ORIGINAL ARTICLE

MALIGNANT FIBROUS HISTIOCYTOMA OF THE HEAD AND

NECK REGION
David W. Clark, MD,1,2 Brian A. Moore, MD,3 Shreyaskumar R. Patel, MD,4
B. Ashleigh Guadagnolo, MD,5 Dianna B. Roberts, PhD,1 Erich M. Sturgis, MD, MPH1,6
1
Department of Head and Neck Surgery, The University of Texas M. D. Anderson Cancer Center, Houston, Texas.
E-mail: esturgis@mdanderson.org
2
Department of OtolaryngologyHead and Neck Surgery, The University of Texas Health Science Center, Houston, Texas
3
Department of OtolaryngologyHead and Neck Surgery, Eglin Hospital, Eglin Air Force Base, Florida
4
Department of Sarcoma Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
5
Department of Radiation Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
6
Department of Epidemiology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas

Accepted 23 February 2010

Published online 13 July 2010 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/hed.21449

Abstract: Background. Malignant brous histiocytoma

(MFH) is 1 of the most common soft tissue sarcomas in the
head and neck.
Methods. We conducted a retrospective review of patients
with MFH of the head and neck region at a large multidiscipli-
nary cancer center between 1973 and 2005.
Results. Ninety-ve patients were included in the study.
The median age at diagnosis was 53 years (range, 390
years); 69% of the patients were men. The parotid or neck was
the most common subsite (35%), and 23% of the cases were
associated with prior radiation exposure. Although there were
no signicant differences in the distribution of age, sex, year of
presentation, tumor location, size, local extension, or treatment
between patients with and without prior radiation exposure,
those with radiation-associated tumors were more likely to
have positive or unclear surgical margins (p .037). With a
median follow-up of 34 months, 32 (39%) of the 83 patients
treated at M. D. Anderson Cancer Center with curative intent
had a recurrence (isolated local recurrence in 18, isolated dis-
tant recurrence in 8, both local and distant recurrence in 5,
and regional recurrence in 1). For patients treated at our insti-
tution with curative intent, 5-year overall, disease-free, and dis-
ease-specic survival rates were 55%, 44%, and 69%,
respectively. Prior radiation exposure and positive margins
were associated with worse survival.
Conclusion. MFH of the head and neck region is often
aggressive and characterized by local and/or distant recur-
rence and poor survival. Radiation-associated tumors seem to
have an especially poor prognosis. Based on a signicant
body of literature, multidisciplinary evaluation and treatment of
such high-grade sarcomas is encouraged. V C 2010 Wiley
Periodicals, Inc. Head Neck 33: 303308, 2011
Keywords: sarcoma; malignant brous histiocytoma; head and
neck neoplasms; radiation-associated malignancies
Correspondence to: E. M. Sturgis
V
C 2010 Wiley Periodicals, Inc.
Malignant brous histiocytoma (MFH) is a soft tis-
sue sarcoma that was rst described in 1964.1 Histor-
ically, it is the most common soft tissue sarcoma.2,3
Sarcomas can arise because of inherited mutations in
tumor suppressor genes, and also because of environ-
mental exposures; the most common environmental
etiology is exposure to ionizing radiation.4 In fact,
MFH seems to be the most common sarcoma diagno-
sis in patients with prior radiation exposure in the
head and neck region.5 The cell of origin for MFH
was originally thought to be broblastic progenitors.6
However, diagnosis of MFH has undergone scrutiny
as a unique pathologic entity, and today many con-
sider MFH to be synonymous with undifferentiated
pleomorphic soft tissue sarcoma.7
Most MFHs are considered high-grade, and their
classic behavior is to recur locally. Regional metasta-
ses are extremely rare, but distant metastases are
common. Approximately 3% to 10% of cases of MFH
occur in the head and neck region.4 However,
although MFH is 1 of the more common soft tissue
sarcomas in the head and neck region, it remains a
rare head and neck malignancy.4 For most MFHs of
the head and neck region, surgery is the mainstay of
treatment, and adjuvant chemotherapy and radio-
therapy seem to be critical for successful treatment.
However, a consistently effective strategy for MFHs
of the head and neck remains elusive in part because,
given the low numbers of patients with this disease,
prospective randomized trials of treatment are dif-
cult to conduct.3 Furthermore, radiation-associated
MFH of the head and neck region seems to be associ-
ated with a particularly grave prognosis.8
The purpose of this study was to review the demo-
graphic and clinical characteristics at presentation
and the general outcomes of patients with MFH of
the head and neck region presenting to a large multi-
disciplinary cancer center.

Malignant Fibrous Histiocytoma of the Head and Neck HEAD & NECKDOI 10.1002/hed March 2011 303
PATIENTS AND METHODS
After institutional review board approval of this ret-
rospective review, M. D. Anderson Cancer Centers
tumor registry was searched for all patients with an
MFH involving the head and neck region seen at our
institution between 1973 and 2005. One hundred
twenty patients were identied. Of these, 25 patients
were excluded owing to a different pathologic diagno-
sis, a nonhead and neck primary tumor site, or
insufcient documentation.
A diagnosis of MFH was considered radiation-asso-
ciated if the patient had previously received radiation
to the entire body or to the head and neck region specif-
ically. Each patients disease status at presentation to
our institution was classied as incident, incompletely
treated, or recurrent. Patients were considered to have
incident disease if they presented to our institution af-
ter imaging studies or a biopsy only. Patients were con-
sidered to have incompletely treated disease if they
presented to our institution after recent surgical resec-
tion and had no evidence of gross residual disease on
imaging studies at the primary tumor site but required
adjuvant therapy. One patient who presented after
recent surgical resection with no evidence of gross dis-
ease on imaging studies but who was not given any ad-
juvant therapy was grouped with those classied as
incident disease. Patients were considered to have
recurrent disease if they presented to our institution
with evidence of recurrent disease after completing
treatment and being free of disease for at least 3
months.
Information on demographics, primary tumor site,
histologic subtype, disease extent, grade, treatment,
and outcomes were extracted from the patients medi-
cal records. Demographic information included the
patients age at rst diagnosis of MFH and their sex.
The primary tumor site was classied as scalp or face,
neck or parotid, upper aerodigestive tract (oral cavity,
oropharynx, hypopharynx, and larynx), or skull base
or orbit (including nasal cavity, paranasal sinuses, and
nasopharynx). In 44 patients, the tumor size was docu-
mented on pathology reports and was dened as the
maximum tumor dimension recorded on gross exami-
nation of the specimen. In 19 patients for whom size
was not documented in the pathology report, size was
determined from CT or MRI studies (10 patients) or
physical examination (9 patients). Tumor size could
not be determined in the remaining 32 patients. The
patients medical records were also reviewed for evi-
dence of local extension of the tumor into the surround-
ing bone, vessel, or nerve. In addition, for patients with
metastatic disease at presentation to our institution,
the involved distant sites were recorded.
Treatment data collected included initial therapy
administered at our institution and, for patients who
underwent surgery, margin status. Margin status was
classied as unclear for 6 patients in which the pa-
thologist did not clearly document that the margins
were negative or positive. Disease-specic outcome
304 Malignant Fibrous Histiocytoma of the Head and Neck
data collected included the type of recurrence (local,
regional, or distant), if any, and the date and disease
status at the last clinic visit. For overall and disease-
specic survival, follow-up time was calculated from
the date of initial diagnosis to the date of the last
clinic visit or death, and for disease-free survival, fol-
low-up time was calculated from the date of comple-
tion of treatment to the date of recurrence or the last
clinic visit or death. The chi-square test was used to
compare categorical variables between groups, and
the Fisher exact test was used when 1 or more sub-
groups had fewer than 10 observations. Kaplan
Meier estimates and the log-rank test were used to
compare survival curves. Statistical signicance was
preset at p .05, and all tests were 2-sided.
RESULTS
Demographic and clinical characteristics of the 95
patients with MFH of the head and neck region are
presented in Table 1. The mean age at diagnosis was
52.0 years (median, 53 years; range, 390 years). The
male:female ratio was approximately 2:1. Consistent
with the overall demographics of patients seen at our
institution, most patients (84%) were non-Hispanic
whites; 6% were Hispanic Americans, 4% African
Americans, and 6% of other groups. Approximately
half of the patients were treated before 1990, and
approximately three-fourths of the patients presented
with incident MFH or shortly after initial surgical
resection (incompletely treated). Twenty-two patients
(23%) had a history of irradiation of the head and neck,
and there were no signicant differences in the distri-
butions of age, sex, ethnicity, year of presentation, or
disease status at presentation between those with and
without a history of radiation exposure (Table 1). Rea-
sons for previous radiation was for treatment of: lym-
phoma (6 patients), sinus/skull base malignancies (6
patients), squamous cell carcinoma of the laryngophar-
ynx (3 patients), squamous and basal cell carcinomas
of the skin (2 patients), and 1 each for salivary cancer,
seminoma, adenoidal hypertrophy, breast cancer, and
unclear reason. The median time from radiation to
MFH diagnosis was 10 years (mean, 15.2 years; range,
338 years). Twenty-eight patients (29%) had a prior
malignancy other than nonmelanoma skin cancer, and
3 patients had undergone radiotherapy for retinoblas-
toma as a child.
The most common presenting symptom was a
mass, which was present in the vast majority of
patients; pain, paresthesias, dysphagia, and nasal
obstruction were also common. Similarly, the over-
whelmingly most common nding on physical exami-
nation was a mass; the most common locations, in
decreasing order of frequency, were parotid or neck,
scalp or face, and anterior skull base or orbit. Tumor
grade was specied for only 42 patients, and 29 of
these had high-grade tumors. Tumor size could be
documented for 63 patients, and of these
HEAD & NECKDOI 10.1002/hed March 2011
 Table 1. Clinical characteristics of patients with MFH of the head and neck region.

Radiation-associated
Total (n 95) Yes (n 22) No (n 73)
Variable No. (%) No. (%) No. (%) p value

Age .587
<53 y 47 (49.5) 12 (54.5) 35 (47.9)

53 y 48 (50.5) 10 (45.5) 38 (52.1)
Sex .599
Male 66 (69.5) 14 (63.6) 52 (71.2)
Female 29 (30.5) 8 (36.3) 21 (28.8)
Year of presentation .984
19731989 45 (47.4) 12 (54.5) 33 (45.2)
19902005 50 (52.6) 10 (45.5) 40 (54.8)
Disease status at presentation to M.D. Anderson Cancer Center .612
Incident* 64 (67.4) 17 (77.3) 47 (64.4)
Incompletely treated 7 (7.4) 1 (4.5) 6 (8.2)
Recurrent 24 (25.2) 4 (18.2) 20 (27.4)
Tumor location .651
Scalp or face 21 (22.1) 4 (18.2) 17 (23.2)
Neck or parotid 33 (34.7) 11 (50) 22 (30.1)
Upper aerodigestive tract 13 (13.7) 2 (9.1) 11 (15.1)
Anterior skull base or orbit 21 (22.1) 4 (18.2) 17 (23.2)
Lateral skull base or ear 7 (7.36) 1 (4.5) 6 (8.2)
Tumor grade .853
Low 6 (14.3) 1 (11.1) 5 (15.2)
Intermediate 7 (16.7) 2 (22.2) 5 (15.2)
High 29 (69.0) 6 (66.7) 23 (69.7)
Tumor size .162
<5 cm 49 (77.8) 13 (92.9) 36 (73.5)

5 cm 14 (22.2) 1 (7.1) 13 (26.5)
Treatment at M. D. Anderson Cancer Center .131
Surgery only 26 (28.9) 9 (42.9) 17 (24.6)
Surgery XRT 23 (25.6) 3 (14.3) 20 (29.0)
Surgery chemo 12 (13.3) 5 (23.8) 7 (10.1)
Surgery XRT chemo 15 (16.7) 1 (4.8) 14 (20.3)
Postoperative XRT  chemo 7 (7.8) 1 (4.8) 6 (8.7)
Chemo only 7 (7.8) 2 (9.5) 5 (7.2)
Surgical margins .037
Negative 54 (71.1) 9 (50.0) 45 (77.6)
Positive or unclear 22 (28.9) 9 (50.0) 13 (22.4)
Local extension .796
None documented 67 (60.5) 16 (72.7) 51 (69.9)
Bone, vessel, or nerve 28 (39.5) 6 (27.3) 22 (30.1)
Abbreviations: chemo, chemotherapy; XRT, radiotherapy.
*One patient presented after recent surgery without clinical evidence of disease and required no additional treatment.

Grade was not specied for 53 patients.

Size was not available for 32 patients.

Five patients received no treatment at M. D. Anderson Cancer Center.

Nineteen patients did not undergo surgery at M. D. Anderson Cancer Center.

approximately 1 in 5 had tumors of 5 cm or larger.
Radiation-associated tumors tended to be found in the
parotid or neck and to be smaller tumors, though
these associations were not statistically signicant
(Table 1).
In the 76 patients who underwent resection at
our institution and for whom detailed pathologic in-
formation was available, positive or unclear margins
were common (particularly for radiation-associated
tumors), as was local extension to bone (25 patients),
vessel (1 patient), or nerve (3 patients; Table 1).
Of the 95 patients who presented to M. D. Ander-
son Cancer Center, 90 underwent treatment at our
institution, whereas 5 were seen in consultation only.
Seventy-six patients underwent surgical resection
with (n 50) or without (n 26) adjuvant therapy
(Table 1). Three patients received preoperative radio-
therapy, and 42 patients received postoperative
adjuvant radiotherapy. Seventeen patients received
neoadjuvant chemotherapy, 14 patients received
chemotherapy concomitant with postoperative radio-
therapy, and 7 patients received chemotherapy as sole
(palliative) treatment.
At a median follow-up time of 34 months (range,
1355 months), 32 (39%) of the 83 patients treated
with curative intent at our institution had a recur-
rence. Most patients (18; 56%) had isolated local
recurrences, 8 (25%) had isolated distant recurrences,
5 (16%) had synchronous local and distant recur-
rences, and 1 had a regional recurrence. Crude local

Malignant Fibrous Histiocytoma of the Head and Neck HEAD & NECKDOI 10.1002/hed March 2011 305
FIGURE 1. Overall (solid line), disease-free (dotted line), and
disease-specic (dashed line) survival of all patients with malig-
nant brous histiocytoma (MFH) of the head and neck region.

recurrence rates were more common in the tumors

with positive/unclear margins (59%) compared to
those with negative margins (20%), and all isolated
distant recurrences occurred in those with negative
margins. The lungs were overwhelmingly the most
common site of distant metastases. The median time
to recurrence was 9 months (range, 164 months; me-
dian, 9.1 months for those with local recurrences and
5.8 months for those with distant recurrences). Sec-
ond primary tumors developed in 3 patients and were
as follows: osteosarcoma, melanoma, and lymphoma.
For the 83 patients who received treatment with
curative intent at our institution, the 5-year overall
survival rate was 55%, the 5-year disease-free sur-
vival rate was 44%, and the 5-year disease-specic
survival rate was 69% (Figure 1). Twenty-ve
patients have died with MFH and 21 patients died of
other causes. Patients with radiation-associated MFH
had signicantly worse 5-year overall survival
than those with no history of radiation exposure (Fig-
ure 2A; 36% vs 62%; p .028). Worse overall survival
was also associated with positive or unclear resection
margins (p .004), and our ndings suggested that
worse overall survival was also associated with tumor
location other than the parotid or neck (p .087),
larger tumor size (p .088), and local extension (p
.064), although these associations were not statisti-
cally signicant. Whereas patients with radiation-
associated tumors also had worse 5-year disease-free
survival, this difference was not signicant (Figure FIGURE 2. (A) Overall survival of patients with malignant -
2B; 30% vs 48%; p .167). Worse disease-free sur- brous histiocytoma (MFH) of the head and neck region with
vival was associated with high tumor grade (p .008) (dashed line) and without (solid line) history of radiation expo-
and with tumor location other than the parotid or sure to this region. (B) Disease-free survival of patients with
neck (p .010), and there was a trend for association MFH of the head and neck region with (dashed line) and with-
of worse disease-free survival with positive or unclear out (solid line) history of radiation exposure to this region. (C)
Disease-specic survival of patients with MFH of the head and
margins (p .083) and local extension (p .063).
neck region with (dashed line) and without (solid line) history of
Whereas patients with radiation-associated tumors
radiation exposure to this region.
had worse 5-year disease-specic survival, this differ-
ence was not signicant (Figure 2C; 65% vs 70%; p

306 Malignant Fibrous Histiocytoma of the Head and Neck HEAD & NECKDOI 10.1002/hed March 2011
.727). Worse disease-specic survival was associated
with positive or unclear margins (p .017), and there
was a trend for association of worse disease-specic
survival with local extension (p .067).
DISCUSSION
Historically, MFH is the most common subtype of soft
tissue sarcoma, accounting for up to 40% of the
tumors in some series.2,3,9,10 However, although MFH
is 1 of the more common soft tissue sarcomas in the
head and neck region, it remains a rare malignancy
in this region.4 This study was a review of demo-
graphics and clinical characteristics of 1 of the largest
series of patients with MFH of the head and neck
region reported to date. Additionally, this study sug-
gests (similar to previously published ndings) that
radiation-associated MFHs are associated with a
worse prognosis.
Since it was rst described in 1964, MFH has pro-
ven to be a difcult clinical entity to dene.1,6 Recent
changes in nomenclature after a systematic reevalua-
tion of the cytologic characteristics of soft tissue sar-
comas have led to the reclassication of MFH as
equivalent to undifferentiated pleomorphic sarcomas,
thus making MFH a diagnosis of exclusion.7,11 The
diagnostic difculty is illustrated by studies in which
specimens initially classied as MFH were often
reclassied as other sarcomas, typically brosar-
coma.10,12 Rather than reecting inaccurate initial
histopathologic classication, these reclassications
point to the advancement in cytopathologic evaluation
techniques, such as immunohistochemistry and cyto-
genetics, which are useful in identifying markers of
tumor differentiation.9 We have not histologically
reevaluated these tumors nor the larger sample of un-
differentiated pleomorphic sarcomas at our institu-
tion, and our ndings/conclusions are in part limited
by these potential misclassications.
Main determinants of negative outcomes among
patients with MFH include positive surgical margins,
tumors located in the head and neck region, tumor
size
5 cm, deep location, and high stage.2,10 Whereas
our sample size precluded meaningful multivariate
analysis and control of confounding, we also found
worse outcomes for patients with tumors having posi-
tive margins, larger size, higher grade, local exten-
sion, or location other than neck/parotid.
Although no single cause of MFH has been con-
sistently documented, MFH is commonly associated
with prior radiotherapy.5 MFH is 1 of the most com-
mon radiation-associated sarcomas, occurring in both
bone and soft tissue and accounting for almost 50%
of radiation-associated soft tissue sarcomas.13 The
head and neck region is a common site for the occur-
rence of radiation-associated MFH, possibly owing to
the frequent use of radiotherapy in the treatment of
many head and neck malignancies.13 Almost 1 in 4
patients in our series reported a history of radiation
exposure in the head and neck region. Poor outcomes
Malignant Fibrous Histiocytoma of the Head and Neck
are typical for such radiation-associated sarcomas,14
and our ndings, though not uniformly signicant, are
consistent with those of prior reports. Most likely due
to limitations in the sample size reported here, the
lack of statistical difference in disease-free and dis-
ease-specic survival for radiation-associated MFH
cannot be interpreted to mean there is no difference.
In fact, we found in addition to a signicantly lower
overall survival, an absolute 18% lower disease-free
survival and an absolute 5% lower disease-specic sur-
vival for those with radiation-associated MFH. Prior
radiation exposure may merely be an effect modier
for subsequent positive margins (as suggested by Table
1) and thus poor outcomes, but the critical point is
that treating clinicians should be particularly wary of
a patient presenting with MFH that is associated with
prior radiation; and in so doing, the clinician should
consider the options of preoperative treatment and
modications to the surgical plan.
In conclusion, radiation-associated MFH accounts
for a signicant proportion of MFH arising in the
head and neck region, and these radiation-associated
tumors may confer an even worse prognosis. A retro-
spective analysis such as this cannot reasonable
assess whether one treatment is better than another
because of inherent biases in whom those with more
aggressive tumors received more aggressive treat-
ments. In fact, we found no signicant differences in
overall, disease-free, or disease-specic survivals
between those patients treated by surgery only as
compared with those patients treated by surgery with
adjuvant therapies (data not shown). Furthermore, it
is difcult to conduct prospective randomized trials of
optimal therapeutic regimens for MFH of the head
and neck region given the low numbers of patients
with this disease. Whereas complete surgical resec-
tion remains the primary therapy for MFH of the
head and neck region, given the high rates of positive
margins and local recurrence, adjuvant radiotherapy
must be considered in all patients, either preopera-
tively or postoperatively. Given the risk of disease
recurrence and in particular distant metastases,
neoadjuvant and/or adjuvant chemotherapy should be
considered in most patients, but scientic support for
this approach will most likely have to come from
broad studies of soft tissue sarcomas of all body sites.
Surgical excision alone is feasible for the rare MFH of
the head and neck region with favorable size, loca-
tion, and histology, but because of the high risk for
local and distant recurrence and death, up-front mul-
tidisciplinary evaluation and treatment planning is
critical to successful treatment of MFH of the head
and neck region.
Acknowledgment. The authors wish to thank
Ms. Stephanie P. Deming for assistance with manu-
script editing.
HEAD & NECKDOI 10.1002/hed March 2011 307

REFERENCES
1. OBrien JE, Stout AP. Malignant brous xanthomas. Cancer
1964;17:14451455.
2. Zagars GK, Ballo MT, Pisters PW, et al. Prognostic factors for
patients with localized soft-tissue sarcoma treated with conserva-
tion surgery and radiation therapy: an analysis of 1225 patients.
Cancer 2003;97:25302543.
3. Asahi T, Kurimoto M, Kawaguchi M, Yamamoto N, Sato S, Endo
S. Malignant brous histiocytoma originating at the site of pre-
vious fronto-temporal craniotomy. J Clin Neurosci 2002;9:704
708.
4. Sturgis EM, Potter BO. Sarcomas of the head and neck region.
Curr Opin Oncol 2003;15:239252.
5. Ko JY, Chen CL, Lui LT, Hsu MM. Radiation-induced malignant
brous histiocytoma in patients with nasopharyngeal carcinoma.
Arch Otolaryngol Head Neck Surg 1996;122:535538.
6. Enzinger FM, Weiss SW. Malignant brohistiocytic tumors. In: Soft tis-
sue tumors, 3rd ed. St. Louis: Mosby; 2001. pp 535569.
7. Fletcher CD. The evolving classication of soft tissue tumours:
an update based on the new WHO classication. Histopathology
2006;48:312.
308 Malignant Fibrous Histiocytoma of the Head and Neck
8. Wang CP, Chang YL, Ting LL, Yang TL, Ko JY, Lou PJ. Malig-
nant brous histiocytoma of the sinonasal tract. Head Neck
2009;31:8593.
9. Dei Tos AP. Classication of pleomorphic sarcomas: where are
we now? Histopathology 2006;48: 5162.
10. Oda Y, Tamiya S, Oshiro Y, et al. Reassessment and clinicopa-
thological prognostic factors of malignant brous histiocytoma
of soft parts. Pathol Int 2002;52: 595606.
11. Al-Agha OM, Igbokwe AA. Malignant brous histiocytoma:
between the past and the present. Arch Pathol Lab Med 2008;
132:10301035.
12. Nakayama R, Nemoto T, Takahashi H, et al. Gene expression
analysis of soft tissue sarcomas: characterization and reclassi-
cation of malignant brous histiocytoma. Mod Pathol 2007;20:
749759.
13. Sheppard DG, Libshitz HI. Post-radiation sarcomas: a review of
the clinical and imaging features in 63 cases. Clin Radiol 2001;
56:2229.
14. Chang SM, Barker FG 2nd, Larson DA, Bollen AW, Prados MD.
Sarcomas subsequent to cranial irradiation. Neurosurgery 1995;
36:685690.
HEAD & NECKDOI 10.1002/hed March 2011