Clinical and Pathologic Review of 48 Cases of Chordoma

TYVIN A. RICH, MD.’.t ALAN SCHILLER, MD,* HERMAN D. SUIT, MD, PHD.’ AND HENRY J. MANKIN, MD9

The results of treatment of 48 patients with the diagnosis of chordoma during the period 1931 to 1981
at the Massachusetts General Hospital were reviewed. Fourteen patients were treated with surgery
alone: eight patients with primary tumors in the sacrococcygeal region were treated with radical surgery
and four are alive with no evidence of disease (NED) with follow-up of 8 to 20 years. Recurrent tumors
in six patients were treated with surgery alone resulting in long palliation (3-25 years). The actuarial
survival rate at 5 years for all patients treated with surgery was 76%. Radiation therapy was used in
patients after either a biopsy (15), partial excision (17). or before radical excision in 2 patients. To
achieve a worthwhile level of palliation, doses greater than 4000 c G y were required. Highdose levels
(>6500 cCy) were achieved in nine cases,by a combination of photon and 160 MeV proton beams.
The results to date of this approach for lesions of the base of skull and cervical vertebral body are
encouraging: high local control and low morbidity. The 5-year actuarial survival rate of all patients
treated with radiation was 50%.
Cancer 56:182-187, 1985.

C HORDOMAS are rare tumors that arise from rests of

notochord cells along the spinal axis. The anatomic
distribution of these tumors is: sacrococcygeal, approx-
imately 50%; vertebral, approximately 15%; and basioc-
ciput, approximately 35%.’-’ Chordomas grow slowly,
producing local destruction of bone, and frequently
extend to adjacent soft tissue. The most common s y m p
tom is pain secondary to destruction of bone and/or to
pressure effects on nerves or adjacent organs. The radio-
logic features of chordomas have been well d e ~ r i b e d . ~ . ’
Metastases are not seen except in those patients whose
tumor has reached massive proportions as occurs com-
monly in patients with tumors of the sacrococcygeal
and vertebral site^.^.'.^
Radical resection for lesions of the distal sacrum (S3-
S S ) and coccyx may be curative, as total removal is
often achieved. In contrast, complete resection of chor-
domas of the base of skull or vertebra is rarely successful
because of the inability to achieve good surgical mar-
gin~.’.~Eradication of chordomas by conventional radia-
tion therapy alone is uncommon because of the large
size and/or the limitation on dose imposed by the
sensitivity of adjacent structures. Where small lesions
From the Departments of *Radiation Medicine. $Pathology, and
§Orthopedic Surgery, Massachusetts General Hospital, Hamard Medical
School, Boston, Massachusetts.
Address for reprints: Tyvin A. Rich, MD. Department of Radio-
therapy, University of Texas System Cancer Center, M. D. Anderson
Hospital and Tumor Institute. 6723 Bertner Drive, Houston, TX
77030.
Accepted for publication July 13, 1984.
can be treated to high dose levels, long-term control can
be expected in a substantial number of patients.
This article presents an assessment of the criteria for
histologic diagnosis of chordoma based upon a review
of the clinical and pathologic features of tumors diag-
nosed as chordoma at the Massachusetts General Hos-
pital from 193 1 to 198 1. We then describe the surgical
and radiotherapeutic approaches employed in the treat-
ment of these patients and the results achieved.
Clinical Material
The clinical records of 53 patients who were listed in
our tumor registry with the diagnosis of chordoma were
reviewed. Five cases were excluded because of a change
of diagnosis and the remaining 48 cases are reported.
The age, sex, and anatomic distributions are shown
in Figures 1 and 2. A history of significant trauma to
the sacrum was recorded in 2 of 19 patients with
sacrococcygeal lesions. In both, this antedated the diag-
nosis by about 5 years. Table I lists the symptoms and
range of duration of symptoms before diagnosis for
chordomas at the basiocciput, vertebral, or sacrococcygeal
regions. For the 34 vertebral column and sacral patients,
pain was the most frequent symptom. Of 14 patients
with chordoma of the base of the skull, diplopia was the
most frequent symptom. Four of the base of skull
chordomas were of the chondroid variety (tumors con-
taining features of either chondrosarcoma or chondroma,
e.g, the stroma resembled hyaline cartilage with neo-
plastic cells in lacunae). The mean durations of symp-

182
 CHORDOMA . Rich et al. 183

n Percent of
Total

28% Base of
MALE FEMALE

Skull II

0 Fornolo
Yolo Cervical

34K Thoracic

10-20 30-39 40-49 50.59 -69 70-79

Age
FIG. 1. Age and sex distribution of patients with chordoma.

Lumbar
toms were 15.6 and 28.5 months for the regular and the
chondroid chordomas of the base of skull, respectively.

Pathology 38To Sacro-

Coccygaal

Grossly, this tumor appears nodular with dense fibrous

FIG. 2. Sex and site distribution of patients with chordoma.
trabeculae surrounding cystic or gelatinous tumor. A
pushing margin can be seen associated with osteolysis
and replacement of bone by diffuse tumor infiltration.
The fresh specimens have frequently been described as nodes, brain, lungs, and abdominal viscera. The histo-
having a deep reddish or purple color and contain logic pattern of the metastases was not different from
grossly hemorrhagic zones. Soft tissue masses are fre- that of its primary lesion. The cause of death usually
quently surrounded by a pseudocapsule.
Five of the original 53 cases of chordoma were
TABLEI . Symptoms of Patients With Chordoma
excluded. The diagnosis was changed to metastatic ad- ~ ~~~~~

enocarcinoma in two cases that were reticulin negative sacral ( 19) Range, 2-24 mo
Sacral pain 14
and had an unusual clinical course for chordoma. In Incontinence 2
two other cases, benign chordoma was found in the Constipation I
dura at the base of the skull on routine autopsy. These Vertebral ( 15) Range, 2-48 mo
Pain 16
lesions were clinically silent, and there was no evidence Bladder dysfunction 1
of invasion of adjacent structures grossly or microscop Hoarseness 1
ically. In one case no tumor was found on histologic Dysphagia 2
Pharyngeal bleeding I
review. The histopathologic slides were available for Lower extremity weakness 4
review in 32 of the 48 cases. Detailed study of these Base of Skull (14) Range, 3-72 mo
slides revealed that mitoses and necrosis were rarely Diplopia 10
Headache 3
found. Reticulin stains were performed on recut material Blindness I
available in 24 cases; in 22, reticulin fibers formed a Ptosis I
loose network around large cells with bubbly cytoplasm Gait disturbance 1
Tinnitus/vertigo 1
(the so-called physaliferous cells). In 16 of 48 cases, Hoarseness/dysphagia I
histologic material was unavailable but the clinical course Decreased visual acuity I
and radiographic features confirmed the diagnosis of No. of patients with Average duration of
chordoma. base of skull symptoms
c hordoma (range in mo) Pathology
The incidence of metastasis was 18% (9/48) and only
occurred late in the course of disease, ( 5 sacrococcygeal, 10 15.6 (1-48) Chordoma
4 28.5 (6-72) Chondroid chordoma
4 vertebral). Metastases were found in bones, lymph
184 CANCERJuly 1 1985
TABLE2. Treatment of Patients With Chordoma According to Site
Base of sacro-
skull Vertebral coccygeal Total
Surgery 4 2 8 14
Surgery (partial exci- 5 10 2 17
sion or total gross
removal) and post-
operative radiation
Biopsy and radiation 5 3 7 15
Preoperative radiation - - 2 2
and radical excision
48
was related to the consequence of the local effect of
relentless growth of the primary lesion.
Treatment
The use of surgery alone (14 cases) or in combination
with radiation (34 cases) according to anatomic site is
summarized in Table 2. In the surgery alone group,
eight patients with tumors located in the distal sacrum
and coccyx had complete resection. One of these eight
patients was treated by total extirpation of the coccyx
and the lower sacrum for recurrent disease, six had
limited resection of the coccyx and a portion of the
sacrum, and one had a hemi-pelvectomy for a massive
tumor of the lower sacrum that extended laterally bencath
the sciatic notch. The morbidity of radical surgery (eight
cases) was directly related to the location and extent of
the tumor. Tumors in the distal sacrococcygeal region
were totally removed without loss of anal or urogenital
function in six cases. In two cases more proximal there
was anal dysfunction resulting from the sacrifice of high
sacral roots. In patients with more advanced tumors in
the sacrococcygeal region treated by radiation therapy,
five had permanent neurologic damage affecting the
bladder and seven had anal dysfunction before treatment.
In six other patients with tumors of the base of skull
(four) or vertebral body (two), resection alone was
subtotal and usually achieved some relief of symptoms.
There was one postoperative death secondary to sepsis
after incomplete removal of tumor from the basiocciput.
Thirty-four patients were treated with radiation. In
17 patients radiation was given postoperatively after a
partial excision (16) and after radical excision but with
tumor spillage (I). Partial excision means that gross
disease was left in the patient. This combined approach
was most frequently used in patients with tumors of the
vertebral column where surgical removal of the tumor
was first attempted. Fifteen other patients were treated
with radiation alone after biopsy. In seven patients
(radiation alone [two], or partial excision and radiation
[five]), residual disease remained after radiation and
Vol. 56
further surgery was performed ( 1-8 additional opera-
tions). In two patients with extensive tumors of the
sacrum, preoperative radiation was used before radical
excision.
Physical characteristics of the radiation equipment
used varied markedly and included: 280 kvp x-ray, 1 to
10 MeV x-ray, 6oCo,and 160 MeV protons. Iodine 125
seeds were placed in the tumor bed along the pedicle of
the second cervical vertebra in one case. The doses of
radiation vaned over the period studied. Treatment was
usually given as five treatments of 180 to 200 cGy per
week. In the earlier years doses of orthovoltage radiation
were usually below 4500 cGy (eight patients). Fourteen
other patients received doses between 4500 and 6000
cGy (1-2 MeV). Three patients received total doses
greater than 6000 cGy with only photons (6100-6300
cGy). Nine patients have received combined photon and
160 MeV proton doses of 6490 to 8040 cGyE (60Co
c G y equivalent, 6oCo cGy + proton c G y X relative
biologic effectiveness [RBE]). For this clinical study, the
RBE of 160 mv protons is accepted as l.10.9 In two
cases treated with orthovoltage the details of radiation
dose were not available.
Results
Palliative radiation doses (<6000 cGy) achieved at
least partial relief of pain in all evaluable cases (n = 18).
There was no discernible difference in the palliation
achieved between those patients having radiation com-
bined with partial excision or a biopsy. Tumor regression
was observed in some cases, but it was usually incomplete
and slow.
Table 3 shows the causes of failure and status for all
patients according to treatment and radiation dose level
(sor 2 6000 cGy). Thirteen of 48 patients failed only
locally and 9 others failed with both local and distant
disease for a total incidence of local failure of 46% (22/
48). Four patients survive free of recurrent disease 8,
1 1, 13, and 20 years after total excision of sacrococcygeal
chordoma. Sixteen patients are alive 1 to 8 years after
radiation therapy, of which ten are alive with no evidcnce
of disease (NED) and six are alive with disease (AWD).
All patients treated with radiation regardless of the
extent of surgery who are free of disease progression
have received >6000 cGy except for one of the two
patients who received preoperative radiation ( 5 0 0 0 cGy).
Nine patients are NED after high-dose radiation therapy
and eight of these nine patients have received a boost
dose of radiation with 160 MeV protons. Table 4
summarizes the types of treatment for patients alive
after either radical surgery or radiation.
The survival of all patients has been analyzed accord-
ing to treatment modality (surgery alone vrrszis radiation)
No. I CHORDOMA Rich et al. 185

TABLE3. Patient Status According to Treatment and Radiation Dose Level

Cause of failure
~

Treatment
No. of Local
Surgery Radiation patients NED AWD Local -+DM DM ID Lost

Radical excision None 8 4 0 2 0 0 I I

Palliative excision None 6 0 0 4 I 0 I 0
Partial excision >6000 cGy 5 4 0 0 1 0 0 0
~ 6 0 0 0cGy 12 0 3 6 2 0 0 I
Biopsy >6000 cGy 9 5 2 0 I 0 I 0
<6000 cGy 6 0 0 I 4 0 0 I
Total excision Preop (5000 cGy) 2 I 1 0 0 0 0 0
Total 48 14 6 13 9 0 3 3

NED: no evidence of disease; AWD: alive with disease: DM: distant metastasis; ID: intercurrent death: Preop: preoperative.

by the life-table method (Fig. 3). Clearly the patients are excision and radiation (range, 2.5-7.5 years) compared
not comparable in the two groups and the results, to patients treated with radiation alone after biopsy
therefore, do not provide a basis for comparison of (range, 2.5-10.5 years). However, the follow-up periods
eficacy of surgery (n = 14) or radiation therapy (n are still brief and the tumor sizes were not the same for
= 34). The median survival is 7 and 5 years for surgery the two groups.
or radiation, respectively. Survival does not appear to
be better in the group treated with partial or subtotal Discussion
Benign and malignant chordomas have been well
TABLE
4. Surviving Patients and Type of Treatment described and both arise from rests of notochord tissue
- along the neural axis. Four percent (2/50) of the chor-
Site Treatment Status Yr
- doma cases reviewed here were benign tumors. This is
not different from the incidence of 2% cited by others
Surgery
Sacrum R-hemipelvectomy NED for ectopic remnants of notochord found on the clivus
Sacrum Partial excision NED II at routine autopsy.lO.ll These benign tumors are histo-
Radical excision segments S I -S2
Sacrum Radical excision segments SI-S2 NED 13
Sacrococcygeal Partial excision NED 20
Radical excision; coccyx and
Pro-oporolivo X R T ( n . 2 )
lower sacrum
Radiation
Therapy
Base of skull Biopsy 6 100 c C y NED 3.5
Base of skull Biopsy 6970 cGy* NED 3.0 80
Base of skull Subtotal excision 6960 cGy* NED 2.0 -I
Base of skull Subtotal excision 7660 cGy* NED 2.0
Base of skull Biopsy 7550 c G y * NED I .5
Base of skull Biopsy 6490 cGy* NED I .o
Base of skull Subtotal excision 6580 cGy* NED 1.o
Base of skull Biopsy 6580 cGy* NED 1 .o
L2 vertebra Subtotal excision 6440 cGy: two AWD 8
more operations. specimen
without tumor P o r t l o l Ercirlon + X R T l n = I ? )
C2 vertebra Biopsy 7600 cGy* NED 2.5
T6 venebra Subtotal excision for recurrent AWD 1.5 Blopry + X R T ( n = IS)
tumor 4000 cGy*
T4 vertebra Subtotal excision 5040 cGy AWD 1.5
Sacrum Preoperative RT 5000 cGy; NED 3.5 0' I I l.1 I 1 I 1
radical excision 4 0 - 12 16 20 24
Sacrum Preoperative RT 5000 c C y AWD 2.5 YEARS
Sacrum Biopsy 6300 cGy AWD I .o
Sacrum Biopsy I200 cGy AWD 0.5 FIG. 3. Life-table plot of survival of all patients with chordoma
according to treatment. The radiation group includes partial excision,
* Mixed photons and 160 MeV protons. 17: biopsy only, 15; preoperative radiation therapy plus total gross
R: right; RT: radiation therapy: AWD alive with diseasc; NED no removal, 2. Surgery alone includes eight patients treated radically and
evidence of disease progression. six with palliative intent.
186 CANCERJuly 1 1985
logically similar to the malignant tumors but they do
not invade surrounding structures, i.e., the distinction
between benign and malignant is not based simply upon
cellular features of the tumor. The two cases in this
study were associated with other solid tumors but a
correlation between notochord rests and other malig-
nancies has not been established.
The histologic review was designed to examine for a
correlation between one or more histologic features and
survival. No correlation was detected between cellular
pleomorphism, mitotic figures or hyperchromatic nuclei
(rarely found), and survival. In three cases, multinucle-
ated giant cells and nuclear inclusion cysts were seen
but the significance of this cytologic finding is unclear.
Reports from earlier literature can be found that both
upp port^*^^ and deny13 the correlation of histologic ob-
servations with prognosis.
The only histologic correlation with survival in this
study was the presence of chondroid elements in the
tumor. Chordomas with chondroid elements tend to
follow a relatively indolent course. A report from the
Mayo Clinic3 described average survival of 15.8 years
in 19 patients with chondroid chordoma arising in the
base of the skull as compared with 4.1 years for 36
patients with pure chordoma. In our series, the patients
with chondroid chordomas generally had a longer du-
ration of symptoms and were treated only by surgery.
The relatively low biologic aggressiveness of these tumors
is reflected in the survival (not disease-free) of 4.8 to 8.7
years for the chondroid variant in these patients treated
by surgery alone. In contrast, the maximum survival for
the ten patients with nonchondroid chordoma of the
base of the skull treated by partial excision and postop
erative irradiation is 3.7 years.
Although chondroid chordomas apparently behave
less aggressively and have a long natural history, the
course of disease ultimately leads to tumor recurrence
and death. We therefore recommend the use of radiation
therapy for these tumors when surgical removal is
incomplete.
In our study the reticulin stain was found to be useful
in differentiating chordomas from other tumors, partic-
ularly adenocarcinoma. In two excluded cases the clinical
course was rapidly fatal and the original diagnosis of
chordoma was questionable. In these two cases the
histologic appearance was more consistent with adeno-
carcinoma and reticulin could not be found, whereas in
all other 22 cases studied that had a clinical course
consistent with chordoma reticulin was observed. There
is supporting evidence for this histologic finding from
in vitro studies that chordoma cells produce extracellular
r e t i c ~ l i n . ' The
~ ~ ' ~positive staining reaction found in this
study differs from a previous report that found the
reticulin stain of no value in the diagnosis of chordoma.'6
Chordoma is considered a relatively radiation-resistant
Vol. 56
tumor. Although there have been occasional cases of
rapid regression13*'7-20 total doses of radiation of 6000
to 7000 cGy, fractionated at 200 cGy per day, appear
to be required to achieve worthwhile palliation. Tumor
eradication requires doses in excess of 6500 to 7000
cGy, but this can only be achieved in selected circum-
stances when treatment volumes are small and include
minimal amounts of normal tissue. In this series we
have been able to achieve such dose levels in nine
patients treated by combinations of photon and proton
beam technique for lesions in the base of the skull
(seven patients, doses from 6500 to 7550 c G y in 38
fractions) and the second cervical vertebrae (two patients,
7660 to 8040 cGy in 37 to 38 fractions). Eight of these
patients are free of disease from 10 months to 3.0 years;
to date there has been no radiation morbidity.
In our opinion, surgery has been and remains the
primary treatment modality where it is technically fea-
sible. Sufficient studies exist with long enough follow-
up which demonstrate that cure by surgery can be
achieved with good functional results in patients with
sacrococcygeal t ~ r n o r s . ~ * ~ .Th
~ *e *role
* ~ ' for
* ~ ~radiation
therapy is less defined, but appears useful in combination
with surgery where margins are involved with tumor or
in small inoperable tumors.
Our approach has been modified by the use of com-
puterized tomography, which now provides much more
precise determination of tumor extent. A carefully
planned effort between the surgeon and radiation ther-
apist is likely to improve the local control by optimizing
the planned utilization of both modalities from the
beginning of treatment. Based on our study and a review
of the literature the following recommendations are
suggested to maximize the results of radical surgery,
high-dose radiation therapy, or the combination of these
two modalities.
Sacrum
Total extirpation alone where this is technically feasible
offers the best chance for cure. In selected cases where
tumor lies close to or involves the second sacral segments,
treatment with a combination of surgery and radiation
is indicated because surgical margins will be close or
there may be cut-through of the tumor. There is major
morbidity associated with removal of both S2 or partial
SI sacral segments, e.g., pelvic instability, lumbar descent,
and anal or bladder dysfunction. In an effort to reduce
this morbidity, two .yatients have been treated with
preoperative radiation followed by sacrococcygeal resec-
tion. Microscopic tumor was present at the resection
margins in both patients: one patient has recurred
locally and the other is NED at 48 months since
treatment.
Repeated palliative operations have sometimes been
successful in prolonging survival. In our series one
No. 1 CHORDOMA
patient lived 25 years with eleven local recurrences that
were treated by surgery alone. Patients with large recur-
rent and primary tumors of the sacrum treated with
high-dose radiation therapy often achieved a useful
palliation and survived 2.5 to 10.5 years.
Vertebra and Base of Skull
In these sites total surgical extirpation is unlikely
except for very small lesions because of severe anatomic
constraints. Postoperative radiation is recommended in
all cases with residual disease. In nine cases, high-dose
precision radiotherapy has been used for nonresectable
or only partially resectable lesions with the proton beam
and results to date are satisfactory (Table 4).’
Survival longer than 10 years in patients with chor-
domas has been reported.1*3*5-73.21.23*24 Average survival
ranges from 5.0 to 6.4 years and survival rates of 50%
at 5 years have been reported by the Mayo Clinic.3The
best survival figures have been achieved for those patients
whose lesions could be treated by radical surgery, some
patients reportedly living free of disease 18 to 20 years.
Four patients in our series are alive 8 to 20 years after
radical excision of a sacrococcygeal chordoma. The
survival at 5 years of the patients treated with surgery
alone is 71%. The patients treated with radiation had a
50% survival at 5 years, and although some patients
survived as long as 11 years, nearly all eventually die
with recurrence.
In conclusion, this review indicates that for large
lesions treatment by surgery with or without radiation
may achieve an occasional long-term disease-free sur-
vival. For small lesions that are not resectable, high-
dose precision therapy can achieve good results in some
patients. Our preference is to employ surgery where
feasible and to combine this with radiation therapy in
all situations where there is concern that the surgical
margin will not be ample.
REFERENCES
I . Faust DB, Gilmore HR Jr, Mudgett CS. Chordomata: A review
of the literature with report of a sacrococcygeal case. Ann Intern Med
1944; 21:678-698.
- Rich et ul. I87
2. Gentil F, Coley B. Sacrococcygeal chordoma. Ann Surg 1948;
127432-455.
3. Heffelfinger MJ, Dahlin DC,MacCarthy CS, Beabout JW. Chor-
domas and cartilagenous tumors at the skull base. Cancer 1973; 23:
4 10-420.
4. Honvitz T. Chordal ectopic and its possible relation to chordomas.
Arch Pathol 1941; 31:354-362.
5 . Mabray RE. Chordoma: A study of 150 cases. Am J Cancer
1935; 25:501-517.
6. Hsieh CK, Hsieh HH. Roentgenologic study of sacrococcygeal
chordoma. Radiology 1936; 27:lOl-108.
7. Kamrin RP, Potanos JN, Pool VL. An evaluation of the diagnosis
and treatment of chordoma. J Neurol Neurosurg Psychiar 1964; 27:
157-165.
8. Ariel IM, Verdu C. Chordoma: An analysis of twenty cases
treated over a twenty year span. J Surg Oncol 1975; 7:27-44.
9. Suit HD, Goitein M, Munzenrider J et al. Definitive radiation
therapy for chordoma and chondrosarcoma of base of skull and
cervical spine. J Neurosurg 1982; 56:377-385.
10. Congdon C. Benign and malignant chordomas: A clinico-
anatomical study of twenty-two cases. Am J Palhul 1952; 28:793-821.
I I . Russel H. Rubinstein DS. Textbook of Neuropathology and
Tumors of the Nervous System. ed. 2. 1963; 222.
12. Godtfriedson E. Eye and nerve symptoms in connection with
cranial chordomas. Acta Ophthalmol 1943; 2 1~224-236.
13. Popper JL, King AB. Chordoma: Experience with thirteen cases.
J Neurosurg 1952; 9:139-163.
14. Fu YS, Pritchett PS, Young HF. Tissue culture study of a
sacrococcygeal chordoma with further ultrastructural study. Acta Neu-
ropathol (Berl) 1975; 23:223-225.
15. Horten BC, Montague SR. In vitro characteristics of a sacrococ-
cygeal chordoma maintained in tissue and organ culture systems. Acra
Neuroparhol (Berl) 1976; 35: 13-25.
16. Crawford T. Staining reactions of chordoma. J CIin Parhol
1958: I I : I 10-1 13.
17. Andrews JR. Spheno-occipital chordoma of unusual radiosen-
sitivity. Radiology 1942; 39~478-479.
18. Pearlman AW, Friedman M. Radical radiation therapy of
chordoma. Am J Roentgenol 1970 108(2):333-341.
19. Tewfik HH, McGinnis WL, Nordstrom DG, Latourette HB.
Chordoma: Evaluation of clinical behavior and treatment modalities.
Int J Radiar Biol Oncol Phys 1977; 2:959-962.
20. Wang CC, James AE Jr. Chordoma: Brief review of the literature
and report of a case with widespread metastases. Cancer 1968; 22:
166- 167.
2 I. Rosenquist H. Saltzman G. Sacrococcygeal and vertebral chor-
doma and their treatment. Acta Radio1 1959; 52:177-193.
22. Gray SW, Singhabhandin B, Smith RA, Skandalikis JE. Sacro-
coccygeal chordoma: Report of a case and review of the literature.
Surgery 1975; 78(5):573-582.
23. Dahlin DC, MacCarthy CS. Chordoma. Cancer 1952; 5:1170-
1178.
24. Higinbottom NL, Phillips RF, Farr HW, Husto HO. Chordoma
thirty-five year study at Memorial Hospital. Cancer 1967: 201841-
1850.