hr. _I. Radidon Oncology Bid. Phyr., Vol. 21, pp.

1501-1508 0360-3016/91 13al + al

printed in the U.S.A. All rights reserved Copyright 0 1991 Pergamon Rcss plc

??Clinical Original Contribution

EWING’S SARCOMA: LOCAL TUMOR CONTROL AND PATTERNS OF FAILURE

FOLLOWING LIMITED-VOLUME RADIATION THERAPY

YOSHIO ARAI, M.D., LARRY E. KUN, M.D., M. TERESA BROOKS, M.D., DIANE L. FAIFCLOUGH,
Dr. P.H., JAMES FONTANESI, M.D., WILLIAM H. MEYER, M.D., F. ANN HAYES, M.D.,
ELIZABETHTHOMPSON,M.D. AND BHASKARN. RAO, M.D.

Departmentsof Radiation Oncology, DiagnosticImaging, Biostatistics, Hematology/Oncology,Surgery, St. Jude Children’s Research
Hospital and the Section of Radiation Oncology, Departments of Radiology and Pediatrics, University of Tennessee College of
Medicine, Memphis, TN

Sixty children with localized osseous Ewing’s sarcoma were treated between 1978 and 1988 with induction che-
motherapy (cyclophosphamide, adriamycin), irradiation and/or surgery, and 10 months of maintenance che-
motherapy (cyclophosphamide, adriamycin, dactinomycin, vincristine). Following induction chemotherapy, 43
patients received primary radiation therapy to lhnlted radiation volumes defined by post-chemotherapy resid-
ual soft tissue tumor extension and initial osseous tumor extent. Irradiation was defined as low dose at 30-36
Gy (median 35 Gy) for 31 cases with objective response to induction chemotherapy and high dose at 50-60 Gy
(median 50.4 Gy) for 12 patients with poor response to induction chemotherapy or with tumors 28 cm. Over-
all event-free survival at 5 years is 59% and local tumor control is 68%. Initial failures have been local (12),
simultaneous local and distant failures (7), and distant (6). In the surgical resection group, 14 patients had
complete resection without radiation therapy, and 3 patients had microscopic residual plus 3-l Gy; 100%
local control has been maintained. In 43 patients with primary radiation therapy group, local tumor control is
58% @ = MM). Despite limited radiation volume, 18/19 local failures occurred centrally within the bone, well
within the radiation volume. Imaging response to induction chemotherapy predicted local tumor control in the
radiation therapy group: 62% with complete response/partial response versus 17% with no response/progres-
sive disease @ < 0.01). Local tumor control related strongly to primary tumor size in the radiation therapy
group; among 31 cases receiving 35 Gy, local tumor control is 90% for lesions <8 cm versus 52% for tumors
r8cm@ = .054). The central pattern of local failure in this experience suggests the effectiveness of limited
radiation volume. The overall local tumor control rate following the tested dose level of 35 Gy appears to be
inadequate, although results in selected cases with tumors <8 cm in greatest tumor dimension indicate poten-
tial effkacy in a yet limited experience.

Ewing’s sarcoma, Radiation therapy, Radiation therapy dose, Radiation volume.

INTRODUCTION
The management of Ewing’s sarcoma involves radiation
therapy and/or surgery to control the primary tumor in con-
junction with systemic chemotherapy (13, 15, 16, 19). Re-
cent controversy has focussed on the relative contributions
of surgery and radiation therapy to local tumor control and
ultimate survival (1, 2, 7, 16, 17, 19, 20). To evaluate
specific components of a combined modality regimen often
including selected use of surgery or radiation therapy, nu-
merous clinical studies have sought to identify tumor pa-
rameters that affect treatment outcome. Primary tumor
size, primary tumor site, and the pre-chemotherapy soft
tissue tumor extension have been correlated with local tu-
mor control and ultimate survival (2, 7, 8, 10, 11, 12, 17).
Response to induction chemotherapy has also been related
to primary disease control (4, 14). Radiation dose and ra-
diation volume have not significantly affected local tumor
control in major clinical series; recent series addressing ra-
diation volume adequacy suggest the use of wide local
fields to encompass the primary tumor in contradistinction
to earlier recommendations that full bone irradiation was
necessary (9, 10, 15, 17).
We report the results of a prospective treatment proto-
col for localized osseous Ewing’s sarcoma which included
induction chemotherapy, a surgical option following initial
tumor response, and radiation therapy using limited radia-
tion volume and, in selected cases, reduced radiation dose

Presented at 32nd Annual Scientific Meeting of the American Lauderdale, P.O. Box 318, Memphis, TN 38101 U.S.A.
Society for Therapeutic Radiology and Oncology, Miami Beach, Acknowledgements-The authors express their appreciation to
FL, 15-19 October, 1990. Carole Obst for manuscript preparation and List Ogle for assis-
Reprint requests to: Larry E. Kun, M.D., Department of Radi- tance in data management.
ation Oncology, St. Jude Children’s Research Hospital, 332 N. Accepted for publication 10 May 1991.
1501
1502 I. J. Radiation Oncology 0 Biology 0 Physics November 1991, Volume 21, Number 6

- Local Evahmtlo”
and
. Local Therapy

- CYClADR Eklpsy “CRlDAC CYUADR - “CWDAC

- Resectlo”
Cyckphosphamde “vlcrlslne
(150 mgmlm21day p.0 x 7) (I.5 mgnlm2 I V. weekly x 11)

AdrlamyCln Dactlnomyane
(35 m@m’ I.“. M day 8) (1.5 mgnlm2 IV q 2 weeks x 6)

- Radlatior Therapy
150 180 cGy. Sllweek

Fig. 1. Treatment schema.

(8). This report emphasizes (a) a detailed analysis of out-
come relative to primary tumor size and pre-chemotherapy
soft tissue tumor extension determined by computerized to-
mography, (b) the impact of initial tumor response to in-
duction chemotherapy, (c) radiation dose and radiation
volume relative to local tumor control and survival, and (d)
the effect of selected surgery upon outcome measures.
METHODS AND MATERIALS
Sixty consecutive patients with localized osseous Ew-
ing’s sarcoma (ES) were treated on a St. Jude Children’s
Research Hospital investigational protocol between 1978
and 1988. The trial was approved by the institutional com-
mittee for clinical investigations, and appropriate informed
consent was obtained in all instances. Analysis was per-
formed in October 1989. Patients ranged in age from 2.9
to 22 years at diagnosis with a median of 15.3 years. There
were 41 boys and 19 girls. Minimum follow-up of surviv-
ing patients is 1 year, ranging to 10 years with median of
5 years post diagnosis; follow-up intervals exceeded 2
years in 92% of cases.
Evaluation of tumor
A systematic, retrospective review of tumor imaging
was performed jointly by two of the authors (TB, YA), in-
cluding all pre-treatment radiographs and computerized to-
mography (CT) scans. Tumor size was measured by CT
criteria. Five patients without CT before resection were se-
lectively included in the analysis based on the measure-
ment obtained by plain film and pathology measurement.
The greatest tumor dimension (GTD) included the osseous
and soft tissue tumor extension, measured as a single di-
mension in the plane of maximum tumor area. Tumors
whose size were marginal at 8 cm were re-reviewed and
grouped, respectively. Soft tissue tumor extension (STE) at
diagnosis was recorded as the greatest radial dimension
perpendicular to the cortex of the involved bone. Tumor
volume (TVL) was estimated as the product of 3-dimen-
sional measurement utilizing previously reported formulas:
Vol = a X b X c X 0.52 for “ellipsoidal tumors” with
significant soft tissue components and Vol = a x b x c
X 0.785 for cylindrical tumors with little or no soft tissue
disease (a: the maximal tumor dimension in a single CT
plane; b: tumor dimension perpendicular to a; c: the longi-
tudinal maximal tumor dimension) (7, 9).
Response to chemotherapy was graded by review of CT
studies obtained following induction chemotherapy. Re-
sponses were based upon radiographic reduction in soft tis-
sue tumor extension (measured by single greatest dimension
perpendicular to the cortex of the bone), and coded as
complete response (CR: complete disappearance of STE),
partial response (PR: more than 50% regression of STE),
no response (NR: O-50% reduction of STE), or progressive
disease (PD: any increase in STE).
Chemotherapy
Induction chemotherapy (IDCTX) consisted of five
courses of cyclophosphamide and vincristine (CYC/ADR):
cyclophosphamide (150 mg/M’/day orally for 7 days) and
adriamycin (35 mg/M2 I.V. on day 8) (Fig. 1). Courses
were begun on weeks 1, 3, 5, 8, and 11. Following assess-
ment of response on weeks 12-13, local therapy was initi-
ated in conjunction with vincristine and dactinomycin
(VCRDAC): vincristine (1.5 mg/M2 I.V. weekly for 11
doses) and dactinomycin (1.5 mg/M2 I.V. every 2 weeks
for 6 doses) on weeks 14-24. Dactinomycin was withheld
during radiation therapy if significant epithelial reactions
were noted. Subsequent therapy included 6 additional
courses of sequential CYC/ADR as above at 3-week inter-
vals between weeks 27-43 and 6 courses of VClUDAC as
above every 2 weeks between weeks 47-57. Carmustine
(BCNU), 50 mg/M2 every 6 weeks for two doses, was
given following IDCTX only in the first 14 patients.
Surgery
Surgical tumor resection with negative microscopic
margins was achieved at diagnosis in three cases. Initial
surgery was confined to biopsy or limited resection in 57
children. Following induction chemotherapy, 36 cases un
 Ewing’s sarcoma: Limited-volume radiation therapy ??Y. ARAIet al. 1503

Table 1. Event free survival and local tumor control at 5 years All other patients received primary irradiation. Prior to
by tumor parameters 1985, low dose irradiation (LDRT: 30-36 Gy, median 35
Gy) was given to patients who achieved CR, those with PR
5 yr 5 yr or NR shown to have no residual tumor at biopsy follow-
EFS LTC
p % ing IDCTX, and cases with no measurable soft tissue tu-
No. % P
mor extension at diagnosis. Irradiation was delivered in
Primary tumor site* 17-29 fractions (median 20). High dose irradiation
Axial (HDRT: 50-60 Gy, median 50.4 Gy) was given to patients
Pelvis with proven residual tumor after PR or NR and to those
Proximal long bone with PD; treatment was delivered in 2849 fractions (me-
Distal long bone dian 29). After 1985, HDRT was given, in addition, to all
Greatest tumor dimension?
patients with primary tumor size at diagnosis exceeding 8
< 8 cm 18 81 0.057 94 0.008
cm in greatest tumor dimension, regardless of response to
2 8 cm 41 49 56
IDCTX. Radiation therapy used 4 or 10 Mev photons at
Tumor volumej:
< 1OOcc 18 74 0.063 82 0.10 150-180 cGy per fraction, 5 times per week. Fraction sizes
2 1OOcc 38 49 59 of 125 cGy and 200 cGy were utilized in 2 respective
Pre-chemotherapy soft tissue tumor extension$ LDRT cases; a hyperfractionated regimen of 120 cGy bid
< 2.0 cm 17 75 0.074 82 0.089 was used in one HDRT case.
2 2.0 cm 37 49 56

*No. = 60: All patients. Statistical analysis

tNo. = 59: Excluding one patient who presented with patho-
logic fracture. Five patients without computerized tomography
(CT) before surgery were included by measurement obtained by
plain X-ray and/or pathology measurement.
$No. = 56: Excluding one patient who presented with patho-
logic fracture and three patients without CT before resection. Two
patients without CT before surgery were included by measure-
ment obtained by plain X-ray and/or pathology measurement.
$No. = 54: Excluding one patient who presented with patho-
logic fracture and five patients without CT before resection.
derwent surgery. In 14 cases, tumor resection was
performed, with total tumor resection (i.e., negative mi-
croscopic margins) in 11 and incomplete removal (i.e.,
positive microscopic margins) in 3. Biopsy only was per-
formed in 22.
The surgical resection group includes the 17 cases
treated by initial (3) or delayed resection (14) with nega-
tive microscopic margins or only microscopically positive
margins.
Radiation therapy
Radiation treatment was given between weeks 14 and
19. The radiation tumor volume was defined by the initial
@e-chemotherapy) osseous tumor extent and residual soft
tissue tumor extension based upon imaging studies at com-
pletion of induction chemotherapy. The radiation volume
included 3 cm margins beyond the defined radiation tumor
volume. Simulation and verification films were reviewed
in conjunction with initial and post-induction chemother-
apy imaging.
Patients with total tumor resection received no irradia-
tion (n = 14). Those with positive microscopic margins
following incomplete removal received 3541 Gy (n = 3).
Survival, local tumor control, local failure, and distant
metastasis were calculated from the time of diagnosis. As-
sessment of local tumor status was rigorously pursued in
cases demonstrating distant metastasis. Simultaneous local
and distant failure was coded when both local failure and
distant metastasis occurred within 3 months of the initial
event signaling relapse. Duration of local tumor control
rate at 5 years (LTC) and event-free survival rate at 5 years
(EFS) were estimated by the Kaplan-Meier method. The
Log Rank test was used to compare the survival of each
group. A Wilcox test was also used to analyze the impact
of tumor measurement parameters in the surgical and radi-
ation therapy groups.
RESULTS
Overall, EFS at 5 years is 59% and LTC is 68%. Treat-
ment failures have occurred in 25 patients. Initial relapse
sites were local (n = 12), simultaneous local and distant
(n = 7), or distant (n = 6). Relapse occurred within 2
years in 16 of the 25 treatment failures.
Primary tumor site and tumor measurement parameters
are shown in Table 1; axial primaries were classified as rib
(n = lo), vertebrae (n = 5) and skull (n = 2). Neither
EFS nor LTC varied significantly with primary tumor site.
EFS is better for smaller tumors measured by greatest tu-
mor dimension, tumor volume, or pre-chemotherapy soft
tissue tumor extension; differences approached statistical
significance in each parameter. Local tumor control is sig-
nificantly better in tumors <8 cm in greatest dimension
(Table 1); local tumor control approaches superior levels in
the other tumor measurement parameters.
Imaging response to IDCTX was evaluable in 45 cases.
Objective response was noted in 37 (82%): CR in 29 (64%)
1504 I. J. Radiation Oncology 0 Biology 0 Physics November 1991, Volume 2 1, Number 6

Table 2. Outcome and patterns of initial failure mor control has been maintained in all surgical resection
group cases. Surgical resection was selected in cases with
Surgical Radiation dispensable bones; comparison of tumor parameters sug-
resection therapy
gests the inclusion of more limited tumors in the surgical
group group
(No. = 17) (No. = 43) group: greatest tumor dimension 3.0-14.5 cm (median =
8.5 cm) versus 5.0-27.0 cm (median 10.1 cm) in the radi-
Estimated event free 75% -p=o.o44- 53% ation therapy group (p = 0.15). All 16 pelvic primaries
survival at 5 years were included in the RT group.
Estimated local tumor 100% -p=o.O04- 58% In 43 radiation therapy group patients, EFS at 5 years
control at 5 years
is 53% and local tumor control is 58% (Table 2). The ini-
Patterns of failure
(initial relapse sites) tial site of failure was local in 19 of 22 failures: 12 as iso-
Local 0 12 lated local failure (O-37 months post-diagnosis, median 20
Simultaneous 0 7 months) and 7 with simultaneous distant metastasis (7-67
Distant 3 3 months post-diagnosis, median 12 months). Distant me-
tastasis alone occurred in three patients with local tumor
control at 29-64 months; two later demonstrated local fail-
ure 16 and 20 months after distant metastasis.
and PR in 8 (18%). NR or PD was apparent in 8 (18%). Analysis of EFS and LTC for the radiation therapy
Fifteen patients were not evaluable for imaging response group by tumor measurement parameters is shown in Ta-
because of absence of STE (n = 9), or lack of measurable ble 3. LTC is significantly superior in tumors <8 cm in
disease secondary to surgically obscured STE (n = 5), or greatest tumor dimension; results suggest better control in
pathologic fracture (n = 1). Patients with CR/PR to those with limited soft tissue tumor extension and target
IDCTX (n = 37) showed EFS of 64%, statistically supe- volume as well. No significant differences were noted by
rior to the 13% EFS in the group of NR/PD cases primary tumor site as an independent parameter.
(p <O.OOl). Within the radiation therapy group, patients with CR/PR
Histologic evaluation following IDCTX in 36 patients to induction chemotherapy had EFS at 5 years of 54%
disclosed identifiable tumor in soft tissue and/or perios- compared to 17% for those with NIVPD (p < .002). Local
teum in 14 at biopsy or resection; no soft tissue tumor was tumor control following CR/PR after induction chemother-
found in 22. The presence of residual tumor did not affect apy is 62%, compared to 17% for cases with NR/PD @ <
the EFS or LTC in the entire group or when analyzed ,001). The outcome of patients with no measurable soft
within the surgical or radiation subsets. tissue extension at diagnosis was similar to those included
The impact of local treatment modality is shown in Ta- in the CR/PR group.
ble 2. In 17 cases selected for surgical resection, EFS is Analysis of local tumor control by tumor measurement
75% at 5 years; results are statistically better than the 53% parameters and radiation dose is shown in Table 4. Depen-
EFS in the irradiation alone group @ = .044). Local tu- dent upon the parameter determining size, LTC in a lim-

Table 3. Event free survival and local tumor control at 5 years in radiation therapy group

5 yr 5 yr
EFS LTC
No. % P % P

Greatest tumor dimension*

< 8 cm 11 73 0.16 94 0.014
2 8 cm 31 44 44
Tumor volume*
< 1OOcc 12 67 0.11 75 0.10
2 1OOcc 30 46 49
Pre-chemotherapy soft tissue tumor extensiont
< 2.0 cm 13 69 0.089 77 0.066
2 2.0 cm 28 42 45
*No. = 42: All primary irradiation patients, excluding one patient who presented with pathologic fracture.
tNo. = 41: All primary irradiation patients, excluding one patient who presented with pathologic fracture and one patient whose soft
tissue extension was not evaluable due to surgery.
 Ewing’s sarcoma: Limited-volume radiation therapy ??Y. ARAIet al. 1505

Table 4. Local tumor control at 5 years by radiation dose and tumor parameters

Low dose irradiation High dose itradiationS

(30-36 Gy) (50-60 GY)

5 yr 5 yr
LTC LTC
No. % P No. %

All primary irradiation patients 31 64 - 12 39

Greatest tumor dimension
< 8 cm 10 90 0.054 1 -
2 8 cm 21 52 10 25
Tumor volume*
< 1OOcc 10 80 0.16 2 -
2 1OOcc 21 57 9 28
Pm-chemotherapy soft tissue tumor extensiont
< 2.0 cm 11 82 0.081 2 -
2 2.0 cm 19 52 9 28

*No. = 42: All primary irradiation patients, excluding one patient who presented with pathologic fracture.
tNo. = 41: All primary irradiation patients, excluding one patient who presented with pathologic fracture and one patient whose soft
tissue extension was not evaluable due to surgery.
fNumbers within the high dose radiation subset do not permit meaningful statistical analysis.

ited number of cases (n = 10-11) is E&90%. For the fined by initial soft tissue tumor extent in the LDRT group
LDRT cases, local tumor control appeared to be superior indicated identical local treatment control of 65% in 17
in cases with greatest tumor dimension <8 cm, as shown whose radiation volumes covered the pre-chemorherupy
in Table 4 and Figure 2. Differences in local tumor control soft tissue tumor extension and 63% in 14 whose radiation
by tumor volume and pre-chemotherapy soft tissue tumor volume had not included the pre-chemotherapy soft tissue
extension approach statistical significance. tumor extent.
The site of local failure in 18 out of 19 instances was
central, within the primary bone and well within the re- Toxicities
duced radiation volumes. A comparison of outcome based Acute radiation reactions included six cases of moist
upon retrospective assessment of radiation volume if de- radioepidermitis, one transient esophageal radioepithelitis,

1.0

1 I\
-4

I I I II I II I
c 8cm

P
A
0
0.8

0.6-
I -7
4
L_

:___.

P
I________,

0 I____L___I___J_____l;
A
2
IJI_LL______________L____________________J 8cm
T
I 0.4s
0
N

0.2s

0.0
I I I I I I I I I I r
0 1 2 3 4 5 6 7 e 9 10 11

Fig. 2. Local tumor control by greatest tumor dimensions in 31 localized osseous Ewing’s sarcoma treated by low
dose irradiation (34-36 Gy) at St. Jude Children’s Research Hospital from 1978 to 1988.
1506 I. J. Radiation Oncology 0 Biology 0 Physics
and one instance of delayed healing at the biopsy site. One
patient with maxillary sinus primary who received 35 Gy
developed a cataract 3 years post-irradiation and decreased
visual acuity to 20/50. Three cases of post-treatment frac-
tures were all associated with documented local relapse.
There were no unanticipated late effects of RT that signif-
icantly impaired function or required surgical correction
except to equalize long bone growth. No second malig-
nancy has occurred to date.
DISCUSSION
The overall event-free survival of 59% at 5 years in the
current study closely parallels results in most major series.
Single institution reports by Sailer and Bacci indicate
5-year relapse-free survival in 50% and 54%, respectively
(1, 14). Multi-institutional trials in the Intergroup Ewing’s
Sarcoma Study (IESS) and the German Cooperative Ew-
ing’s Sarcoma Study (CESS) have shown relapse-free rates
of 47% (IESS-I), 62% (IESS-2), and 55% (CESS)(3,
5, 9, 13).
The St. Jude trial investigated induction chemotherapy
and modifications of radiation therapy delivery. The high
rate of objective response (CR and PR) to induction che-
motherapy noted in the earlier report of this trial has been
maintained (8). A strong correlation between imaging re-
sponse to induction chemotherapy and outcome is noted,
with marked differences in EFS (64% following CR/PR vs
13% with NR/PD) comparable to those described in Ober-
lin et al (57% vs 9%) (14). A similar effect based upon
histologic response following induction chemotherapy in
Delepine’s report was not apparent in the present
series (4).
Local tumor control at 5 years is 68% in the reported
trial. Comparable single-institution reports indicate 64%
local disease control in Bacci et al.‘s experience and 78%
in Sailer’s review (1, 17). The local failure rates in inter-
institutional series appear to be less frequently docu-
mented: 15% in IESS-1, 9% for extremity primaries and
28% for pelvic primaries in IESS-2, and 23% in the CESS
trial (3, 5, 9, 13, 18). The difficulty in evaluating local
tumor control versus overall outcome has recently been
highlighted by Gelman et al. (6).
The present study includes a high proportion of local
failures: 19 of 25 cases (76%) failing therapy showed iso-
lated local failure or simultaneous local and distant failure.
Only reports by Brown et al. and the CESS show an ex-
cess of local over distant failures (2, 9, 18). In the IESS
studies and single center series by Bacci et al. and Sailer
et al., comparable in size and time interval to this trial,
local failures comprise only 2542% of all failures (1, 3,
5, 13, 17).
Local tumor control amongst the irradiated patients in
this report is 58%. Similar results have been noted by
Brown et al. (55%) and Sailer et al. (66%), although us-
ing more conventional radiation doses and radiation vol-
umes. The investigational use of 35 Gy in patients with
November 1991, Volume 21, Number 6
documented response to induction chemotherapy has re-
sulted in overall local tumor control in a selected subset of
only 64% (Table 4). Although a dose response has not
been demonstrated for Ewing’s sarcoma, it is difficult to
recommend building upon this low-dose trial without iden-
tifying parameters predictive of more consistent local tu-
mor control.
Tumor size appears to be a significant factor both
for local tumor control and survival. Marcus and Million
noted disease-free survival in 80% of cases with greatest
tumor dimension <8 cm compared to 10% for larger tu-
mors (11). Significant soft tissue tumor extension has also
predicted outcome (12). The CESS study confirmed a sig-
nificant correlation between primary tumor size and out-
come; using tumor volume determinations, survival was
65% for lesions cl00 ml compared to 32% for larger tu-
mors (18). In the St. Jude experience, primary tumor size
correlated strongly with EFS (81% vs 49% for tumors <8
cm vs ~8 cm, respectively) and LTC (94% vs 56%, re-
spectively). Tumor volume and pre-chemotherapy soft tis-
sue tumor extension related similarly. Among the patients
treated with 35 Gy, local tumor control has been main-
tained in 9 of 10 with lesions less than 8 cm in greatest tu-
mor dimension. Although the number of cases is limited,
the analysis suggests that lower radiation doses may be
appropriately tested in highly selected cases of osseous
Ewing’s sarcoma.
The local failure rate of 52% with 35 Gy for larger tu-
mors, although differing insignificantly from rates of 45%
in the irradiated groups reported both by Brown et al and
the CESS, is high in comparison to the 15-28% noted in
the IESS studies (2, 3, 5, 9, 13). The suggestion by Sailer
et al. and Marcus and Million that higher radiation doses
and more aggressive chemotherapy may result in improved
local tumor control for larger tumors has stimulated more
recent trials in this center (11, 17). The incidence of patho-
logic fracture at the primary tumor site after treatment in
the current study is lower than in previous reports, conjec-
turally related to initial healing in response to chemother-
apy and lower radiation dose.
Also tested in the reported trial is the use of limited ra-
diation volume. Perez et al. suggested that minor varia-
tions from the standard of full-bone irradiation for Ewing’s
sarcoma did not compromise local control when using ef-
fective systemic therapy (15). Marcus and Million have
also used target volumes based upon initial tumor extent
rather than the standard full bone coverage (11). The cur-
rent study reports the use of radiation volumes limited to
the initial osseous tumor extent and residual soft tissue tu-
mor extension following induction chemotherapy. Analysis
of local failures indicates central recurrence confined to the
irradiated bone site in 18 of 19 instances; one child had
concurrent central and marginal (scar) failure. Retrospec-
tively comparing the radiation volume to the pre-chemo-
therapy soft tissue tumor extent, we noted “marginal
adequacy” (i.e., inclusion of soft tissue tumor extension
by CT with 1 cm margin) in 17 of the 31 cases receiving
 Ewing’s sarcoma: Limited-volume radiation therapy 0 Y. ARAIer al.
LDRT. The frequency of local tumor control was not af-
fected by the radiation volume adequacy measured retro-
spectively in this manner. A trial of limited-versus-full
bone irradiation, based upon initial tumor extent, has re-
cently been concluded with the Pediatric Oncology Group.
The lack of marginal failures in the St. Jude experience
suggests that target volumes may be significantly reduced
without incurring a substantial risk of marginal recurrence.
Primary tumor site did not correlate with outcome in
this series. Other reports indicate a significant difference
favoring extremity or other central lesions over pelvic pri-
maries (1, 13, 15, 19). The correlation between large tu-
mor volume and pelvic primary has been suggested in other
series (5, 7, 19). Tumor volume alone was predictive in
the present report.
The impact of surgical resection is difficult to derive
from this analysis. Rosen et al. suggested significant im-
provement in both local tumor control and survival related
to surgery; local failure were noted only among the non-
operated cases in their report (16). Retrospective reviews
indicating improved outcome with surgery also demon-
strate the case selection process, reporting surgical inter-
vention more often in extremity tumors and smaller
primary tumor size (2, 7, 17, 20). The surgical resection
REFERENCES
1. Bacci, G.; Toni, A.; Avella, M.; Manfrini, M.; Sudanese,
A.; Ciaroni, A.; Boriani, S.; Emiliani, E.; Campanacci, M.
Long-term results in 144 localized Ewing’s Sarcoma patients
treated with combined therapy. Cancer 63: 1477-1486; 1989.
2. Brown, A. P.; Fixsen, J. A.; Plowman, P. N. Local control
of Ewing’s sarcoma: an analysis of 67 patients. Br. J. Ra-
diol. 60: 261-268; 1987.
3. Burgert, E. 0.; Nesbit, M. E.; Garnsey, L. A.; Gehan, E.
A.; Herrmann, J.; Vietti, T. J.; Cangie, A.; Tefft, M.;
Evans, R.; Thomas, P.; Askin, F. B.; Kissane, J. M.; Prit-
chard, D. J.; Neff, J.; Makley, J. T.; Gilula, L. Multimodal
therapy for the management of nonpelvic, localized Ewing’s
sarcoma of bone: an Intergroup Study IESS-II. J. Clin. On-
col. 8: 1514-1524; 1990.
4. Delepine, N.; Delepine, G.; Jasmin, C.; Desbois, J. C.;
Volsin, M. D.; Misset, J. L.; Mathe, G. Predictive value
of complete histological response (HR) to sequential cyclo-
phosphamide (CPX) and adriamycine (ADR) induction ther-
apy (IT) in Ewing’s sarcoma (ES). Proc. ASCO 6: 136;
1987.
5. Evans, R. G.; Nesbit, M. E.; Gehan, E. A.; Gamsey, L. A.;
Burgert, 0.; Vietti, T. J.; Cangir, A.; Teftt, M.; Thomas, P.;
Askin, F. B.; Kissane, J. M.; Pritchard, D. J.; Neff, J.;
Makley, J. T.; Gilula, L. Multimodal therapy for the man-
agement of localized Ewing’s sarcoma of pelvic and sacral
bones: a report from the Second Intergroup Study (IESS-II).
J. Clin. Oncol. 9:1173-1180; 1991.
6. Gelman, R.; Gelber, R.; Henderson, C.; Coleman, C. N.;
Harris, J. R. Improved methodology for analyzing local and
distant recurrence. J. Clin. Oncol. 8: 548-555; 1990.
7. Gobel, V.; Jurgens, H.; Etspuler, G.; Kemperdick, H.; Jung-
blut, R. M.; Stienen, U.; Gobel, U. Prognostic significance
of tumor volume in localized Ewing’s sarcoma of bone in
children and adolescents. J. Cancer Res. Clin. Oncol. 113:
187-191; 1987.
1507
group in the St. Jude trial included tumors that were
smaller in greatest tumor dimension and non-pelvic in lo-
cation. Systematic local tumor control and superior EFS
likely reflect both the bias of selection and treatment effi-
cacy; inclusion of 10 cases with primary rib lesions, often
with considerable soft tissue tumor extension, suggests the
effectiveness of this approach in such settings. The CESS
study also reported improved outcome following surgery;
in part, the difference between 10% failure after surgery
and 47% failure after radiation therapy alone was related to
marginal radiotherapy techniques (9, 18). Sailer er al.
noted reduction in local failure from 34% to 0 with surgery
in addition to irradiation; a significant impact upon local
tumor control showed only a marginal survival benefit from
surgery in multifactorial analysis ( 17).
The relative roles of radiation therapy and surgery in
Ewing’s sarcoma are controversial and evolving. The
present series confirms the value of primary tumor size and
response to induction chemotherapy as factors important in
analyzing, and potentially determing, local therapy. Use of
more limited radiation volume appears to be effective for
osseous Ewing’s sarcoma. Studies testing reduced radia-
tion dose or altered fractionation may be appropriate in se-
lected presentations.
8. Hayes, F. A.; Thompson, E. I.; Meyer, W. H.; Kun, L.;
Parham, D.; Rao, B.; Kumar, M.; Hancock, M.; Parvey, L.;
Magil, L.; Hustu, H. 0. Therapy for localized Ewing’s sar-
coma of bone. J. Clin. Oncol. 7: 208-213; 1989.
9. Jurgens, H.; Exner, U.; Gadner, H.; Harms, D.; Michaelis,
J.; Sauer, R.; Treuner, J.; Voute, T.; Winkelmann, W.;
Winkler, K.; Gobel, U. Multidisciplinary treatment of pri-
mary Ewing’s sarcoma of bone. A 6-year experience of a
European Cooperative Trial. Cancer 61: 23-32; 1988.
10. Marcus, R. B.; Graham-Pole, J. R.; Springfield, D. S.; Fort,
J. A.; Gross, S.; Mendenhall, N. P.; Elfenbein, G. J.;
Weiner, R. S.; Enneking, W. F.; Million, R. R. High-risk
Ewing’s sarcoma: end-intensification using autologous bone
marrow transplantation. Int. J. Radiat. Oncol. Biol. Phys.
15: 53-59; 1988.
11. Marcus, R. B.; Million, R. R. The effect of primary tumor
size on the prognosis of Ewing’s sarcoma. Int. J. Radiat.
Oncol. Biol. Phys. 10: 88; 1984.
12. Mendenhall, C. M.; Marcus, R. B.; Enneking, W. F.;
Springfield, D. S.; Thar, T. L.; Million, R. R. The prognos-
tic significance of soft tissue extension in Ewing’s sarcoma.
Cancer 51: 913-917; 1983.
13. Nesbit, M. E.; Gehan, E. A.; Burgert, E. 0.; Vietti, T. J.;
Cangir, A.; Tefft, M.; Evans, R.; Thomas, P.; Askin, F. B.;
Kissane, J. M.; Pritchard, D. J.; Herrmann, J.; Neff, J.;
Makley, J. T.; Gilula, L. Multimodal therapy for the man-
agement of primary, non-metastatic Ewing’s sarcoma of
bone: a long-term follow-up of the First Intergroup Study. J.
Clin. Oncol. 8: 1664-1674; 1990.
14. Oberlin, 0.; Patte, C.; DeMeocz, F.; LaCombe, M. J.; Bru-
nat-Mentigny, M.; Demaille, M-C.; Tron, P.; Bui, B. N.;
Lemerle, J. The response to initial chemotherapy as a prog-
nostic factor in localized Ewing’s sarcoma. Eur. J. Cancer
Clin. Oncol. 21: 463-467; 1985.
15. Perez, C. A.; Tefft, M.; Nesbit, M.; Burgert, E. 0.; Vietti,
1508 I. J. Radiation Oncology 0 Biology 0 Physics
T.; Kissane, J.; Pritchard, D. J.; Gehan, E. A. The role of
radiation therapy in the management of non-metastatic Ew-
ing’s sarcoma of bone. Report of the Intergroup Ewing’s
Sarcoma Study. Int. J. Radiat. Oncol. Biol. Phys. 7: 141-
149; 1981.
16. Rosen, G.; Caparros, B.; Nirenberg, A.; Marcove, R. C.;
Huvos, A. G.; Kosloff, C.; Lane, J.; Murphy, M. L. Ew-
ing’s sarcoma: ten-year experience with adjuvant chemother-
apy. Cancer 47: 2204-2213; 1981.
17. Sailer, S. L.; Harmon, D. C.; Mankin, H. J.; Truman, J. T.;
Suit, H. D. Ewing’s sarcoma: surgical resection as a prog-
nostic factor. Int. J. Radiat. Oncol. Biol. Phys. 15: 43-52;
1988.
November 1991, Volume 21, Number 6
18. Sauer, R.; Jurgens, H.; Burgers, J. M. V.; Dunst, J.; Haw-
licek, R.; Michaelis, J. Prognostic factors in the treatment of
Ewing’s sarcoma. The Ewing’s Sarcoma Study Group of the
German Society cf Paediatric Oncology CESS 8 1. Radiother.
Oncol. 10: 101-110; 1987.
19. Tepper, J.; Glaubiger, D.; Lichter, A.; Wackenhut, J.; Glat-
stein, E. Local control of Ewing’s sarcoma of bone with ra-
diotherapy and combination chemotherapy. Cancer 46:
1969-1973; 1980.
20. Wilkins, R. M.; Pritchard, D. J.; Burgert, E. 0.; Unni, K.
K. Ewing’s sarcoma of bone. Experience with 140 patients.
Cancer 58: 2551-2555; 1986.