RESEARCHHUMANCLINICAL STUDIES

RESEARCHHUMANCLINICAL STUDIES

Arman Jahangiri, BS*

Aaron T. Chin, BS*
Jeffrey R. Wagner, BS
Sandeep Kunwar, MD
Christopher Ames, MD
Dean Chou, MD
Igor Barani, MD
Andrew T. Parsa, MD, PhDk
Michael W. McDermott, MD

Factors Predicting Recurrence After Resection of

Clival Chordoma Using Variable Surgical
Approaches and Radiation Modalities
BACKGROUND: Clival chordomas frequently recur because of their location and invasiveness.
OBJECTIVE: To investigate clinical, operative, and anatomic factors associated with clival

chordoma recurrence.
METHODS: Retrospective review of clival chordomas treated at our center from 1993 to

Arnau Benet, MD#

2013.

Ivan H. El-Sayed, MD**

RESULTS: Fifty patients (56% male) with median age of 59 years (range, 8-76) were newly

Manish K. Aghi, MD, PhD

diagnosed with clival chordoma of mean diameter 3.3 cm (range, 1.5-6.7). Symptoms included
headaches (38%), diplopia (36%), and dysphagia (14%). Procedures included transsphenoidal
(n = 34), transoral (n = 4), craniotomy (n = 5), and staged approaches (n = 7). Gross total resection
(GTR) rate was 52%, with 83% mean volumetric reduction, values that improved over time. While
the lower third of the clivus was the least likely superoinferior zone to contain tumor (upper
third = 72%/middle third = 82%/lower third = 42%), it most frequently contained residual tumor
(upper third = 33%/middle third = 38%/lower third = 63%; P , .05). Symptom improvement
rates were 61% (diplopia) and 53% (headache). Postoperative radiation included proton beam
(n = 19), cyberknife (n = 7), intensity-modulated radiation therapy (n = 6), external beam (n = 10),
and none (n = 4). At last follow-up of 47 patients, 23 (49%) remain disease-free or have stable
residual tumor. Lower third of clivus progressed most after GTR (upper/mid/lower third = 32%/
41%/75%). In a multivariate Cox proportional hazards model, male gender (hazard ratio [HR] =
1.2/P = .03), subtotal resection (HR = 5.0/P = .02), and the preoperative presence of tumor in the
middle third (HR = 1.2/P = .02) and lower third (HR = 1.8/P = .02) of the clivus increased further
growth or regrowth, while radiation modality did not.
CONCLUSION: Our findings underscore long-standing support for GTR as reducing chordoma recurrence. The lower third of the clivus frequently harbored residual or recurrent tumor,
despite staged approaches providing mediolateral (transcranial 1 endonasal) or superoinferior (endonasal 1 transoral) breadth. There was no benefit of proton-based over photonbased radiation, contradicting conventional presumptions.

Center for Minimally Invasive Skull Base

Surgery (MISB), University of California at San
Francisco, San Francisco, California; Department of Neurosurgery, Center for Minimally
Invasive Skull Base Surgery (MISB), University of
California at San Francisco, San Francisco,
California; Department of Radiation Oncology,
University of California at San Francisco, San
Francisco, California; kDepartment of Neurosurgery, Northwestern University, Chicago,
Illinois; #Skull Base and Cerebrovascular Laboratory, University of California at San Francisco,
San Francisco, California; **Department of
Otolaryngology, University of California at San
Francisco, San Francisco, California
*These authors contributed equally to this
work.
Correspondence:
Manish K. Aghi, MD, PhD,
Associate Professor of Neurological Surgery,
Neurosurgical Director,
Center for Minimally Invasive Skull Base
Surgery (MISB),
University of California at San Francisco (UCSF),
505 Parnassus Ave, Rm M779,
San Francisco, CA.
E-mail: AghiM@neurosurg.ucsf.edu
Received, June 26, 2014.
Accepted, October 3, 2014.
Published Online, December 29, 2014.
Copyright 2014 by the
Congress of Neurological Surgeons.

KEY WORDS: Chordoma, Cyberknife, Endoscopic, Proton beam, Recurrence

Neurosurgery 76:179186, 2015

DOI: 10.1227/NEU.0000000000000611

hordomas are extra-axial tumors originating

from embryologic remnants of the notochord. This leads to a tendency for occurrence at extreme ends of the vertebral axis, with 40%
ABBREVIATIONS: GTR, gross total resection; IMRT,
intensity-modulated radiation therapy; STR, subtotal resection; UCSF, University of California, San
Francisco
Supplemental digital content is available for this article.
Direct URL citations appear in the printed text and are
provided in the HTML and PDF versions of this article on
the journals Web site (www.neurosurgery-online.com).

NEUROSURGERY

www.neurosurgery-online.com

occurring at the clivus or cervical spine and 60%

occurring in the sacrococcygeal junction of the
spine.1,2 Within the clivus or cervical spine, most
chordomas involve the clivus with occasional caudal
extension to involve C1 or C2, although chordomas
isolated to the midcervical spine occur as well.3
Although chordomas are generally slow growing
and are histologically considered low-grade tumors,
their high recurrence rate, even after postoperative
radiation, renders them difficult to treat. This is
particularly true for clival chordomas, whose deep
anatomic location and proximity to vital anatomic
structures makes surgical resection challenging.4,5

VOLUME 76 | NUMBER 2 | FEBRUARY 2015 | 179

Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

JAHANGIRI ET AL

Surgical approaches for clival chordoma described in the

literature include craniotomies, transoral6 and endoscopic endonasal approaches.5,7 Recent reports indicate that complete removal
of clival chordomas provides the best long-term survival; however,
controversy remains regarding the optimal surgical approaches, the
associated complications, and the progression-free or overall
survival benefit achieved by aggressive resection. The role of
adjuvant radiation and the type of radiation that should be used
also remain a subject of discussion; several investigators advocate
proton beam radiation, because its sharp dosimetry decline at the
tumor edge minimizes the radiation dose to which the surrounding
normal structures are subjected.8-10
Here, we retrospectively review 50 newly diagnosed clival
chordomas managed at our center through a variety of surgical
approaches, including staged approaches, and several postoperative radiation modalities to better identify factors influencing
progression-free survival.

METHODS
Study Design and Population
We retrospectively reviewed University of California, San Francisco
(UCSF) department of neurological surgerys database and identified 50
examples of patients undergoing initial resection of clival chordoma
between 1993 and 2013. The UCSF Committee on Human Research
approved this study in advance.

Parameters Assessed
Original pathology and operative reports were reviewed to confirm the
diagnosis of chordoma and the surgical approach used. The age at diagnosis
was documented as the patients age at the time of the initial operation. All
parameters were reviewed separately by 2 members of the study team.

Imaging
Preoperative and postoperative MRIs were reviewed by the study team
to assess (1) tumor size, which was recorded at its longest in all 3
dimensions (anteroposterior, left to right, and superoinferior) along with
the average of the 3 dimensions; (2) tumor volume, measured using
AWServer software (GE Healthcare; San Francisco, California) to outline
areas of the tumor on sequential sagittal images with each volume
calculation repeated by 2 members of the study team to confirm
concordance; (3) superoinferior zones of involvement; and (4) the
presence of tumor medial/lateral to the carotid artery. Superoinferior
zones of involvement were classified as (1) upper third of the clivus
from dorsum sellae and posterior clinoids to the plane of Dorello canal,
including the paraclinoidal and intracavernous segments of the carotid
artery, the interpeduncular cistern, the third cranial nerve, the basilar
apex, the superior cerebellar artery, and posterior cerebral artery; (2)
middle third of clivusfrom the petrous apex at Dorello canal to the
pars nervosa of the jugular foramen, including the paraclival carotid
artery, the sixth cranial nerve, the basilar trunk, the prepontine cistern,
and Meckel cave out laterally; and (3) lower third of clivusfrom
the pars nervosa to the foramen magnum, including down to the
parapharyngeal carotid segment, the hypoglossal canal, vertebral
arteries, and premedullary cistern. Postoperative MRIs were retrospectively reviewed by the study team until the last known MRI or the first

180 | VOLUME 76 | NUMBER 2 | FEBRUARY 2015

MRI showing tumor recurrence or progression of residual disease. Areas

of recurrence or progression of residual disease were classified into the
anatomic zones described above.

Cadaveric Analysis
Two human postmortem specimens were used to illustrate the
anatomic location of the residual/recurrent chordomas in our series.
The specimens were prepared for surgical simulation as previously
described by our group.11 An extended endoscopic endonasal approach
to the skull base and a transcranial far lateral approach were performed
and recorded.

Statistical Analysis
Kaplan-Meier analysis was used to assess and compare actuarial
radiographic recurrence rates in the group as a whole and in the various
patient subcategories, with time until radiographic recurrence or time
until last MRI showing no radiographic recurrence recorded for each
patient. Analysis of variance was used for parametric comparisons of more
than 2 variables when the dependent variable was continuous while a x2
test was used to compare more than 2 proportions. The Student t test
was used to perform parametric comparison between 2 variables. The
Fisher exact test was used to compare 2 proportions. P values are 2-tailed
and P , .05 was considered statistically significant.

RESULTS
Patient Population
Fifty patients (56% male) with a median age of 59 years (range,
8-76) were newly diagnosed with clival chordoma with mean
diameter 3.3 cm (range = 1.5-6.7). Presenting symptoms included
headaches (38%), diplopia (36%), and dysphagia (14%).
Preoperative Tumor Characteristics
Of the 39 tumors with digitized MRIs that could undergo
volumetric analysis, mean preoperative tumor volume was 18.4
cm3. Review of images of all 50 cases revealed that 72% occupied
the upper third of the clivus, 82% occupied the mid third of the
clivus, and 42% occupied the lower third of the clivus (Figure
1A). A more detailed radiographic review of specific structures in
and around the upper, mid, and lower third of the clivus revealed
the prepontine cistern behind the mid third of the clivus to be the
most involved area, at 66% of the cases, followed by the dorsum
sellae and posterior clinoids in the upper third of the clivus, which
were involved in 60% of the cases (Figure 1B).
Surgical Procedures Performed as First Treatment
For 43 patients undergoing a single surgical procedure, the most
common approach was endonasal transsphenoidal, performed in
34 patients (79%) by using either a microscope (n = 15) or
endoscope (n = 19). Four tumors (9%) were accessed exclusively
through transoral approaches owing to significant involvement of
the lower third of the clivus. Five tumors (12%) were resected
exclusively through a craniotomy (far lateral/suboccipital n = 2,
frontotemporal/orbitozygomatic n = 2, and middle fossa trans- or

www.neurosurgery-online.com

Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

FACTORS INFLUENCING CLIVAL CHORDOMA RECURRENCE

FIGURE 1. Locations of clival chordomas and representative MRIs. A, percentage of cases located in the upper, middle, and lower third of the clivus. Tumors were less
frequently located in the lower third of the clivus than in other locations (P , .05). B, percentage of cases located in the following specific structures found within or adjacent to
(lateral or posterior to) each third of the clivus: Upper third(1) dorsum sellae and posterior clinoids; (2) interpeduncular cistern; (3) basilar apex; (4) paraclinoidal/
intracavernous carotid; (5) cranial nerve III; (6) Dorello canal; and (7) posterior cerebral arteries. Middle third(1) prepontine cistern; (2) basilar trunk; (3) petrous apex; (4)
Dorello canal; (5) cranial nerve VI; (6) Meckel cave; (7) jugular foramen; and (8) paraclival carotid. Lower third(1) premedullary cistern; (2) foramen magnum; (3)
parapharyngeal carotid; (4) hypoglossal canal; and (5) vertebral artery. When separating clival and paraclival structures into those that were medial (1-3 in the upper third; 1-2
and 4 in the mid third; and 1-3 in the lower third) vs lateral (4-7 in the upper third; 3 and 5-8 in the mid third; and 4-5 in the lower third), medial structures were more likely
to contain tumor (P , .05). C, sagittal and coronal illustrations of the upper, mid, and lower thirds of the clivus with key locations in each third illustrated with the same
numbering system used in B with circles in each zone the size of which correspond to the frequency with which tumor occurred in that location. D, representative images of cases
treated and approaches used. First row shows 57-year-old man with diplopia and tumor in all 3 thirds of the clivus, specifically Dorello canal, cavernous carotid arteries, petrous
apices, cranial nerve VI, basilar trunk, prepontine cistern, and hypoglossal canal. Patient underwent endoscopic endonasal surgery with GTR followed by proton beam
radiation. Recurrence of tumor in the right hypoglossal canal 2 years later was treated with reoperation and external beam radiation. Second row shows 52-year-old woman
with shoulder and neck spasms and tumor in the lower third of the clivus, specifically in the hypoglossal canal, vertebral artery, premedullary cistern, and extending out the
foramen magnum. A transoral approach with tumor resection and odontoidectomy was performed. There was trace residual tumor encasing the pharyngeal carotid arteries
which the patient elected to observe and it has not progressed to date. Third row shows a 22-year-old man with shoulder and neck pain and tumor in the lower third of the clivus,
particularly the jugular foramen, prepontine cistern, and extending out beyond the foramen magnum. A combined transsphenoidal and transoral approach was performed,
followed by a retrosigmoid craniotomy. Resection was subtotal, followed by IMRT, with recurrence at 18 months. GTR, gross total resection; IMRT, intensity-modulated
radiation therapy.

retrolabyrinthine n = 1). Seven patients (16%) underwent

combined approaches, which were done in multiple stages in 6
patients and in the same setting in 1 patient. Three of these
combined approaches targeted the superoinferior extent by either
combining transsphenoidal and transoral approaches or transoral approaches with upper cervical laminectomies; 3 targeted the
mediolateral extent through transsphenoidal surgery followed
by craniotomy, and 1 spanned both directions through 3
approaches (transsphenoidal, transoral, and craniotomy). When
assessing preoperative tumor sizes in 3 linear dimensions
(anteroposterior, transverse, and craniocaudal), tumors undergoing staged approaches were larger in all 3 dimensions than
tumors undergoing single approaches (P , .05), and tumors
undergoing craniotomy were larger in the transverse dimension

NEUROSURGERY

than tumors undergoing other single-stage procedures (P , .05)

(Figure 2A).
Surgical ResultsExtent of Resection
Gross total resection (GTR) was achieved with 52% of cases.
There was significant improvement in GTR rates over time (Figure
2B). GTR was accomplished in 2 of the 7 cases undergoing staged
approaches. Volumetric analysis revealed an average of 83%
tumor removal for all patients (range = 18%-100%). Mean
percentages of tumor removed by approach were 57% (transoral),
88% (endonasal), and 57% (craniotomy) (P , .05) (Figure 2C).
Staged approaches allowed 60% tumor removal after the first
procedure, and an additional 30% of the original tumor volume
was removed after the second procedure for a total average of

VOLUME 76 | NUMBER 2 | FEBRUARY 2015 | 181

Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

JAHANGIRI ET AL

FIGURE 2. Preoperative and postoperative chordoma sizes associated with

different types of surgical procedures. A, maximal diameter in cm in all 3
dimensions (AP, anteroposterior; TR, transverse left to right; CC, craniocaudal)
on preoperative MRI. B, percent of cases with GTR during 4 sequential time
intervals: 1993 to 1997, 1998 to 2004, 2005 to 2010, and 2011 to 2013.
There was increased GTR during each interval (P , .05). C, mean tumor
volumes in cm3 before and after different types of surgical procedures, including
before and after first and second stage for 2 staged procedures. Error bars
represent standard errors of the mean. GTR, gross total resection; op, preoperative; preop, preoperative.

90% tumor removal, higher than achieved after the first stage
alone (P , .05) (Figure 2C).
Sites of Residual Tumor
Although the lower third of the clivus was the least likely
superoinferior clival zone to contain tumor (upper third = 72%,
middle third = 82%, lower third = 42%; Figure 1A-D), the rate
with which residual tumor occurred was highest in the lower
third of the clivus (upper third = 33%, middle third = 38%, lower
third = 63%; P , .05) (Figure 3A). More medial structures were
more likely to contain tumor than more lateral structures in and
around the clivus (P , .05; Figure 1B-C), consistent with the
medial origin of chordoma. Twenty-two chordomas extended
lateral to the carotid artery (44%), of which 9 (41%) received
GTR, comparable to the overall GTR rate of 52% (P . .05). Eight
of 24 residual chordomas (33%) occurred lateral to the carotid
artery, with 5 of these residual chordomas occurring exclusively

182 | VOLUME 76 | NUMBER 2 | FEBRUARY 2015

FIGURE 3. Relationship between residual and recurrent chordoma and

function of location in the mid sagittal plane and extent of resection. A, shown is
the percentage of residual chordomas for subtotally resected cases that were located
in each third of the clivus. Residual tumor tended to be located in the lower third
(P , .05). B, shown are GTR rate, recurrence rate after GTR, and recurrence
rate after STR for chordomas located in each third of the clivus. GTR rate
dropped as tumor location progressed from upper to middle to lower third of the
clivus (P , .05). Recurrence after GTR was more common in the lower third of
the clivus (P , .05), whereas there was no significant trend for recurrence after
STR (P . .05). GTR, gross total resection; STR, subtotal resection.

lateral to the carotid artery. Of 13 clival chordomas subtotally

resected endonasally, sites of residual tumor included cavernous
sinus (n = 6), lateral prepontine cistern (n = 2), occipital condyles
(n = 2), lateral dorsum sellae (n = 2), petrous apices (n = 2),
foramen magnum (n = 2), hypoglossal canal (n = 2), and encasing
supraclinoid carotid artery (n = 1) (see Figure, Supplemental
Digital Content 1, http://links.lww.com/NEU/A701). Of 4 clival
chordomas subtotally resected through craniotomy, sites of
residual tumor included maxillary or sphenoid sinus (n = 3),
cavernous sinuses (n = 3), Meckel cave (n = 2), petrous apices (n =
1), orbit (n = 1), encasing the supraclinoid artery (n = 1), prepontine
cistern (n = 1), encasing the basilar artery (n = 1), premedullary
cistern (n = 1), and C1 arch (n = 1) (see Figure, Supplemental
Digital Content 2, http://links.lww.com/NEU/A702).
Rates of Symptomatic Improvement and Morbidity
Rates of symptomatic improvement were 61% (diplopia), 53%
(headache), and 43% (dysphagia). Surgical complications

www.neurosurgery-online.com

Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

FACTORS INFLUENCING CLIVAL CHORDOMA RECURRENCE

occurring in more than 1 patient were cerebrospinal fluid leakage

in 6 cases (12%), meningitis in 6 cases (12%), transient cranial
neuropathies in 3 cases (6%), and diabetes insipidus in 2 cases
(4%).

TABLE. Results of Multivariate Cox Proportional Hazards Model

Correlating Clinical Parameters for Chordomas With Their Risk of
Recurrence After GTR or Radiographic Progression After STRa
Variable

Radiation Modalities Used

Modalities of postoperative adjuvant radiation therapy
included fractionated proton beam therapy delivered over 2
months (n = 19), cyberknife delivered over 5 sessions (n = 7),
intensity-modulated radiation therapy (IMRT) over 35 fractions (n = 6), and external beam radiation therapy (n = 10). Four
patients did not receive radiation: 3 after GTR and 1 after near
GTR.
Recurrence Rates
Follow-up imaging after the immediate postoperative MRI was
retrievable and reviewable by the study team for 47 of 50 patients.
For these patients, the median time from diagnosis to last followup MRI was 41 months. Of these patients, 24 (51%) had tumor
recurrence or progression of residual tumor. In these 24 patients
who had tumor recurrence or progression of residual tumor, the
median time to recurrence or progression of residual tumor was 21
months (range = 5-101 months). The 23 patients who did not
have a recurrence (49%) were followed by imaging for a median
of 24 months (mean, 38 months; range = 3-225 months), or
longer than the median time to recurrence in the patients who
had a recurrence. Two of 4 patients who had not received
postoperative radiation after GTR or near GTR had a recurrence
(50%), while 8 of 23 (35%) receiving postoperative radiation
after GTR had a recurrence (P . .05). All progressions occurring
after subtotal resection (STR) occurred in areas of residual tumor
(13% upper third, 53% mid third, and 60% lower third), with all
involving tumor medial to the carotid and 53% having tumor
lateral to the carotid artery. Recurrence after GTR always
occurred medial to the carotid and 40% of the time occurred
lateral to the carotid artery but preferentially occurred in the
lower clivus (30% upper third, 50% mid third, 70% lower third)
and was more likely to occur when the tumor originally involved
the lower clivus (Figure 3B). In a multivariate Cox proportional
hazards model including age, sex, preoperative tumor size,
superoinferior anatomic zone, mediolateral anatomic zone,
surgical approach, extent of resection, and type of adjuvant
radiation used, male sex (hazard ratio [HR] = 1.2/P = .03), STR
(HR = 5.0/P = .02), and the preoperative presence of tumor in
the middle third (HR = 1.2/P = .02) and lower third (HR = 1.8/
P = .02) of the clivus increased the risk of further growth or
regrowth (Table; Figure 4).
Management of Recurrences or Progressions
Eleven recurrences were treated with reoperation alone. Seven
recurrences were treated with reoperation followed by gamma
knife radiosurgery. One recurrence was treated with gamma knife
radiosurgery alone. Five recurrences were small and continued to

NEUROSURGERY

Older age
Male sex
Approach
STR relative to GTR
Radiation modality
Size (cm diameter)
Upper third clivus
Mid third clivus
Lower third clivus
Lateral to carotid

Hazard Ratio

1.1
1.2
1.1
5.0
0.7
1.9
0.7
1.2
1.8
2.0

.06
.03
.7
.02
.4
.2
.7
.02
.02
.5

GTR, gross total resection; STR, subtotal resection.

P Values less than .05 are marked in bold.

be observed. Six deaths attributable to chordoma progression were

noted a median of 18 months (range, 2-35 months) after diagnosis
and 7 months after recurrence (range, 2-16 months).

DISCUSSION
In this series of clival chordomas, we were able to support the
importance of achieving GTR in reducing recurrence, while
stressing other risk factors for recurrence including male sex and
tumor in the middle and lower third of the clivus as other risk
factors for recurrence. The importance of GTR in reducing
recurrence is consistent with previous studies, including studies
describing exclusive utilization of the endoscopic endonasal5,12,13
or transcranial12-14 approaches. Although a cancer registry found
age and tumor size to influence overall survival of chordoma
patients,15 these variables did not influence recurrence in our
single-institution study.

FIGURE 4. Kaplan-Meier curves showing results of univariate analyses

investigating influence of extent of resection and location in the midsagittal plane
on chordoma recurrence rates. Subtotal resection (A) and tumor in the lower
third of the clivus (B) increased recurrence in univariate analyses (P , .05).
GTR, gross total resection; STR, subtotal resection.

VOLUME 76 | NUMBER 2 | FEBRUARY 2015 | 183

Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

JAHANGIRI ET AL

The increased recurrence risk associated with tumor in the mid

and lower third of the clivus in our multivariate analysis and the
increased presence of residual tumor and recurrence after GTR in
the lower third of the clivus is consistent with a recent study
describing increased recurrence associated with tumor in the lower
third of the clivus after an exclusively endoscopic endonasal
approach.5 Our study suggests that this area remains a more
frequent site of residual tumor even when multimodal approaches
are used, including the combination of endoscopic endonasal and
transoral approaches. Our finding that progression after STR
occurred in areas of residual tumor further supports the
importance of GTR for chordoma, and also suggests that our
finding of increased recurrence in the lower third of the clivus
after GTR may reflect this area being prone to small tumor
remnant not detectable by MRI.
The use of multiple approaches in a staged or simultaneous
fashion was used for 14% of our cases. Chordomas undergoing
staged procedures tended to be larger than those undergoing single
procedures. Despite the increased size of chordomas undergoing
staged procedures, the use of such staged procedures enabled GTR
in 29% of these cases that would have otherwise undergone STR,
with the remainder undergoing a level of additional tumor
resection during the second stage that we were able to quantify
by volumetric analysis. This level of resection achieved after
completing staged procedures was significantly higher than that
achieved after the first stage alone. Other series have made more
frequent use of multimodal surgical approaches for clival chordomas than we report here, but they did not specifically state the
frequency with which the multimodal approach enabled
GTR.3,9,14,16
The choice of combined approaches depends on the specific
plane being addressed. For superoinferior extension in the mid
sagittal plane, we found a transoral approach to provide valuable
supplementation of the endoscopic endonasal approach. We have
previously described this combined endoscope-assisted transnasal
and transoral approach for other pathologies of the craniovertebral
junction.17 Although the nasopalatal line, a line from the nasal
bone to the posterior edge of the hard palate, has been
conventionally regarded as the lower extent of resection achievable by the use of an endoscopic endonasal approach, we recently
reported that the palatal line, a line drawn parallel to and along
the floor of the hard palate, can better guide the choice between
purely endonasal, combined endonasal and transoral, and purely
transoral approaches.18
Our finding that the type of radiation did not influence recurrence
runs contrary to the perception that proton beam radiation is the
treatment of choice for chordomas. However, this perception has
been born out of small series10 published since the efficacy of proton
beam radiation for chordoma was first reported in 1989,19 whereas
other smaller series have supported our finding that the type of
radiation did not impact progression-free survival.3,20 Furthermore,
other studies have found that there are dosimetric advantages to
using IMRT or combining proton beam with IMRT compared
with proton beam alone,21-23 and a role for gamma knife

184 | VOLUME 76 | NUMBER 2 | FEBRUARY 2015

radiosurgery as adjuvant treatment for chordoma has recently been

suggested from work done by 6 participating centers in the North
American Gamma Knife Consortium.24
Limitations
There are limitations to our study. As with most retrospective
studies, the lack of prospective enrollment creates a selection bias
and limits the ability to mandate established postoperative imaging
protocols for imaging, ultimately reducing median length of
follow-up. Verification of the generalizability of our findings may
ultimately require a large prospective multi-institutional study
with central radiology review and sufficient sample size to be
empowered to address the issues we have raised, particularly the
role of radiation modality in preventing recurrence.

CONCLUSION
Our findings underscore long-standing support for GTR as
reducing chordoma recurrence. We also found tumor in the mid or
lower third of the clivus to increase the risk of recurrence, with the
latter location a particularly frequent site of residual tumor and
recurrence after GTR, despite incorporating staged approaches
providing medio lateral (transcranial 1 endonasal) or superoinferior (endonasal 1 transoral) breadth. Our findings also
suggest no benefit of proton-based over photon-based radiation,
contradicting conventional presumptions and underscoring the
need for randomized trials addressing radiation modality.
Disclosure
Dr Ames is a consultant for DePuy, Stryker, and Medtronic. He has received
royalties from Aesculap and Biomet Spine. He owns stock and stock options in
Doctors Research Group, and Baxano Surgery, and he owns a patent with Fish and
Richardson, P.C. Dr Chou is a consultant for Globus, Medtronic, and Orthofix.
The other authors have no personal, financial, or institutional interest in any of the
drugs, materials, or devices described in this article.

REFERENCES
1. Eriksson B, Gunterberg B, Kindblom LG. Chordoma. A clinicopathologic and
prognostic study of a Swedish national series. Acta Orthop Scand. 1981;52(1):
49-58.
2. McMaster ML, Goldstein AM, Bromley CM, Ishibe N, Parry DM. Chordoma:
incidence and survival patterns in the United States, 1973-1995. Cancer Causes
Control. 2001;12(1):1-11.
3. Sen C, Triana AI, Berglind N, Godbold J, Shrivastava RK. Clival chordomas:
clinical management, results, and complications in 71 patients. J Neurosurg. 2010;
113(5):1059-1071.
4. Fernandez-Miranda JC, Gardner PA, Snyderman CH, et al. Clival chordomas:
a pathological, surgical, and radiotherapeutic review. Head Neck. 2014;36(6):
892-906.
5. Koutourousiou M, Gardner PA, Tormenti MJ, et al. Endoscopic endonasal
approach for resection of cranial base chordomas: outcomes and learning curve.
Neurosurgery. 2012;71(3):614-624; discussion 624-625.
6. Singh H, Harrop J, Schiffmacher P, Rosen M, Evans J. Ventral surgical
approaches to craniovertebral junction chordomas. Neurosurgery. 2010;66
(3 suppl):96-103.
7. Dehdashti AR, Karabatsou K, Ganna A, Witterick I, Gentili F. Expanded
endoscopic endonasal approach for treatment of clival chordomas: early results
in 12 patients. Neurosurgery. 2008;63(2):299-307; discussion 307-309.

www.neurosurgery-online.com

Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

FACTORS INFLUENCING CLIVAL CHORDOMA RECURRENCE

8. Jahangiri A, Jian B, Miller L, El-Sayed IH, Aghi MK. Skull base chordomas:
clinical features, prognostic factors, and therapeutics. Neurosurg Clin N Am. 2013;
24(1):79-88.
9. Yasuda M, Bresson D, Chibbaro S, et al. Chordomas of the skull base and cervical
spine: clinical outcomes associated with a multimodal surgical resection combined
with proton-beam radiation in 40 patients. Neurosurg Rev. 2012;35(2):171-182;
discussion 182-183.
10. Fuji H, Nakasu Y, Ishida Y, et al. Feasibility of proton beam therapy for
chordoma and chondrosarcoma of the skull base. Skull Base. 2011;21(3):201206.
11. Benet A, Rincon-Torroella J, Lawton MT, Gonzlez Snchez JJ. Novel embalming
solution for neurosurgical simulation in cadavers. J Neurosurg. 2014;120(5):
1229-1237.
12. Colli B, Al-Mefty O. Chordomas of the craniocervical junction: follow-up review
and prognostic factors. J Neurosurg. 2001;95(6):933-943.
13. Sen CN, Sekhar LN, Schramm VL, Janecka IP. Chordoma and chondrosarcoma
of the cranial base: an 8-year experience. Neurosurgery. 1989;25(6):931-940;
discussion 940-941.
14. Gay E, Sekhar LN, Rubinstein E, et al. Chordomas and chondrosarcomas of the
cranial base: results and follow-up of 60 patients. Neurosurgery. 1995;36(5):887896; discussion 896-897.
15. Jones PS, Aghi MK, Muzikansky A, Shih HA, Barker FG II, Curry WT Jr.
Outcomes and patterns of care in adult skull base chordomas from the SEER
database. J Clin Neurosci. 2014;21(9):1497-1502.
16. Tzortzidis F, Elahi F, Wright D, Natarajan SK, Sekhar LN. Patient outcome at
long-term follow-up after aggressive microsurgical resection of cranial base
chordomas. Neurosurgery. 2006;59(2):230-237; discussion 230-237.
17. El-Sayed IH, Wu JC, Ames CP, Balamurali G, Mummaneni PV. Combined
transnasal and transoral endoscopic approaches to the craniovertebral junction.
J Craniovertebr Junction Spine. 2010;1(1):44-48.
18. Burke K, Benet A, Aghi MK, El-Sayed I. Impact of platybasia and anatomic
variance on surgical approaches to the craniovertebral junction. Laryngoscope.
2014;124(8):1760-1766.
19. Austin-Seymour M, Munzenrider J, Goitein M, et al. Fractionated proton
radiation therapy of chordoma and low-grade chondrosarcoma of the base of the
skull. J Neurosurg. 1989;70(1):13-17.
20. Di Maio S, Temkin N, Ramanathan D, Sekhar LN. Current comprehensive
management of cranial base chordomas: 10-year meta-analysis of observational
studies. J Neurosurg. 2011;115(6):1094-1105.
21. Torres MA, Chang EL, Mahajan A, et al. Optimal treatment planning for skull
base chordoma: photons, protons, or a combination of both? Int J Radiat Oncol
Biol Phys. 2009;74(4):1033-1039.
22. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the skull base.
Strahlenther Onkol. 1999;175(suppl 2):57-63.
23. Staab A, Rutz HP, Ares C, et al. Spot-scanning-based proton therapy for
extracranial chordoma. Int J Radiat Oncol Biol Phys. 2011;81(4):e489-e496.
24. Kano H, Iqbal FO, Sheehan J, et al. Stereotactic radiosurgery for chordoma:
a report from the North American Gamma Knife Consortium. Neurosurgery. 2011;
68(2):379-389.

Supplemental digital content is available for this article. Direct URL citations
appear in the printed text and are provided in the HTML and PDF versions of this
article on the journals Web site (www.neurosurgery-online.com).

COMMENTS

rom various publications, and from a systematic analysis of the publications, we know some facts about the treatment of chordomas, and
their recurrence (or progression). The 1 factor that has been shown to be
important in regard to recurrence is total tumor resection. So, any surgical
approach(es) which can improve this in a particular patient is important.1
In our hands, the now well-established skull base approaches work very
well to achieve total resection, either in one, or a two staged operations.
At present, in primary (previously unoperated) cases, we are able to
achieve this in better than 95% of cases, but the rate of total removal

NEUROSURGERY

drops in the case of previously operated/or irradiated patients. We do use

endonasal endoscopic approaches (especially the extended transnasal
approach) for predominantly extradural, mid clival lesions. The
limitations of this approach are in tumors which involve the cavernous
sinus extensively, encase the carotid artery, or the basilar artery, or invade
the brain stem extensively. I have also found that the extreme lateral trans
condylar approach is the best one to achieve gross total resection of lower
clivalforamen magnum-upper cervical chordomas, with a low risk of
cerebrospinal fluid leakage and meningitis.
In regards to radiotherapy, proton beam therapy has the best long term
follow up, and efficacy. Carbon ion therapy may be better than proton
beam therapy, but this has not yet been proven in long-term follow up
studies. Whether to treat smaller residual tumors with radiosurgery, or
intensity modulated radiotherapy (photon treatments) is an open question, and as yet there has not been a randomized trial that compares
radiotherapy modalities. Also, the issue of whether to irradiate patients
who have had a total tumor resection upfront is controversial, however,
our recent experience suggests that this improves survival.2
There have been many recent advances in our understanding of
genetics, and the epigenetics of chordomas. This has led to some promising drug treatments in small numbers of patients with advanced chordomas. These treatments may in future become an important adjuvant to
surgical tumor resection.3
Laligam N. Sekhar
Seattle, Washington

1. Di Maio S, Temkin N, Ramanathan D, Sekhar LN. Current comprehensive

management of chordomas: a 10 year Meta analysis of observational studies. J
Neurosurg. 2011;115(6):1094-1105.
2. Di Maio S, Rostomily R, Sekhar LN. Current surgical outcomes for cranial base
chordomas: cohort study of 95 patients. Neurosurgery. 2012;70(6):1355-1360.
3. Di Maio S, Kong E, Yip S, Rostomily R. Converging paths to progress for skull base
chordoma: review of current therapy and future molecular targets. Surg Neurol Int.
2013;4:72.

he authors present a large series of patients surgically treated for clival

chordoma. Their data support the tenet that gross total resection of
these tumors confers the greatest survival advantage for these patients and
that this should be the goal of surgical treatment. The need for, and benefit
of, staged procedures to achieve this goal is emphasized. In the authors
series the modality of radiation used did not confer any further advantage,
although the series is inadequately powered to make this conclusion. This
well-described and well-reported series emphasizes the need for surgical
teams to be facile with all surgical approaches to the clivus. At our center
the most frequently used approach for these lesions is the endonasal
endoscopic approach. It is commonly used as part of a staged procedure
along with the far lateral/transcondylar approach or the orbitocranial
approach with an expanded middle fossa extension. These added approaches deal with tumor lateral to the carotid and jugular foramen which
is not best addressed from a medial approach. The authors did not comment on the need for craniocervical stabilization in their patients with
lower clival disease. The need, timing, and nature of this procedure needs
to be identified up front and planned appropriately so that it does not
interfere with the surgical access for tumoral resection.
Franco DeMonte
Houston, Texas

VOLUME 76 | NUMBER 2 | FEBRUARY 2015 | 185

Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.

JAHANGIRI ET AL

his article presents the experience from the UCSF Medical center in the
treatment of chordomas. It adds no novel information to our current
knowledge, but still it illustrates several aspects of the management of skull
base chordomas that are worth emphasizing. First, the authors provide
further evidence of the critical importance of achieving gross total resection
by showing a faster and significantly increased risk of recurrence/regrowth
after subtotal resection. It is our current strategy at UPMC to always perform
a fine-cut T1 and T2 MRI on postoperative day 1 or 2 to identify any
residual tumor that could be amenable for further resection. Our current
resection rates for original clival chordomas is greater than 90%, which takes
us to the second important aspect nicely illustrated in this report: the
introduction of the endoscopic endonasal approach and the learning curve
for the surgical treatment of chordomas.
Their reported gross total resection rate of 52% is well below current
standards, but when looking carefully at their data, in particular, Figure
2B, it is clear that, since the introduction of endoscopic endonasal
surgery, their resection rates have progressively improved from 64% to
80%, which represents a remarkable improvement compared with the
initial rate of 31%.1 These data clearly exemplify the dramatic impact
that the introduction of endoscopic endonasal surgery has had in the
management of skull base chordomas. In our current practice, the
endoscopic endonasal approach represents the single most important
approach for clival chordomas, while transcranial microsurgical approaches remain as additional options for large and/or recurrent tumors
that cannot be fully resected via the endonasal route.
Yet, another critical point of this article is the excessively high rate of
residual tumor at the lower clival area (63%), something that was also
found on our clinical series where we observed an initial (2003-2007)
residual tumor rate at the lower clivus of 75% that was improved to 38%
for the period 2008 to 2011. Tumors extending to the jugular foramen,
hypoglossal canal, occipital condyle, and occipitocervical junction are
challenging and may cause significant morbidity. Not only the learning
curve of the surgical technique, but also a better understanding of the
complex surgical anatomy of the lower clivus is paramount to increase
resection rates and minimize surgical morbidity.2
Finally, the authors show that proton- and photon-based radiotherapy
yielded similar results. This study is obviously neither designed nor valid
to address this important debate, but it is just another one that raises the
urge to perform multicentric randomized trials evaluating proton- vs
photon-based radiotherapy for patients with chordoma.
Juan C. Fernandez-Miranda
Pittsburgh, Pennsylvania

1. Koutourousiou M, Gardner PA, Tormenti MJ, et al. Endoscopic endonasal

approach for resection of cranial base chordomas: outcomes and learning curve.
Neurosurgery. 2012;71(3):614-624; discussion 624-625.
2. Fernandez-Miranda JC, Morera VA, Snyderman CH, Gardner P. Endoscopic
endonasal transclival approach to the jugular tubercle. Neurosurgery. 2012;71(1
suppl operative):146-59.

186 | VOLUME 76 | NUMBER 2 | FEBRUARY 2015

CME QUESTIONS:
1. What is the cell of origin of chordomas?

A. Physaliferous cells
B. Notochord remnants
C. Arachnoid cap cells
D. Rathkes pouch epithelium
2. A 35-year-old male presents with headaches, diplopia and right
abducens nerve palsy; his MRI is shown below. What is the most
likely diagnosis?

A. Chondrosarcoma
B. Chordoma
C. Plasmacytoma
D. Pituitary macroadenoma
E. Clival meningioma
3. What theoretical advantage does proton beam therapy offer for the
treatment of chordomas?

A. Proton particles are more effective than photons

B. Fewer treatment sessions to achieve treatment dose
C. Lower dose delivered to surrounding normal tissue
D. Better tolerance of the brainstem to protons

www.neurosurgery-online.com

Copyright Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.