CLINICAL STUDIES

PATIENT OUTCOME AT LONG-TERM FOLLOW-UP

AFTER AGGRESSIVE MICROSURGICAL RESECTION
OF CRANIAL BASE CHORDOMAS

Fortios Tzortzidis, M.D.
Department of Neurosurgery,
University of Washington,
Seattle, Washington
Foad Elahi, M.D.
Department of Neurosurgery,
University of Washington,
Seattle, Washington
Donald Wright, M.D.
Department of Neurosurgery,
Virginia Hospital Center,
Arlington, Virginia
Sabareesh K. Natarajan, M.S.,
M.D.
Department of Neurosurgery,
University of Washington,
Seattle, Washington
Laligam N. Sekhar, M.D.
Department of Neurosurgery,
University of Washington,
Seattle, Washington
Reprint requests:
Laligam N. Sekhar, M.D., F.A.C.S.
Department of Neurosurgery,
325 Ninth Avenue,
Box 359924,
Seattle, WA 98104.
Email: lsekhar@uwashington.edu
Received, August 1, 2005.
Accepted, April 10, 2006.
OBJECTIVE: In this study, we evaluated patients’ clinical outcome and recurrence rates at
long-term follow-up after aggressive microsurgical resection of cranial base chordomas.
METHODS: Seventy-four patients with chordomas underwent operations during a
16-year period from 1988 to 2004. The philosophy was to perform complete resection
whenever possible and to provide adjuvant radiotherapy for remnants. Staged opera-
tions were performed for extensive tumors or if a sizable tumor remnant was noted
after the first resection. Patients included primary (previously untreated) and previously
operated or irradiated cases. Information was prospectively gathered concerning the
patients’ neurological condition, Karnofsky Performance Scale score, and tumor status
on magnetic resonance imaging scans.
RESULTS: There were 47 primarily operated patients (63.5%) and 27 patients (36.5%) who
had previously undergone surgery or radiotherapy. A total of 121 procedures were performed
in 74 patients. The mean follow-up period was 96 months, with a range of 1 to 198 months.
A single stage removal was performed in 41 (55.4%) of the patients and multiple stage removal
was performed in 33 (44.5%) of the patients. Gross total removal was accomplished in 53
(71.6%) of the patients, and subtotal resection was accomplished in 21 (28.4%) of the patients.
During the follow-up period, 24 (32%) of the patients had no evidence of disease, 37 (50%) of
the patients were alive with evidence of disease, 11 (14.8%) of the patients died of disease, and
two (2.7%) of the patients died of complications. Recurrence-free survival at 10 years was 31%
for the whole group, 42% for the primarily operated patients, and 26% for the reoperation
cases (P ⫽ 0.0001). The average Karnofsky Performance Scale score was 80 ⫾ 11.7 preoper-
atively, 84 ⫾ 8.9 at the 1-year follow-up, and 86 ⫾ 12.8 at the last follow-up in surviving
patients. No conclusion could be drawn regarding the value of radiotherapy because of the
treatment philosophy and the small number of patients.
CONCLUSION: Aggressive microsurgical resection of chordomas can be followed by
long-term, tumor-free survival with good functional outcome. A more conservative
strategy is recommended in reoperation cases, especially after previous radiotherapy,
to reduce postoperative complications.
KEY WORDS: Chordoma, Clinical outcome, Follow-up
Neurosurgery 59:230-237, 2006 DOI: 10.1227/01.NEU.0000223441.51012.9D www.neurosurgery-online.com

C
hordomas represent fewer than 0.1% of
all cranial base tumors. Chordomas de-
velop from notochord remnants, which
exist in normal adults as the nucleus pulposus
of the intervertebral discs. However, other
remnants may be found in the clival bone
marrow, accounting for the extradural loca-
tion of most chordomas. Primary intradural
chordomas have been reported, but they are
rare (11, 13, 22, 25, 26).
On short repetition time (TR)/short echo
time (TE) images, chordomas generally have
a low-to-intermediate signal. On long TR/
long TE images, chordomas generally have a
very high signal that is heterogeneous in
more than two-thirds of the images (9, 18).
After gadolinium administration, all chor-
domas demonstrate some degree of contrast
enhancement. The T2-weighted imaging sig-
nal intensity of chordomas is significantly

230 | VOLUME 59 | NUMBER 2 | AUGUST 2006 www.neurosurgery-online.com

 FOLLOW-UP AFTER RESECTION OF CRANIAL BASE CHORDOMAS

higher than nasopharyngeal carcinomas and pituitary ade-

nomas (18). TABLE 2. Tumor size distribution
Chordomas are radiosensitive only in the 70- to 80-Gy dose Tumor size distributiona No. of tumors (%)
range (4, 19, 24). Multiple delivery methods have been used to
⬍2 cm 14 (19%)
treat chordomas. The most common approaches include
2– 4 cm 32 (43.2%)
charged particle (protons, helium, or neon ions) radiation and
⬎4 cm 28 (37.8%)
stereotactic radiosurgery (4, 15, 19, 20, 21).
Surgical tumor resection plays a major role in the treatment a
Size shown as tumor equivalent diameter, 3公 D1D2D3.
of chordomas. However, the long-term usefulness of gross
total surgical resection has not been defined, and there are
differing philosophies regarding the treatment of these tumors base approaches most commonly used for tumor resection
with respect to surgical resection and radiotherapy (2, 7, 12, included extended subfrontal, frontotemporal orbitozygo-
16, 23, 27). In this study, we evaluated patients’ clinical out- matic transcavernous, subtemporal-infratemporal, extreme
comes and recurrence rate at long-term follow-up after micro- lateral transcondylar, and subtemporal-transzygomatic ap-
surgical resection of cranial base chordomas. proaches. Other approaches, such as extended transsphenoi-
dal and transmaxillary, were used occasionally when neces-
MATERIALS AND METHODS sary (Table 3).
The philosophy of treatment was to perform complete re-
From 1988 to 2004, 74 patients with clival and cranial base section when possible and to provide adjuvant radiotherapy
chordomas underwent surgery. A prospective database of the for remnants after attempted complete removal, as seen on
patients was maintained with data entered by a trained nurse magnetic resonance imaging (MRI) scans 2 to 3 months after
practitioner or physicians’ assistant. A total of 121 operative the surgery. Staged operations were performed for extensive
procedures were performed. Forty-seven patients (63.5%) tumors. Information was prospectively gathered concerning
were operated primarily and 27 patients (36.5%) had previous patients’ neurological condition, Karnofsky Performance Scale
surgery or radiation therapy. Forty-one patients (55.4%) had a (KPS) score, and tumor status on MRI examination.
single operation and 33 patients (44.6%) had more than one After adequate tumor exposure, the core of the tumor was
operation (Table 1). removed, followed by drilling of the bony tumor margins
Most patients had large or giant tumors (Table 2). Cranial until normal-appearing bone marrow was reached. If there
was intradural tumor invasion, complete tumor resection
was attempted if it was possible to perform safely. Al-
TABLE 1. Summary of data, January 1989 to January 2005a though encased major arteries, such as the internal carotid
artery, could be dissected away in the majority of patients,
Total patients 74 in some patients with recurrent tumors, arterial resection
Male (%) 36 (49%) with graft repair was necessary. If perforating arteries were
Female (%) 38 (51%) encased by intradural tumor in some patients, it was essen-
Primary patients (%) 47 (63.5%) tial to leave tumor behind.
Reoperation cases (%) 27 (36.5%) At the conclusion of tumor resection, layered repair of the
Overall surgical resection cranial base was performed using a combination of dural
Gross total resection (%) 53 (71.6%) repair (if necessary), free fat grafts, and vascularized flaps,
Subtotal resection (%) 21 (28.4%) such as the pericranial, temporalis, or rectus abdominis free
Recurrence after gross total resection 26 cases flap. After the conclusion of the first operation, an MRI scan
Treated by reoperation or radiosurgery was performed. If a significant (⬎2 cm) tumor remnant was
No. of operative procedures noted, reoperation (usually by a different approach) was
One operation (%) 41 (55.4%) performed, especially in primary cases. Otherwise, the pa-
Two or more operations (%) 33 (44.6%)
Total procedures 121
Follow-up period TABLE 3. Operative approaches
Range (mo) 1–198 Operative approaches Cases (%)
Mean (mo) 96 Frontotemporal orbitozygomatic, transcavernous 14.5%
Revascularization Subtemporal-infratemporal 26.7%
Bypass surgery (no.) 6 Subtemporal-transzygomatic 25.6%
Gross total resection (no.) 4 Extended subfrontal 14.5%
Subtotal resection (no.) 2 Extreme lateral transcondylar 16.2%
a
GTR, gross total resection. Others 8%

NEUROSURGERY VOLUME 59 | NUMBER 2 | AUGUST 2006 | 231

TZORTZIDIS ET AL.

tient was treated with adjuvant radiotherapy. A more con- section. They were treated by irradiation only (six cases, 26%),
servative approach was used in recurrent cases. Such radio- resection followed by irradiation (19 cases, 73%), or no treat-
therapy usually consisted of proton beam radiotherapy or ment (one case, 4%) (Table 5). The types of radiation adminis-
radiosurgery. Although it was our goal to refer such pa- tered include proton beam radiotherapy (38%), radiosurgery
tients for proton beam radiotherapy, many patients could (43%), and fractionated external beam radiotherapy (19%) (Fig.
not afford the travel and the long-term stay that was re- 1).
quired. Therefore, both radiosurgery and fractionated ra- At the last follow-up examination, 24 patients (32%) had no
diotherapy were used as adjuvants. In some patients, a evidence of disease, 37 patients (50%) were alive with evi-
multistage operative approach was planned preoperatively, dence of disease, 11 patients (14.8%) died of disease, and two
based on the extent of tumor present in the initial MRI patients (2.7%) died of complications (Table 5).
scans. Recurrence-free survival was observed to be 1 to 36 months
All tumors were examined pathologically with immunohis- in 26 patients (56%), 37 to 84 months in 21 patients (45%), 85
tochemistry, especially for epithelial membrane antigen and to 132 months in 19 patients (41%), and longer than 132
cytokeratin (CK), to distinguish the tumor from chondrosar- months in 15 patients (31%). Recurrence-free survival for pre-
coma. viously treated cases was much less than for the primary
patients (log-rank significance, 0.0001) (Fig. 2). At 10 years, the
Follow-up recurrence-free survival for primarily operated patients was
Follow-up examinations were performed in patients annu- 42%; for reoperation cases, it was 26% (Table 7).
ally (whenever possible) by direct examination and MRI scan- This study was not designed to compare irradiated and
ning. At the conclusion of the study in December 2004, further nonirradiated patients after complete tumor removal because
follow-up was conducted by a standardized questionnaire and none of the patients received irradiation after complete resec-
a telephone interview by a nurse practitioner or physicians’ tion. The number of patients irradiated was not large enough
assistant. Information was gathered regarding patients’ qual- to compare the different radiation modalities after subtotal
ity of life, neurological deficits, and history of other treat-
ments. The most recent MRI examination and/or the radiolo-
gist’s report were obtained for review. The MRI scans were TABLE 4. Complications after surgerya
reviewed by radiologists who were not part of the study. If no Complication No. of patients
tumor was visible on MRI scans, the patient was considered CSF leakage 1
free of tumor. Survival curves were plotted. CN palsy, permanent 3
Hydrocephalus 1
Deep vein thrombosis 2
RESULTS Pulmonary emboli 1
Hemiparesis 1
Postoperative complications in these patient series are sum-
Postoperative stroke 1
marized in Table 4. Two patients died postoperatively of com-
Sphenoid sinus infection 1
plications. The mean follow-up period was 96 months, with a
Hematoma at the fat graft harvesting site 1
range of 1 to 198 months. A single stage removal was per-
Death 2
formed in 41 patients (55.4%) and multiple stage removal was
Total 12
performed in 33 patients (44.5%). Gross total resection was
a
accomplished in 53 patients (71.6%), and subtotal resection CSF, cerebrospinal fluid; CN, cranial nerve.
was accomplished in 21 patients (28.4%). In primarily oper-
ated patients, gross total re-
section was accomplished in
39 patients (83%), and, in re-
operated patients, gross total TABLE 5. Patients’ condition at the last follow-up examinationa
resection was accomplished No. Recurrence NED AWD DOD DOC
in 14 patients (30%) (Table 5). Primary cases (n ⫽ 47)
The average KPS score was Gross total resection 39 17 21 17 — 1
80 ⫾ 11.7 preoperatively, 84 Subtotal resection 8 — — 5 3 —
⫾ 8.9 at the 1-year follow-up Reoperation cases (n ⫽ 27)
examination, and 86 ⫾ 12.8 at Gross total resection 14 9 3 9 4 —
the last follow-up examina- Subtotal resection 13 — — 6 4 1
tion (in surviving patients) Total 26 24 37 11 2
(Table 6). a
NED, no evidence of disease; AWD, alive with evidence of disease DOD, died of disease; DOC, died of complications.
Twenty-six patients had re-
currence after gross total re-

232 | VOLUME 59 | NUMBER 2 | AUGUST 2006 www.neurosurgery-online.com

 FOLLOW-UP AFTER RESECTION OF CRANIAL BASE CHORDOMAS

TABLE 6. Karnofsky Performance Scale scorea TABLE 7. Overall recurrence-free survival

3 yr 5 yr 10 yr
Preoperative (mean ⫾ SD) 80 ⫾ 11.7
Primary cases 56% 47% 42%
1-year follow-up (mean ⫾ SD)b 84 ⫾ 8.9
Reoperated cases 41% 39% 26%
Last follow-up (mean ⫾ SD)c 86 ⫾ 12.8
All cases 46% 41% 31%
a
SD, standard deviation.
b
All patients.
c
Surviving patients.

FIGURE 3. Preoperative axial (A) and sagittal (B) contrast MRI scans
FIGURE 1. Flow chart showing patient characteristics and follow-up.
revealed a large petroclival tumor with extensive cavernous sinus involve-
ment.

FIGURE 2. MRI scans showing the recurrence-free survival of our series

of patients with chordomas. There was a significantly better survival in
primary operated patients.
resection, particularly proton beam radiotherapy and radio-
surgery.
Illustrative Cases
Patient 1
This case is presented to illustrate the operative approach to exten-
sive tumors, with multistage surgery.
A 43-year-old woman presented with facial numbness, VIth cranial
nerve palsy, and swallowing problems. Preoperative MRI scans
showed a petroclival, destructive cranial base lesion in the petroclival
area with extensive cavernous sinus involvement (Fig. 3). The lesion
was removed by a subtemporal infratemporal approach. However, on
the first postoperative day, an MRI examination revealed a sizeable
tumor remnant in the cavernous sinus. A re-exploration was per-
formed by a frontotemporal transcavernous approach, and the tumor
FIGURE 4. Follow-up axial T1-weighted MRI scan with contrast (A) and
sagittal T2-weighted MRI scan (B) revealing complete removal of tumor,
and no recurrence at 16 months.
was totally resected. The patient recovered completely and remains
free of tumor recurrence at 24 months with a KPS score of 90 (Fig. 4).
No radiotherapy was administered because the tumor was completely
resected.
Patient 2
This case is presented to illustrate the long-term recurrence-free
survival of a completely resected tumor.
A 61-year-old man presented with severe ataxia and headache. A
preoperative MRI scan revealed an upper and mid-clival chordoma
extending into the right cavernous sinus (Fig. 5). He underwent total
tumor resection by an extended subfrontal, transbasal approach. No
radiotherapy was administered because of complete tumor resection.
The patient was followed for 150 months, and, at his last MRI scan,

NEUROSURGERY VOLUME 59 | NUMBER 2 | AUGUST 2006 | 233

TZORTZIDIS ET AL.

FIGURE 5. Preoperative mid-sagittal (A) and axial (B) T2-weighted MRI

scans with contrast revealing an upper clival tumor.
FIGURE 7. Preoperative T2-weighted sagittal (A) and postcontrast coronal
T1-weighted MRI scans (B) revealing a large tumor extending from the pituitary
fossa to the upper and mid-clival region.

a tumor recurrence in the lower

clival area (Fig. 9) with new on-
set of hemiparesis. This tumor
was partially removed by a
transcondylar approach for pal-
liative purposes because he had
maintained good function until
then. In December 1996, the pa-
tient died because of multiple
tumor recurrences.

FIGURE 6. Follow-up mid-sagittal (A) and axial (B) T2-weighted MRI
scans revealing complete removal of tumor and no recurrence at 138
months after the operation.
there was no evidence of recurrence (Fig. 6). He is not able to smell
and has marked visual impairment in the right eye, but, until recently,
worked as a university professor. He has now retired.
Patient 3
This case illustrates an example of an extremely aggressive chor-
doma.
A 25-year-old man presented with severe headache, progressive
IIIrd nerve palsy, and complete VIth nerve palsy. He also had total
pituitary dysfunction. On MRI scans, a destructive tumor was re-
vealed extending from the pituitary fossa to the upper and mid-clivus
(Fig. 7). His preoperative KPS score was 90. In March 1994, the tumor
was partially removed using an extended subfrontal approach with
residual tumor in left cavernous sinus, which was inaccessible
through this approach. On the second stage surgery, the tumor was
completely removed via a transsylvian approach. After 5 months from
his first admission, a third operation was performed because of ag-
gressive tumor recurrence.
Six months after the first operation, tumor recurred significantly in
the same area. He underwent reoperation for tumor excision, and he
subsequently underwent proton beam radiotherapy. In May 1995, he
developed back pain and urinary incontinence. The MRI scan revealed
a sacral tumor, which was partially removed via a sacral laminectomy,
and he received local radiotherapy (Fig. 8). In August of 1996, he had
DISCUSSION
Chordomas are locally ag-
gressive tumors that arise
from remnants of the noto-
chord. The predominance of
these tumors occurs in the sa-
crococcygeal and cranial base
areas, although they do in-
volve other spinal areas ( 6–8, FIGURE 8. Sagittal T2-weighted
10, 21). Their histological ap- lumbar MRI revealing a sacral tumor.
pearance is typical, with areas The patient experienced a seeding
of cartilage or bone destruc- along the neuraxis or de novo tumor
tion and bubble-like or “phys- formation in the sacral area 14 months
aliphorous” cells (25). A vari- after initial resection.
ant of chordoma called the
“chondroid chordoma” is recognized with features resembling
chondrosarcoma (5, 25). Although some series have reported a
better long-term outcome with chondroid chordoma, this was
not borne out in other series (2, 4, 12, 19). On immunohistochem-
istry, chordomas stain positively for S100, vimentin, epithelial
membrane antigen, and CK 8/18. They also express other CKs,
such as CK 1/10, CK 7, CK 20, CK 19, and CK 12 to 17. Chon-
drosarcomas of the cranial base do not stain positively for epi-
thelial membrane antigen and CK. In reviewing any series of
these tumors, it is important to elucidate whether the patients

234 | VOLUME 59 | NUMBER 2 | AUGUST 2006 www.neurosurgery-online.com

 FOLLOW-UP AFTER RESECTION OF CRANIAL BASE CHORDOMAS

were studied by immunohis-
tochemistry. Chondrosarco-
mas of the cranial base have a
much more benign course
than chordomas, and mixing
these two types of tumors in
the report may give a false
idea regarding the long-term
results of a particular treat-
ment modality used (11, 13,
19). Chordomas seem to have FIGURE 9. T1-weighted postcon-
two distinct biological behav- trast MRI revealed the last recurrence
iors. The first group, which of this extremely aggressive tumor in
comprises the majority of tu- the upper and lower clival area in Au-
mors, are slow growing, and, gust 1996.
in rare cases, may not even
grow at all (the senior author, LNS, is following two such pa-
tients). Such tumors also remain locally confined rather than
metastasize to other areas of the remains or other body areas (3,
6, 8). The second group has a more aggressive behavior, with
rapid local recurrence spread to other areas of the neuroaxis or
metastasis to the lung, liver, or bone (as seen in Patients 3 and 4).
Some primary tumors and many postradiation recurrent tumors
behave in a more aggressive fashion (14, 16, 17).
The treatment of cranial base chordomas is based entirely
on Class III evidence, as there are no randomized trials re-
garding the efficacy of surgery or radiotherapy (1). Because
chordomas are known to be minimally radiosensitive, high-
dose therapy with proton beam radiation and radiosurgery
using the gamma knife and CyberKnife have been used (4,
19–21, 23, 24). There are no reports regarding long-term results
after gamma knife or CyberKnife radiosurgery, but there are
reports regarding proton beam radiotherapy (4, 19). On the
basis of the results of treatment with proton beam radiother-
apy, it seems that chondrosarcomas respond much better than
chordomas, and the results are better for chondroid chordo-
mas.
The term “chondroid chordoma” is based on the findings of
large amounts of cartilaginous material in the matrix of chor-
domas, but not all pathologists are convinced regarding this
distinction (5). In a previous report, we did not find a better
prognosis in “chondroid chordomas” (12). In this report, our
patients were not classified into chondroid and nonchondroid
varieties. However, it is obvious that there is a distinct group
of chordomas with a more aggressive behavior, as exemplified
by Patient 3. Most chordomas that have recurred after radio-
therapy also belong to this group.
There are three different philosophies regarding the treat-
ment of chordomas: aggressive surgical resection, with radio-
therapy given only in patients who have remnants, aggressive
resection followed by radiotherapy, and partial resection fol-
lowed by radiotherapy. After aggressive surgical resection, we
have also observed a third group of patients with delayed
recurrence that may do well for a number of years with
reoperation. The senior author (LNS) and Crockard et al. (7)
have followed the policy of aggressive surgical resection and
no radiotherapy unless distinct remnants remain. Al-Mefty
and Borba (2) administer radiotherapy to all patients postop-
eratively, regardless of resection. Other neurosurgeons have
adopted varying philosophies. Our series presents the infor-
mation regarding long-term follow-up, which is not yet avail-
able with other series (Table 8). None of the previously re-
ported patient series have had the length of follow-up as our
series. For a disease with low incidence and recurrence even
after 10 years, it would be very difficult to conduct a random-
ized trial to compare different treatment modalities.
In our series of patients, we have used a variety of cranial
base approaches for the resection of tumors on the basis of the
tumors’ location and the senior author’s (LNS) preference.
Certain approaches were used less frequently, in particular,
the transphenoidal, transmaxillary, and transoral approaches.
We also did not use endoscopic resection of these tumors as is
being performed presently in some centers (endoscopic assis-
tance was used during resection of some tumors). Frameless

TABLE 8. Previously published seriesa

Recurrence-free survival Mean
Series No. of Sex Tumor-
Therapy GTR STR PR RT follow-up
(ref. no.) patients (M/F) related death
3 mo 3 yr 5 yr 7 yr 11 yr >11 yr (mo)
Al-Mefty and 25 15/10 Radical surgery 43 48 9 17 16/21 25.4 2 (10%)
Borba, 1997 (2) and proton-photon
beam therapy
Gay et al., 1995 46 Radical surgery 47 43 10 12 65% 46 5 (10.8%)
(12) and RT for ST only
Forsyth et al., 51 Surgery and RT 40 11 39 51% 35%
1993 (11)
Current studies 74 36/38 Radical surgery 53 21 0 25 56% 45% 41% 32% 96 11 (14.9%)
and RT for ST only
a
GTR, gross total resection; STR, subtotal resection; PR, partial resection; RT, radiotherapy.

NEUROSURGERY VOLUME 59 | NUMBER 2 | AUGUST 2006 | 235

TZORTZIDIS ET AL.
stereotactic guidance has been extensively used for extradural
tumors during the past 6 years and was found to be particu-
larly useful for tumor removal from certain areas. The use of
an individual surgical approach is based on the experience of
the surgeon and the surgeon’s confidence with the safety and
efficacy of the approach for tumor resection. As such, there is
no correct or incorrect approach; one has to focus on the
results and complications.
On the basis of the experience gained from this series, it may
be stated that tumor resections are much easier if the patient is
seen initially, before the patient has had a previous resection or
radiotherapy. Recurrent tumors are not only more difficult to
remove, but also carry a higher complication rate. Thus, the
surgeon’s aim should be to administer the optional treatment
during the initial treatment session, which may consist of tumor
resection and/or radiotherapy. The senior author (LNS) thinks
that, whenever possible, radiotherapy should be reserved for
recurrences rather than initial therapy. However, the philosophy
should be tempered by not creating a significant disability in the
patient. In this series, better long-term results were obtained in
primary tumors than recurrent tumors. However, this may be
caused by more aggressive biological behavior (recurrent tu-
mors) rather than complete removal (in primary patients).
When a patient presents with recurrent tumors, the approach
should be different and more conservative. If the tumor can be
removed completely without causing disability, it should be
removed. However, the surgeon (and the patient) should be
aware that the chances of long-term survival and the potential for
complications (resulting in disability) are higher in reoperation
cases. The treatment strategy should be planned appropriately.
CONCLUSION
A series of patients with chordomas treated by aggressive
surgical resection and long-term follow-up is presented. The
survival was better in primary operation patients than in
reoperation cases. We could not draw any conclusions regard-
ing the value of radiotherapy, particularly after complete re-
moval. New approaches to treatment of aggressive chordomas
are necessary. We also could not compare the different radi-
ation modalities because of the small numbers.
REFERENCES
1. Anonymous: The periodic health examination: 2. 1987 update. Canadian
Task Force on the Periodic Health Examination. CMAJ 138:618–626, 1998.
2. Al-Mefty O, Borba LA: Skull base chordomas: A management challenge.
J Neurosurg 86:182–189, 1997.
3. Arnautovic KI, Al-Mefty O: Surgical seeding of chordomas. J Neurosurg
95:798–803, 2001.
4. Austin-Seymour M, Munzenrider J, Goitein M, Verhey L, Urie M, Gentry R,
Birnbaum S, Ruotolo D, McManus P, Skates S, Ojemann RG, Rosenberg A,
Schiller A, Koehler A, Suit HD: Fractionated proton radiation therapy of
chordoma and low-grade chondrosarcoma of the base of the skull.
J Neurosurg 70:13–17, 1989.
5. Brooks JJ, LiVolsi VA, Trojanowski JQ: Does chondroid chordoma exist?
Acta Neuropathol (Berl) 72:229–235, 1987.
236 | VOLUME 59 | NUMBER 2 | AUGUST 2006
6. Colli B, Al-Mefty O: Chordomas of the craniocervical junction: Follow-up
review and prognostic factors. J Neurosurg 95: 933–943, 2001.
7. Crockard HA, Steel T, Plowman N, Singh A, Crossman J, Revesz T, Holton
JL, Cheeseman A: A multidisciplinary team approach to skull base chordo-
mas. J Neurosurg 95:175–183, 2001.
8. Dalpra L, Malgara R, Miozzo M, Riva P, Volonte M, Larizza L, Fuhrman
Conti AM: First cytogenetic study of a recurrent familial chordoma of the
clivus. Int J Cancer 81:24–30, 1999.
9. Erdem E, Angtuaco EC, Van Hemert R, Park JS, Al-Mefty O: Comprehensive
review of intracranial chordoma. Radiographics 23:995–1009, 2003.
10. Eriksson B, Gunterberg B, Kindblom LG: Chordoma: A clinicopathologic
and prognostic study of a Swedish national series. Acta Othop Scand
52:49–58, 1981.
11. Forsyth PA, Cascino TL, Shae EG, Scheithauer BW, O’Fallon JR, Dozier JC,
Piepgras DG: Intracranial chordomas: A clinicopathological and prognostic
study of 51 cases. J Neurosurg 78:741–747, 1993.
12. Gay E, Sekhar LN, Rubinstein E, Wright DC, Sen C, Janecka IP, Snyderman
CH: Chordomas and chondrosarcomas of the cranial base: Results and
follow-up of 60 patients. Neurosurgery 36:887–897, 1995.
13. Heffelfinger MJ, Dahlin DC, MacCarty CS, Beabout JW: Chordomas and
cartilaginous tumors at the skull base. Cancer 32:410–420, 1973.
14. Holton JL, Steel T, Luxsuwong M, Crockard HA, Revesz T: Skull base
chordoma: Correlation of tumor doubling time with age, mitosis and Ki67
proliferation index. Neuropathol Appl Neurobiol 26:497–503, 2000.
15. Hug EB, Fitzek MM, Liebsch NJ, Munzenrider JE: Locally challenging osteo-
and chondrogenic tumors of the axial skeleton: Results of combined proton
and photon radiation therapy using three-dimensional treatment planning.
Int J Radiat Oncol and Biol Phys 31:467–476, 1995.
16. Kamrin RP, Potanos JN, Pool JL: An evaluation of the diagnosis and treat-
ment of chordoma. J Neurol Neurosurg Psychiatry 27:157–165, 1964.
17. Korten A, Berg H, Spincemaille GH, Laan RT, van de Wel AM: Intracranial
chondrosarcoma: Review of the literature and report of 15 cases. J Neurol
Neurosurg Psychiatry 65:88–92, 1998.
18. Meyers SP, Hirsch WL Jr, Curtin HD, Barnes L, Sekhar LN, Sen C: Chordo-
mas of the skull base: MR features. AJNR Am J Neuroradiol 13:1627–1636,
1992.
19. Munzenrider JE, Liebsch NJ: Proton therapy for tumors of the skull base.
Strahlenther Oncol 175 [Suppl 2]:57–63, 1999.
20. Muthukumar N, Kondziolka D, Lunsford LD, Flickinger JC: Stereotactic
radiosurgery for chordoma and chondrosarcoma: Further experiences. Int J
Radiation Oncology Biol Phys 41:387–392, 1998.
21. Noël G, Feuvret L, Ferrand R, Boisserie G, Mazeron JJ, Habrand JL: Radio-
therapeutic factors in the management of cervical-basal chordomas and
chondrosarcomas. Neurosurgery 55:1252–1262, 2004.
22. Nishigaya K, Kaneko M, Ohashi Y, Nukui H: Intradural retroclival chor-
doma without bone involvement: No tumor regrowth 5 years after opera-
tion: Case report. J Neurosurg 88:764–768, 1998.
23. O’Connell JX, Renard LG, Liebsch NJ, Efird JT, Munzenrider JE, Rosenberg
AE: Base of skull chordoma: A correlative study and clinical features of 62
cases. Cancer 74:2261–2267, 1994.
24. Pearlman AW, Friedman M: Radical radiation therapy of chordoma. Am J
Roentgenol Radium Ther Nucl Med 108:333–341, 1970.
25. Russel DS, Rubinstein LJ (eds): Pathology of Tumors of the Nervous System.
London, Edward Arnold, 1989, pp 664–765.
26. Warnick RE, Raisanen J, Kaczmar T, Davis RL, Prados MD: Intradural
chordoma of the tentorium cerebelli: Case report. J Neurosurg 74:508–511,
1991.
27. Watkins L, Khudados ES, Kaleoglu M, Revesz T, Sacares P, Crockard HA:
Skull base chordomas: A review of 38 patients, 1958–1988. Br J Neurosurg
7:241–248, 1993.
Acknowledgments
We thank Ramin Rak, M.D., Susan Buchholtz, N.P., and Oleg Rivkin, P.A., for
their assistance with data collection.
www.neurosurgery-online.com
 FOLLOW-UP
COMMENTS
A large series of 74 patients with chordomas who underwent a total
of 121 surgical procedures is presented in this study. With a mean
follow-up period of 96 months, the authors present 12 complications
after surgery and two deaths in previously operated and irradiated
patients.
Chordomas present a challenging problem for the neurosurgeon.
The biological behavior of chordomas is not uniform, and two patients
with the same diagnosis may have a completely different prognosis
independent of the surgical strategy. In our experience, aggressively
growing chordomas can rarely be completely removed by microsur-
gical means, and most will show a large recurrence within 1 year.
These patients will have a fatal outcome, even with repeated surgical
procedures. Conversely, slow-growing chordomas present a better
prognosis after radical resection of the tumor. The crucial question
when treating the latter group is what amount of morbidity has to be
taken into account for a radical resection. In cases of recurrent tumors,
we agree with the authors to perform another resection. Afterwards,
radiation therapy should be performed in suspicious regions in which
surgeons could expect a recurrence.
Recurrences after surgery and radiation present a difficult and
challenging problem for the neurosurgeon because these cases are
prone to develop complications, such as cranial nerve and vascular
injuries. With increased experience, surgical removal of the tumors
can be more complete. In our opinion, we should not perform a radical
resection if it impedes on the patient’s quality of life. In this context,
“aggressive” surgical treatments do not entail damaging important
structures, but they do require radical, yet gentle, microsurgical tech-
niques.
Wolf Lüdemann
Madjid Samii
Hannover, Germany
T zortzidis et al. provide an important contribution for defining the
role of surgery in the management of cranial base chordomas. The
authors report on a large cohort of 74 patients harboring cranial base
chordomas who underwent aggressive microsurgical resection be-
tween 1988 and 2004. There is a long follow-up period (mean, 96 mo;
range, 1–198 mo), and the results were unaffected by radiotherapy,
which was given only in two cases. Gross total resection was accom-
plished in 53 patients (71.6%), and subtotal resection was accom-
plished in 21 patients (28.4%). Twenty-six patients (35%) had recur-
rences after gross total resection. At the 10-year follow-up
examination, 19% of patients were disease free. The average Karnof-
sky Performance Scale score was 80 ⫾ 11.7 preoperatively and 86 ⫾
12.8 at the time of the last follow-up visit in surviving patients.
NEUROSURGERY
AFTER RESECTION OF CRANIAL BASE CHORDOMAS
The authors’ results are comparable to previously reported findings
in which radiotherapy was administered (Table 8), thus supporting
aggressive tumor resection to obtain long survival with good func-
tional preservation.
The authors report the differences in the biological behavior of
chordomas. The majority are slow growing, some do not grow, and
some rare cases are aggressive, with rapid local recurrence and neural
and /or extra neural metastasis. There is currently no grading system
to define the malignancy of this type of tumor, and, in my opinion,
with the absence of similar criteria, aggressive management of chor-
domas is mandatory.
Giorgio Frank
Bologna, Italy
T his article reports clinical outcome and rates of recurrence in a
large series of 74 patients managed with a consistent strategy of
attempted complete resection and reoperation or radiation for any
residual tumor. Complete removal was achieved by one or more
operation in 53 out of 74 patients (71.6%), with an operative mortality
rate of 2.7%. At the time of the last follow-up visit, 35% were alive
with disease, 38% were alive without disease, and 15% had died of
disease. The median duration of postoperative survival was approx-
imately 8 years.
This excellent outcome in a large series supports the authors’ con-
tention that their strategy has merit. However, the data do not estab-
lish the superiority of this approach compared with the alternatives of
aggressive surgery with postoperative radiation, even if radiograph-
ically complete removal is achieved, conservative surgery with post-
operative radiation, or either radiotherapy or radiosurgery alone.
Griffith R. Harsh IV
Stanford, California
T he authors present a rather extensive cranial base experience in the
surgical treatment of chordomas in 74 patients. The authors note
that, although chordomas may look the same histologically, they may
be biologically different and have a different clinical course. They
clearly indicate the benefit of aggressive surgical resection in the
majority of patients, irrespective of the type of tumor. We agree with
the authors that the best chance for a more radical resection is at the
first surgery. For this reason, patients with the presumptive radiolog-
ical diagnosis of chordoma are better served in specialized centers that
have a high volume and experience with these relatively uncommon
lesions. It would have been interesting if the authors reported how
their experience influenced the outcomes by comparing recently
treated patients with those treated earlier.
Ali F. Krisht
Little Rock, Arkansas
VOLUME 59 | NUMBER 2 | AUGUST 2006 | 237