Postradiation sarcoma: Morphological findings on fine-needle aspiration with clinical correlation.

Abstract

The current study was conducted to describe the clinical features and presentation, cytomorphological characteristics with histological correlation, and prognosis of patients who undergo fine-needle aspiration (FNA) for postradiation sarcoma (PRS).
A retrospective review was performed of 13 individual patients who were pooled from the FNA services of 3 academic institutions between 2001 and 2012. Cases were reviewed for the primary tumor, radiation history, latency period, and other distinguishing clinical features. The frequency of the various cytological preparations as well as the use of immunohistochemistry (IHC) on this material were reviewed. The cytopathology diagnosis was compared with the resection diagnosis, and the survival time was reviewed.
The median age of the patients was 61 years (range, 35 years-94 years) and no significant gender predilection was noted. The median latency period was 11 years (range, 5 years to > 50 years). Patients generally presented with large tumors (median, 8 cm [range, 3 cm-12 cm]), and the median survival was 14 months (range, 6 months-46 months). Nine of 13 patients died of their disease and 1 was lost to follow-up. The tumors were morphologically heterogeneous. IHC played an important role in excluding other diagnoses in those cases in which sufficient material was available.
PRS is a morphologically heterogeneous entity that can be diagnosed by FNA. It is a diagnosis of exclusion that requires a history of therapeutic radiation and often requires IHC to rule out locally recurrent malignancy. Cancer (Cancer Cytopathol) 2012. © 2012 American Cancer Society.

Full text

Page 1

Postradiation Sarcoma: Morphological Findings on
Fine-Needle Aspiration With Clinical Correlation
Matthew T. Olson, MD1; Paul E. Wakely, Jr, MD2; Kristy Weber, MD3;
Momin T. Siddiqui, MD4; and Syed Z. Ali, MD1,5
BACKGROUND: The current study was conducted to describe the clinical features and presentation, cytomorphological
characteristics with histological correlation, and prognosis of patients who undergo fine-needle aspiration (FNA) for post-
radiation sarcoma (PRS). METHODS: A retrospective review was performed of 13 individual patients who were pooled
from the FNA services of 3 academic institutions between 2001 and 2012. Cases were reviewed for the primary tumor,
radiation history, latency period, and other distinguishing clinical features. The frequency of the various cytological prepa-
rations as well as the use of immunohistochemistry (IHC) on this material were reviewed. The cytopathology diagnosis
was compared with the resection diagnosis, and the survival time was reviewed. RESULTS: The median age of the patients
was 61 years (range, 35 years-94 years) and no significant gender predilection was noted. The median latency period was
11 years (range, 5 years to > 50 years). Patients generally presented with large tumors (median, 8 cm [range, 3 cm-12
cm]), and the median survival was 14 months (range, 6 months-46 months). Nine of 13 patients died of their disease and 1
was lost to follow-up. The tumors were morphologically heterogeneous. IHC played an important role in excluding other
diagnoses in those cases in which sufficient material was available. CONCLUSIONS: PRS is a morphologically heterogene-
ous entity that can be diagnosed by FNA. It is a diagnosis of exclusion that requires a history of therapeutic radiation and
often requires IHC to rule out locally recurrent malignancy. Cancer (Cancer Cytopathol) 2012;120:351-7. V
C 2012 American
Cancer Society.
KEY WORDS: postradiation sarcoma, postirradiation sarcoma, radiation-associated sarcoma, radiation-induced sarcoma,
soft tissue sarcoma, fine-needle aspiration, cytopathology, cytomorphology, oncology.
INTRODUCTION
Postradiation sarcoma (PRS) is a nosologic entity comprised of any sarcoma that arises in an irradiated
body site after a latency period.1-5The term is interchangeable with ‘‘postirradiation sarcoma’’ and ‘‘radia-
tion-associated sarcoma.’’ Because the evolution of this phenomenon is multifactorial and poorly under-
stood, most investigators do not favor the overly simplistic term ‘‘radiation-induced sarcoma.’’6,7The
most common primary tumors for which radiotherapy is associated with PRS include breast carcinoma,
Hodgkin lymphoma, cervical carcinoma, and bone and soft tissue sarcomas.8Histologically, PRS can be
any type of sarcoma, including malignant fibrous histiocytoma, angiosarcoma, leiomyosarcoma, fibro-
sarcoma, malignant peripheral nerve sheath tumor, myxofibrosarcoma, chondrosarcoma, and osteo-
sarcoma.9-12Distinguishing PRS from spontaneous sarcoma is important because PRS tends to present at
Received: February 6, 2012; Revised: March 2, 2012; Accepted: March 9, 2012
Published online April 19, 2012 in Wiley Online Library (wileyonlinelibrary.com)
DOI: 10.1002/cncy.21200, wileyonlinelibrary.com
Corresponding author: Syed Z. Ali, MD, Department of Pathology, The Johns Hopkins Hospital, Path 406, 600 N Wolfe St, Baltimore, MD 21287;
Fax: (410) 614-9556; sali@jhmi.edu
1Department of Pathology, The Johns Hopkins Hospital, Baltimore, Maryland;
Columbus, Ohio;3Department of Orthopedic Surgery, The Johns Hopkins Hospital, Baltimore, Maryland;4Department of Pathology, Emory University
Hospital, Atlanta, Georgia;5Department of Radiology, The Johns Hopkins Hospital, Baltimore, Maryland
2Department of Pathology, Ohio State University Medical Center,
Cancer Cytopathology
October 25, 2012
351
Original Article

Page 2

a more advanced stage and behaves more aggressively
thanamorphologically
sarcoma.8,13To our knowledge, this is the first cyto-
morphologic study on a seriesof PRS.
equivalent spontaneous
MATERIALS AND METHODS
Case Selection and Information
All available electronic records from the pathology data
systems at The Johns Hopkins Hospital (7 cases), Emory
Medical Laboratories (2 cases), and Ohio State University
Medical Center (4 cases) were queried for cytological
specimens that were signed out as PRS or were derived
from tumors that were signed out as PRS on the resection
specimen. The cases identified originated between the
years 2001 and 2012. If a patient had undergone repeated
biopsies, only the first diagnosis of PRS by fine-needle
aspiration (FNA) was reviewed and included for the
purposesofthe featuresnotedatpresentation.
The history of radiotherapy and chemotherapy,
latency period, primary tumor type, and the stage of the
PRS at the time of presentation were derived from the
electronic medical record. In 1 case, the radiotherapy
history was so remote that the age of the patient was not
known definitively, so the latency period was approxi-
matedasatleast50years.Thetumorsizewasderivedfrom
the dimensions of the tumor noted at the time of resection
(10 cases), by radiology (2 cases), or by clinical examina-
tion (1 case). Of the 9 patients who died, all deaths were
due to disease. The survivaltime was calculated as the time
elapsedfromthefirstdiagnosisofPRStothedateofdeath.
For cases from The Johns Hopkins Hospital, the date of
death was either documented by hospital clinicians or
obtainedfromtheSocialSecurityDeathIndex.
Cytopathology
The FNA procedures were performed by interventional
radiologists (9 cases) and pathologists (4 cases) using
ultrasound guidance with on-site evaluation of adequacy.
Smears evaluated on-site for adequacy were prepared in
duplicate and included at least 1 air-dried slide for Diff-
Quik staining and on-site evaluation and an alcohol-fixed
slide for Papanicolaou staining for each pass as material
allowed. Additional aspirated material was diverted for
studies such as cell block, immunohistochemistry (IHC),
and cytospin as demonstrated in Table 1. IHC was
performedonthecorebiopsy(2cases),cellblock(5cases),
or destained direct smear (1 case) according to standard
procedures. After the procedure, all the slides were
previewed by a cytotechnologist and signed out by a
board-certified cytopathologist. If a core biopsy was taken
(4 of 11 cases), the specimen was interpreted separately by
an expert surgical pathologist with expertise in soft tissue
tumorsincommunication
cytopathologist.
withtheattending
RESULTS
Thirteen FNA cases from PRS were identified at 3 institu-
tions. Of these, 3 cases were aspirations performed on a
known PRS, and 10 were aspirations performed for the
purposes of establishing a first diagnosis. For 1 institution
(Johns Hopkins Hospital), the percentage of total known
PRS cases (23 cases) that underwent FNA (7 cases) is
approximately 30%. PRS cases that did not undergo FNA
most commonly included tumor sites in the bone (4
cases), bladder (2 cases), and brain (2 cases). By contrast,
only 1 tumor in this series was in the bone, and the other
12 tumors were present in soft tissue at sites close to the
radiation fields of the primary tumor. Hence, this study
Table 1. Preparations and Techniques Used for the 13 Cases of PRS in the Current Study
Diff-Quik PapanicolaouCell BlockCytospinCore
Biopsy
IHC
No. of cases
Median no. of
slides/case (range)
13
4 (3-8)
11
4 (0-10)
10
1 (0-3)
3
0 (0-2)
4
0 (0-1)
7
1 (0-13)
aa
Abbreviations: IHC, immunohistochemistry; PRS, postradiation sarcoma.
aThe median values for the number of cytospin and core needle biopsy preparations are both 0 because > 50% of the
cases did not undergo these procedures.
Original Article
352
Cancer Cytopathology
October 25, 2012

Page 3

clearly represents a subset of PRS that included tumor
sites that underwent FNA at the study institutions. The
median age of the patients at the time of the diagnosis of
PRS was 61 years (range, 35 years-94 years) with no gen-
der predilection. The median latency period was 11 years
(range, 5 years to > 50 years). Of these cases, 46% (6 of
13 cases) had a history of chemotherapy accompanying
the radiation. The primary tumors, cytopathologic diag-
noses, andresection diagnosesaregiven inTable2.
An imaging study of 1 case of PRS is shown in Fig-
ure 1. The median tumor size at the time of presentation
was 8 cm (range, 3 cm-12 cm), and 23% of patients (3 of
13 patients) presented with metastatic disease. Not
surprisingly,the overallsurvivalin thisgroupwaspoor. At
the time of last follow-up, 9 of the 13 patients (69%) had
died of disease with a median survival of 14 months
(range, 6 months-46 months). One patient was lost to
follow-upandthe3remainingpatientswere<36months
from thetimeof aspiration.
Overall, a various amount of cellularity was noted,
with the majority of cases displaying abundant material.
The common denominator was the presence of a highly
pleomorphicspindlecellpopulation.Inthecaseofepithe-
lioid angiosarcoma, there were scattered cells with round
to oval shapes and eccentric nuclei (Fig. 2). Long and
tapering cytoplasmic processes were also evident. Myxofi-
brosarcoma predominantly contained spindle cells em-
bedded in a loose myxoid stroma. Focal marked
anisonucleosis was apparent (Fig. 3), as well as occasional,
curvilinear, fine capillary vessels. Osteosarcoma displayed
pleomorphic naked nuclei or plasmacytoid cells juxta-
posed or embedded in immature osteoid matrix (Fig. 4).
Because the observation of immature osteoid matrix is
difficult to make with certainty on routine direct smears,
the cases of osteosarcoma were diagnosed as such after
resection. Malignant fibrous histiocytoma demonstrated
bizarre multinucleation, brisk abnormal mitoses, and
highly pleomorphic spindle cells focally associated with
loose matrix material (Fig. 5). A histologic section of
MFH showed sheets of large multinucleated cells with
pleomorphic nuclei and macronucleoli, abundant bizarre
mitoses,and focalnecrosis(Fig.6).
As indicated in Table 1, soft tissue tumors often
require multiple cytological modalities for proper
workup. The higher number of Diff-Quik slides reflects
the use of the Diff-Quik inspection of direct smears for
the on-site evaluation of adequacy. A cell block was made
in 10 cases (77%), and IHC was performed in 4 of these
cases. Two cases were found to have sufficient material on
core needle biopsy for IHC staining, and 1 case was
destained for IHC staining. The most common IHC
stains are listed in Table 3 and demonstrate the common
practice of staining for pancytokeratins to rule out carci-
noma in these tumors. The use of the vascular markers
cluster of differentiation 31 (CD31) and CD34 reflects
theclinicalsuspicionforangiosarcomainthisclinicalcon-
text. IHC was also used for establishing smooth muscle
markers in cases of rhabdomyosarcoma and for ruling
Table 2. Primary Tumor, Cytopathologic Diagnosis, and Resection Diagnosis of PRS
Primary Tumor Cytopathology DiagnosisResection Diagnosis
BrCA
BrCA
BrCA
Hodgkin disease
Juvenile angiofibroma
Kaposi sarcoma
Larynx SCC
Lymphoepithelial cavernous sinus tumor
PrCA
PrCA
Rhabdomyosarcoma
Synovial sarcoma
Synovial sarcoma
Atypical cells
Epithelioid angiosarcoma
PRS
MFH
Poorly differentiated SCN
MFH
High-grade MFS
SCN
Spindle and epithelioid high-grade sarcoma
MFH
High-grade PRS
High-grade sarcoma
High-grade sarcoma
Angiosarcoma
—
Grade 2 MFH
PRS osteosarcoma
MFH
PRS osteosarcoma
High-grade MFS
MFH
Spindle and epithelioid high-grade sarcoma
—
High-grade PRS
Chondroblastic osteosarcoma
Sarcoma with rhabdomyosarcomatous differentiation
a
bc
b
b
Abbreviations: BrCA, breast adenocarcinoma; MFH, malignant fibrous histiocytoma; MFS, myxofibrosarcoma; PrCA, prostate adenocarcinoma; PRS, postra-
diation sarcoma; SCC, squamous cell carcinoma; SCN, spindle cell neoplasm.
aSurgery was not performed.
bCytopathologic diagnosis was made on an aspiration sample of a known PRS.
cGrade 2 by the NCI grading system.
FNA of Postradiation Sarcoma/Olson et al
Cancer Cytopathology
October 25, 2012
353

Page 4

out local recurrences of the primary tumors with tumor-
specific markers.
DISCUSSION
PRS is a rare disease that comprises < 5% of all sarcomas;
the overall risk of PRS after radiotherapy for any reason is
estimated to be between 0.03% to 0.8% of all patients
treated with radiotherapy.10,14These statistics are compli-
cated by the variability in radiotherapy regimens and la-
tency periods; the postradiation latency interval varies
from only a few months to > 50 years.15,16Older radio-
therapy regimens comprised of low doses of orthovoltage
radiation for superficial diseases appear to be associated
with longer latency intervals compared with currently
favored radiotherapy protocols, which use higher doses of
deeply penetrating megavoltage radiation. The latency
periods with current protocols is shorter, but there does
not appear to be a difference in other parameters such as
histological grade or overall survival.17,18The incidence
ofPRSappearstoberelatedmostcloselytothedose;there
FIGURE 2. Highly pleomorphic spindle cells of an epithelioid
angiosarcoma are shown (H & E, original magnification ?
40).
FIGURE 3. The pleomorphic spindle cells and myxoid matrix
of a myxofibrosarcoma are shown (Diff-Quik stain, original
magnification ? 40).
FIGURE 1. (A) A typical plain radiograph of postradiation sar-
coma in the arm demonstrates the bony destruction of the
adjacent soft tissue tumor. (B) The T2-weighted magnetic
resonance image demonstrates an infiltrating soft tissue
mass that is hyperintense to muscle.
Original Article
354
Cancer Cytopathology
October 25, 2012

Page 5

does not appear to be a difference between orthovoltage
and megavoltageregimensifthedose is equal.10
The identification of a soft tissue tumor as PRS is
relevant to an FNA service for 2 main reasons. First, PRS
presents at a higher stage of disease and may have a poorer
prognosis than de novo sarcoma,8,13and therefore identi-
fying it can have treatment implications. Second, the con-
sideration of PRS in a patient and the subsequent
clarification of the clinical history can lead to a faster and
more accurate diagnosis with less need for ancillary test-
ing. Thus, a history of radiotherapy and malignancy are
important components of the patient record in the
workup of soft tissue neoplasms by FNA. The current
study presents the key clinical and pathological features of
patients who have presented for the first diagnosis of PRS
by FNA; to the best of our knowledge, the clinical and
pathological features of PRS in patients who present for
FNAhavenotbeendescribedinacomprehensivepublica-
tion to date. This could be due to several factors. First,
these are rare tumors; the largest studies are comprised of
a few hundred cases,12,19and the majority of others have
smaller numbers. Second, the morphological spectrum of
PRS is very broad, and therefore there is not a constella-
tion of findings that can be used to definitively identify
a tumor as PRS instead of a spontaneous sarcoma.
Third, the malignancy and radiotherapy history can be so
FIGURE 4. Pleomorphic naked nuclei, plasmacytoid cells, and
osteoid are seen in an osteosarcoma specimen (Papanicolaou
stain, original magnification ? 40).
FIGURE 5. Pleomorphic spindle cells and multinucleated cells
in a malignant fibrous histiocytoma are shown (Papanicolaou
stain, ? 40).
FIGURE 6. A histologic section of malignant fibrous histiocy-
toma from the resection specimen is shown and demon-
strates multinucleated cells, nuclear and cellular anisocytosis,
and atypical mitotic figures (H & E, original magnification ?
40).
Table 3. Most Commonly Ordered IHC Stains and Their
Frequencies
a
IHC AntigenNo. of
Cases
AE1/AE3
CD31 or CD34
S-100 protein
Cam 5.2
5
3
3
2
Abbreviations: CD31, cluster of differentiation 31; CD34, cluster of differen-
tiation 34; IHC, immunohistochemistry;
A total of 7 cases underwent IHC. Stains that were ordered only once are
not listed and include cytokeratin (CK) 7, CK20, desmin, myogenin, B-cell
lymphoma 2 (Bcl-2), mouse anti-CD99 (O13), prostate-specific antigen
(PSA), prostatic-specific acid phosphatase (PSAP), prostein (P501S), p63,
thyroid transcription factor-1 (TTF-1), leukocyte common antigen (LCA), and
renal cell carcinoma (RwCC).
FNA of Postradiation Sarcoma/Olson et al
Cancer Cytopathology
October 25, 2012
355

Page 6

remote that they are often not available at the time of
FNA or core needle biopsy, and therefore PRS may be
underreported.
The study data indicate that these cases are com-
prised of aspirates that are readily diagnosed as malignant.
Only 1 of the 13 cases received a cytopathologic diagnosis
thatwasmerelyatypicalratherthandefinitiveforthediag-
nosis of sarcoma, and the cellularity of this aspirate was
poor. No false-negative results were found. Because these
neoplasms are high grade, the diagnosis of a malignant
processishighlyreproducibleacrossthe3institutionspar-
ticipating in this study. The more difficult distinction in
this contextis theclassificationof the malignant processas
a sarcoma and more importantly as a PRS. In this regard,
the combination of the clinical history and IHC is effec-
tive.Giventhevariedradiologicalandhistologicalappear-
ances of PRS, the need to pursue a history of radiotherapy
may not be consistently apparent. Because the malignancy
history is often known at the time of FNA, awareness of
this uncommon entity is important for pursuing the treat-
ment history for any neoplasm. As the experience in the
current study demonstrates, the role of IHC in the initial
diagnosis of PRS can be useful in that it can rule out
pseudosarcomatous changes in a poorly differentiated
carcinoma with cytokeratin immunostains. However, in
the absence of highly specific antigens for sarcoma, the
valueofIHCliesmore inits exclusionarypower.
Imaging studies are known to have a high degree of
interpatientvariabilityduetotheheterogeneous histologi-
cal subtypes of PRS that are possible.20,21The current
study was no exception. The radiographic appearance of
these neoplasms ranged from that of a soft tissue sarcoma
to osteosarcoma with ossification. This detail reflects a
key observation in the tumor presentation, specifically
that the malignancy and radiotherapy history in combina-
tion with the cytopathologic findings is the most appro-
priate means of arriving at the diagnosis of PRS. The
radiographic studies are more important for staging than
fordeterminingthe originaldiagnosis.
Although 2 of the cases were typical of the classic
connection between angiosarcoma and radiotherapy for
breast cancer, the numbers were insufficient to make any
association between primary tumor type and the mor-
phology of the PRS. Moreover, the observation that one
of the cases of irradiated breast cancer led to the diagnosis
of MFH serves as a reminder of the extremely heterogene-
ous nature of PRS. Thus, radiotherapy history and the
exclusion of carcinoma remain the most important fea-
tures in the workup of these cases. There is currently no
standardized method for predicting which primary
tumorsleadtoaparticularmorphologyofPRS.
The current study has selection bias both because
therealnumberofPRScasesmostlikelyexceedsthenum-
ber of recognized PRS cases and because the number of
PRS cases clearly exceeds the number of FNA procedures
on PRS. The bias reflects practice limitations of FNA. For
example, PRS in the brain has been described,7,20,22but
this site is not aspirated and therefore no cases exist in this
series. Similarly, osteosarcoma in the bone after irradia-
tion is the oldest example of PRS.4Although the current
series had 3 cases of osteosarcoma, only 1 was aspirated
from bone. These examples underlie the importance of
characterizing patients with PRS who undergo FNA so
that the identification of other potential cases of PRS can
be identifiedmorepromptlyin thefuture.
PRS cases have a higher likelihood of IHC features
with poor prognostic implications such as p53 and a high
proliferative index.23Although not conclusive proof, these
findings, in conjunction with the worse clinical behavior of
PRS, indicate that these tumors may be genetically distinct
from the primary sarcomas they resemble morphologically.
This is an important consideration given that the patients
who develop PRS have already received maximal doses of
chemotherapy and radiotherapy. Therefore, they receive less
extensivetherapythanpatientswithdenovosarcoma.Thus,
the search for molecular diagnostic modalities and targeted
therapies could provide helpful improvements. Recently,
investigatorshaverevealedtargetablemutationsinpostradia-
tion angiosarcoma such as fms-like tyrosine kinase receptor
(FLT) and kinase insert domain receptor (a type III receptor
tyrosine kinase) (KDR),24which may provide a rationale for
better therapeutic guidance in these tumors in the future. It
has yet to be shown whether the tumors accessible to FNA
harbor these mutations. Given the relative ease with which
FNAmaterialcanbeobtainedandusedformutationalanal-
ysis, aspiration may prove to be a more prominent diagnos-
tic modality in PRS if these findings are maintained in
subsequentinvestigations.
Conclusions
To the best of our knowledge, the current series is the first
of its kind to detail the cases of PRS that undergo FNA.
Original Article
356
Cancer Cytopathology
October 25, 2012

Page 7

Commensurate with the findings of other publications
that are comprised of histopathological findings, the PRS
cases in the current study were heterogeneous with regard
to their clinical, radiographic, and histological presenta-
tions. Thus, the findings reinforce the concept that a
history of cancer and radiotherapy should be part of the
clinical knowledgebase. When feasible,IHCwasuseful in
establishingthisdiagnosisof exclusion.
FUNDING SUPPORT
No specific funding was disclosed.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
REFERENCES
1. Cahan WG, Woodard HQ, et al. Sarcoma arising in irradiated
bone; report of 11 cases. Cancer. 1948;1:3-29.
Sabin FR, Doan CA, Forkner CE. The production of osteogenic
sarcomata and the effects on lymph nodes and bone marrow of
intravenous injections of radium chloride and mesothorium in
rabbits. J Exp Med. 1932;56:267-289.
Martland HS. Occupational poisoning in manufacture of luminous
watch dials. JAMA. 1929;92:466-473.
Martland HS, Humphries R. Osteogenic sarcoma in dial painters
using luminous paint. Arch Pathol. 1929;7:406-417.
Martland HS, Conlon P. Some unrecognized dangers in the use and
handling of radioactive substances. JAMA. 1925;85:1769-1776.
Kondziolka D. Radiation-associated sarcoma? Neurosurgery. 2011;69:
E786; author reply 787.
Loeffler JS, Niemierko A, Chapman PH. Second tumors after
radiosurgery: tip of the iceberg or a bump in the road? Neurosur-
gery. 2003;52:1436-1440; discussion 1440-1442.
Robinson E, Bar-Deroma R, Rennert G, Neugut AI. A comparison
of the clinical characteristics of second primary and single primary
sarcoma: a population based study. J Surg Oncol. 1992;50:263-266.
des Guetz G, Chapelier A, Mosseri V, Dorval T, Asselain B, Pouil-
lart P. Postirradiation sarcoma: clinicopathologic features and role
2.
3.
4.
5.
6.
7.
8.
9.
of chemotherapy in the treatment strategy. Sarcoma. 2009;2009:
764379.
Mark RJ, Poen J, Tran LM, Fu YS, Selch MT, Parker RG. Postir-
radiation sarcomas. A single-institution study and review of the lit-
erature. Cancer. 1994;73:2653-2662.
Brady MS, Garfein CF, Petrek JA, Brennan MF. Post-treatment
sarcoma in breast cancer patients. Ann Surg Oncol. 1994;1:66-72.
Brady MS, Gaynor JJ, Brennan MF. Radiation-associated sarcoma
of bone and soft tissue. Arch Surg. 1992;127:1379-1385.
Gladdy RA, Qin L-X, Moraco N, et al. Do radiation-associated
soft tissue sarcomas have the same prognosis as sporadic soft tissue
sarcomas? J Clin Oncol. 2010;28:2064-2069.
Amendola BE, Amendola MA, McClatchey KD, Miller CH Jr.
Radiation-associated sarcoma: a review of 23 patients with postra-
diation sarcoma over a 50-year period. Am J Clin Oncol. 1989;
12:411-415.
Wiklund TA, Blomqvist CP, Raty J, Elomaa I, Rissanen P, Mietti-
nen M. Postirradiation sarcoma. Analysis of a nationwide cancer
registry material. Cancer. 1991;68:524-531.
Davidson T, Westbury G, Harmer CL. Radiation-induced soft-tis-
sue sarcoma. Br J Surg. 1986;73:308-309.
Laskin WB, Silverman TA, Enzinger FM. Postradiation soft tissue
sarcomas. An analysis of 53 cases. Cancer. 1988;62:2330-2340.
Le Vu B, de Vathaire F, Shamsaldin A, et al. Radiation dose,
chemotherapy and risk of osteosarcoma after solid tumours during
childhood. Int J Cancer. 1998;77:370-377.
Robinson E, Neugut AI, Wylie P. Clinical aspects of postirradia-
tion sarcomas. J Natl Cancer Inst. 1988;80:233-240.
Debnam JM, Guha-Thakurta N, Mahfouz YM, et al. Radiation-
associated head and neck sarcomas: spectrum of imaging findings.
Oral Oncol. 2012;48:155-161.
Sheppard DG, Libshitz HI. Post-radiation sarcomas: a review of
the clinical and imaging features in 63 cases. Clin Radiol. 2001;56:
22-29.
Zampieri P, Zorat PL, Mingrino S, Soattin GB. Radiation-associ-
ated cerebral gliomas. A report of two cases and review of the liter-
ature. J Neurosurg Sci. 1989;33:271-279.
McHugh JB, Thomas DG, Herman JM, et al. Primary versus radi-
ation-associated craniofacial osteosarcoma: biologic and clinicopath-
ologic comparisons. Cancer. 2006;107:554-562.
Guo T, Zhang L, Chang NE, Singer S, Maki RG, Antonescu CR.
Consistent MYC and FLT4 gene amplification in radiation-
induced angiosarcoma but not in other radiation-associated atypical
vascular lesions. Genes Chromosomes Cancer. 2011;50:25-33.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
FNA of Postradiation Sarcoma/Olson et al
Cancer Cytopathology
October 25, 2012
357