by the tumour size and surgical resectability. Pallini et al. ( 2003 ) reported a 5 year recurrence-free survival of 65%
after surgery and radiotherapy.
Analysis of survival in 97 patients revealed two distinct prognostic groups: those with a higher morbidity and mortality, early recurrence and death; and those with long-term survival and late recurrence (Choi et al. 2010 ) . Watkins et al.
( 1993 ) reviewed an earlier series of 38 patients who received surgery between 1958 and 1988, which also demonstrated a group of patients who died within 5 years of surgery, and another group with a better life expectancy.
A study of tumour gene expression in a smaller group of patients revealed that there was a greater chance of tumour recurrence in patients with an increased expression of human telom- erase reverse transcriptase mRNA and mutation of p53 protein (Pallini et al. 2003 ) . A high Ki67 proliferation index (greater than 6%) can be associated with a poor prognosis, but low Ki67 proliferation index correlates less reliably with a good prognosis, probably due to sampling error and heterogeneous expression throughout the tumour (Holton et al. 2000 ) .
Surgery remains the mainstay of treatment, and without treatment the mean survival can be as low as 1 year (Eriksson et al. 1981 ) . Maximal tumour
resection at fi rst presentation is the goal, as suggested by our data (Fig. 11.4 ) in which better survival was seen in patients who underwent radi- cal surgery from the outset, compared to patients who presented to our unit and received surgery for a recurrence after suboptimal primary resec- tion (Choi et al. 2010 ) .
Carpentier et al. in 2002 supported this fi nding in their series of 22 patients who underwent pri- mary surgery in Paris, and 14 who were treated for recurrence after primary surgery. There was a signifi cantly worse outcome in the latter group (log rank test, p = 0.049), supporting the role for aggressive surgery at fi rst presentation.
Tzortzidis et al. ( 2006 ) studied 74 patients, but more commonly performed lateral and tran- scranial approaches, rather than transoral surgi- cal approaches described above. This may be due partly to differences in the location of the chordoma, and partly due to surgeon’s prefer- ence for one particular technical approach.
The overall mortality and complication rates of these approaches are similar when comparing the outcomes of larger series (Gay et al. 1995 ; Colli and Al-Mefty 2001 ; Tzortzidis et al. 2006 ; Choi et al. 2010 ) .
Fig. 11.4 Survival of patients after surgery and radio- therapy, comparing patients who received primary surgery at a specialist centre, and those who were operated else- where and subsequently referred for revision surgery.
(Primary operation in specialist centre – black . Surgery for recurrence – light grey )
Operative Complication Rates
Colli and Al-Mefty ( 2001 ) achieved complete excision in almost 50% of their patients, and greater than 90% resection in 78%, but with a surgical complication rate of 60%, mainly neurological complications, but also 7.9% inci- dence of CSF leakage, 4.8% hydrocephalus, 3.2% meningitis, and 3.2% oronasal fi stulae.
They found that there was a correlation of better survival with excision of more than 90% of tumour.
Gay et al. ( 1995 ) had a CSF leakage rate of 30%, but this may have been due to more exten- sive tumour from the outset, already penetrating the dura. They documented a 6.7% mortality rate in their series.
In their series of 36 patients, Carpentier et al.
( 2002 ) described three post-op deaths, three CSF leaks, and one case of meningitis, whereas Pallini et al. had 17 signifi cant complications in 26 patients including three CSF fi stulae, and Harbour et al. ( 1991 ) documented three post-operative deaths in 11 patients.
Survival
Choi et al. ( 2010 ) found 5 and 10 year survival times of 55% and 36% respectively. Forsyth et al.
( 1993 ) previously had very similar 5 and 10 year overall survivals of 51% and 35% after surgery, with disease free survivals of 33% and 24%, while Carpentier et al. ( 2002 ) found a signifi cant difference between primary surgery cases (80%
and 65% survivals) and in cases of recurrent tumour (50% and 0% respectively).
Colli and al-Mefty ( 2001 ) documented a longer 5 year survival of 85.9%, but their series included chondroid chordomas which have a better overall prognosis compared to standard chordomas and have often been misdiagnosed in the past. The diagnosis of chordoma should be confi rmed by immunohistochemical labelling of antibodies for S100, cytokeratin, epithelial membrane antigen, carcinoembryonic antigen, and vimentin (Watkins et al. 1993 ) .
Carpentier et al. ( 2002 ) looked at outcome of patients with and without proton beam irradiation after surgery, and in their series they did not fi nd a signifi cant difference in survival between the two groups. Other groups have demonstrated some survival advantage with proton beam therapy, but outcome is probably more dependent on the extent of resection and tumour type (Hug et al.
2000 ) . Proton beam therapy probably improves outcome signifi cantly compared to other radia- tion modalities, but defi nitive evidence is awaited (Foweraker et al. 2007 ; Hug et al. 2000 ) .
References
Carpentier A, Polivka M, Blanquet A, Lot G, George B (2002) Suboccipital and cervical chordomas: the value of aggressive treatment at fi rst presentation of the dis- ease. J Neurosurg 97:1070–1077
Choi D, Melcher R, Harms J, Crockard A (2010) Outcome of 132 operations in 97 patients with chordomas of the craniocervical junction and upper cervical spine.
Neurosurgery 61(1):59–65
Colli BO, Al-Mefty O (2001) Chordomas of the skull base:
follow-up review and prognostic factors. Neurosurgery 95:933–943
Crockard HA (1985) The transoral approach to the base of the brain and upper cervical cord. Ann R Coll Surg Engl 67:321–325
Crockard HA, Johnston F (1993) Development of tran- soral approaches to lesions of the skull base and cran- iocervical junction. Neurosurg Q 3:61–82
Eriksson B, Gunterberg B, Kindblom LG (1981) Chordoma:
a clinicopathological and prognostic study of a Swedish National series. Acta Orthop Scand 52:49–58 Forsyth PA, Cascino TL, Shaw EG, Scheithauer BW,
O’Fallon JR, Dozier JC, Piepgras DG (1993) Intracranial chordomas: a clinicopathological and prognostic study of 51 cases. J Neurosurg 78:741–747 Foweraker KL, Burton KE, Maynard SE, Jena R, Jefferies SJ, Laing RJC, Burnet NG (2007) High-dose radio- therapy in the management of chordoma and chond- rosarcoma of the skull base and cervical spine: part 1- clinical outcomes. Clin Oncol 19:509–516 Gay E, Sekhar LN, Rubinstein E, Wright DC, Sen C,
Janecka IP, Snyderman CH (1995) Chordomas and chondrosarcomas of the cranial base: results and fol- low-up of 60 patients. Neurosurgery 36:887–897 Greenberg AD, Scoville WB, Davey LM (1968) Transoral
decompression of atlanto-axial dislocation due to odontoid hypoplasia. Report of two cases. J Neurosurg 28:266–269
Grison C (1967) Direct surgical approach by oral route to the fi rst 2 cervical vertebrae. J Fr Otorhinolaryngol Audiophonol Chir Maxillofac 16:271–273
Harbour JW, Lawton MT, Criscuolo GR, Holliday MJ, Mattox DE, Long DM (1991) Clivus chordoma: a report of 12 recent cases and review of the literature.
Skull Base Surg 1:200–206
Holton JL, Steel T, Luxsuwong M, Crockard HA, Revesz T (2000) Skull base chordomas: correlation of tumour doubling time with age, mitosis and Ki67 proliferation index. Neuropathol Appl Neurobiol 26:497–503
Hug EB, Loredo LN, Slater JD, DeVries A, Grove RI, Schaefer RA, Rosenberg AE, Slater JM (2000) Proton radiation therapy for chordomas and chondrosarcomas of the skull base. J Neurosurg 91:432–439
James D, Crockard HA (1991) Surgical access to the base of the skull and upper cervical spine by extended max- illotomy. Neurosurgery 29:411–416
Menezes AH, Van Gilder JC, Graf CJ, McDonnell DE (1980) Craniocervical abnormalities. A comprehensive surgical approach. J Neurosurg 53:444–455
Pallini R, Maira G, Pierconti F, Falchetti ML, Alvino E, Cimino-Reale G, Fernandez E, D’Ambrosio E, Larocca LM (2003) Chordoma of the skull base: predictors of tumor recurrence. J Neurosurg 98:1256–1262 Rock PJ, Tomecek FJ, Ross L (1993) Transoral surgery:
an anatomic study. Skull Base Surg 3(3):109–116 Tzortzidis F, Elahi F, Wright D, Natarajan SK, Sekhar LN
(2006) Patient outcome at long-term follow-up after aggressive microsurgical resection of cranial base chordomas. Neurosurgery 59:230–237
Watkins L, Khudados ES, Kaleoglu M, Revesz T, Sacares P, Crockard HA (1993) Skull base chordomas: a review of 38 patients, 1958-88. Br J Neurosurg 7:241–248
Imaging
91 M.A. Hayat (ed.), Tumors of the Central Nervous System, Volume 6: Spinal Tumors (Part 1),
Tumors of the Central Nervous System 6, DOI 10.1007/978-94-007-2866-0_12,
© Springer Science+Business Media B.V. 2012