Journal of Surgical Research 160, 184–189 (2010)

doi:10.1016/j.jss.2009.05.030

ASSOCIATION FOR ACADEMIC SURGERY

Pediatric Solid Tumors and Second Malignancies: Characteristics
and Survival Outcomes
Vanitha Vasudevan, M.D.,* Michael C. Cheung, M.D.,* Relin Yang, M.D.,* Ying Zhuge, M.D.,*
Anne C. Fischer, M.D., Ph.D.,† Leonidas G. Koniaris, M.D.,* and Juan E. Sola, M.D.*,1
*DeWitt Daughtry Family Department of Surgery, University of Miami Miller School of Medicine, Miami, Florida;
and †Department of Surgery, University of Texas Southwestern Medical Center, Dallas, Texas

Submitted for publication February 23, 2009

Key Words: pediatrics; multiple cancers; solid tumor;

Background. To examine the incidence, character-
istics, and outcomes for second malignancies following
the diagnosis of a primary solid tumor in pediatric
patients.
Methods. The Surveillance, Epidemiology, and End
Results (SEER) database was queried from 1973 to
2005, excluding recurrences, in patients <20 y.
Results. A total of 31,685 cases of pediatric solid
malignancies were identified. Overall, 177 patients
were diagnosed with a unique second malignancy
before the age 20 (0.56%) The mean follow-up was for
8.5 y (2 mo–30.8 y). Mean age at diagnosis of the primary
tumor was 7.7 y. The most common primary malignan-
cies were CNS tumors (22.5%), followed by soft tissue
sarcoma (15.8%), retinoblastoma (14.1%), and bone
tumors (13%). Hematologic malignancies (35.5%) were
the most common second malignancies noted in the
cohort, followed by bone tumors (18%) and soft tissue
sarcomas (15%). Hematologic malignancies had
a shorter latency (3.1 y) compared with solid second
tumors (11.6 y). The overall 10-y survival for the entire
cohort was 41.5%. For most tumor categories, develop-
ment of a secondary malignancy was associated with
lower 5- and 10-y survival than expected.
Conclusions. CNS tumors, retinoblastoma, and soft
tissue sarcomas in children are the most common solid
primary tumors, with an increased risk of a second
malignancy. Leukemia is the most common second ma-
lignancy seen in pediatric solid tumors. Second malig-
nancies are associated with significantly reduced
survival rates compared with the general childhood
cancer population. Ó 2010 Elsevier Inc. All rights reserved.
outcomes; SEER.
INTRODUCTION
The treatment of pediatric cancer is becoming
increasingly successful with the aggressive use of mul-
timodal therapy, yielding an overall 5-y survival of
78%. This has brought forward the emerging spectrum
of long-term sequelae among cancer survivors, of which
the most life threatening are subsequent second malig-
nancies, congestive cardiac failure, and pulmonary
complications [1]. The development of a second malig-
nancy in childhood has been shown to cause a signifi-
cant increase in excess mortality among cancer
survivors. Previous studies have indicated that the
cumulative incidence of a second neoplasm is 3.2% at
20 y from the original diagnosis [2]. This low frequency
necessitates the use of large, multicentric studies or
population-based cancer registries to achieve adequate
sample sizes and follow-up.
We utilized the SEER population based cancer regis-
try to examine the incidence, characteristics, and out-
comes of pediatric second malignancies following the
diagnosis of a solid tumor.
MATERIAL AND METHODS
Case Selection

The surveillance, epidemiology, and end results (SEER) April 2008

1
To whom correspondence and reprint requests should be
addressed at DeWitt Daughtry Family Department of Surgery, Uni-
versity of Miami Miller School of Medicine, 1611 N.W. 12th Avenue,
East Tower 3019, Miami, FL 33136. E-mail: jsola@med.miami.edu.
release was used to identify all malignancies developing in patients
under the age of 20 y. From this set of patients, ID numbers that
appeared more than once were selected. These ID numbers repre-
sented patients who were diagnosed with a primary malignancy,
and then subsequently diagnosed with another, unique malignancy

0022-4804/09 $36.00 184

Ó 2010 Elsevier Inc. All rights reserved.
 VASUDEVAN ET AL.: PEDIATRIC SOLID TUMORS AND SECOND MALIGNANCIES
at a later time point before the age of 20 y. Primary and secondary
cases were identified using date of diagnosis provided in the SEER
database. From this dataset, all primary hematologic malignancies
were removed, thus leaving only primary solid organ tumors. Second
hematologic malignancies were kept in the analysis. This dataset was
then reviewed, and patients with second tumors that were not unique
from the primary tumors were removed, thus eliminating patients
with tumor recurrence from the dataset.
Survival Calculations and Statistical Analysis
Only the percentages based on available data for each individual
variable are given. Patients with missing data were excluded from
each respective univariate and multivariate analysis. Statistical
analysis was performed with SPSS Statistical Package version 15.0
(SPSS Inc., Chicago, IL). A P value less than 0.05 was considered sig-
nificant. Correlations between categorical variables were made using
the c2 test. Survival was calculated by the Kaplan-Meier method, and
was abstracted from the time of the initial diagnosis to the date of last
contact (or the date of death, if the patient was deceased). Observed
survival represents survival of patients from our dataset, while
expected survival time represents that of patients under the age of
20 with the same tumors calculated from SEER.
RESULTS
A total of 31,685 children with solid malignancies
diagnosed before 20 y of age were identified from the
SEER data and followed up for an average of 8.5 y
(range 2 mo to 30.8 y, mean 8.5 y). The age-adjusted
to year 2000 annual incidence for nonhematologic
malignancies was 9.29 cases per 100,000 population,
with an annual percent change of 0.73 over the study
period (P < 0.05). A total of 177 patients (0.56%) were
noted to develop a second malignancy with a unique
histology before the age of 20 y. The demographics
and clinical characteristics of this cohort are described
in Table 1. The mean age at diagnosis of the primary
malignancy was 7.7 y (range 0–18 y). The male to
female ratio was 1.2:1. The majority of the patients
were non-Hispanic and white. The overall 10-y survival
of the cohort was 41.1%.
Table 2 shows the histologic characteristics of the pri-
mary solid malignancy. The most common solid tumors
were central nervous system (CNS) tumors (22.6%),
sarcomas (15.8%), and retinoblastoma (14.1%). Retino-
blastoma was the most common solid tumor for patients
less than 1 y of age. Between 1 and 5 y of age, the major-
ity of tumors were CNS tumors, retinoblastoma, neuro-
blastoma, and Wilms tumor. Between the ages of 5 and
15 years, malignant bone tumors, including osteosar-
coma and Ewing’s sarcoma were the most common solid
malignancy. After 15 y of age, carcinomas were the pre-
dominant primary tumor type.
Overall, 183 second malignancies were observed in
this cohort of 177 patients before they attained the
age of 20 y, including the six patients who developed
a third primary malignancy. Table 3 shows the age dis-
tribution at the time of diagnosis of secondary tumors.
185
The incidence of second malignancies was noted to
increase temporally with increase in age; 44% of second
malignancies developed in the 15–19 y age group. Inter-
estingly, 9.2% of second tumors were seen in patient-
s < 5years of age, most likely reflecting a genetic
susceptibility to cancer or familial syndromes predis-
posing to multiple cancers in these patients.
Table 3 also shows the histologic subtypes of the sec-
ond malignancies observed in the cohort. The most com-
mon histologic subtype in children was hematologic
tumors (35.5%), most of which were leukemia, followed
by malignant bone tumors (18.6%), and sarcomas
(14.8%). Analysis of age distribution among patients
with these tumors subtypes revealed that bone tumors
were more common in the 5–15 y age group, while sarco-
mas were more common in the 10–19 y age group. The
incidence of leukemia increased with the age of the child,
suggesting the possible effects of chemotherapy, radio-
therapy, or both. The incidence of carcinomas as a group
was highest in the 15–19 y age group, corresponding to
a background increase in the incidence of carcinoma in
this age group. The most common histology in children
less than 1 y of age was pineoblastoma (n ¼ 3), following
a primary retinoblastoma, again indicating the genetic
predisposition to cancer as a possible association.
An attempt was made to study the effect of the type of
primary solid malignancy and the risk of a subsequent
second malignancy (Table 4). CNS, bone tumors, and
sarcomas developed leukemia as the most common sec-
ond malignancy, while retinoblastoma and neuroblas-
toma were typically associated with subsequent solid
malignancies like malignant bone tumors and sarcomas.
We studied the relation of time from the diagnosis of
primary tumor to the development of a second tumor (la-
tent period) with the number and type of second tumors
shown in Table 5. Overall, 13.6% of second malignancies
occurred within 1 y of the first tumor, and 48% devel-
oped within 1 to 4 y from diagnosis of primary tumor.
Table 5 also shows the histologic subtypes and their dis-
tribution according to the latent period. Approximately
85% of the hematologic second malignancies were seen
within the first 5 y from the diagnosis of primary solid
tumor (mean latent period of 3.1 y). In contrast, solid
second tumors had a mean latency of 11.6 y.
The association of the latent period and radiotherapy
as a treatment modality for the primary tumor was stud-
ied using c2 analysis and revealed that radiotherapy was
seen to have a significant association with the develop-
ment of a second malignancy after 5 y of the primary solid
tumor (60%) compared with within 5 y (38%), P < 0.005.
Survival Statistics
Cumulative survival were calculated for the various
histologic subtypes of second malignancies using the
186 JOURNAL OF SURGICAL RESEARCH: VOL. 160, NO. 2, MAY 15, 2010

TABLE 1 TABLE 3
Demographics and Clinical Characteristics Histological Subtypes of Second
Malignancies by Age
Entire cohort
Histologic type 0–4 y 5–9 y 10–14 y 15–19 y Total (%)
n ¼ 177
Carcinoma 2 1 4 16 23 (12.6)
Mean age at diagnosis initial tumor (years) 7.7 Sarcoma 1 5 9 12 27 (14.8)
Mean follow-up (years) 8.5 Bone tumors 1 13 9 11 34 (18.6)
Cumulative 10-y survival (%) 41.1 CNS tumors 3 1 6 7 17 (9.3)
n % of total Hematologic 5 14 15 31 65 (35.5)
Gender Skin 1 2 3 1 7 (3.8)
Male 96 54.2 Other 4 1 2 3 10 (5.5)
Female 81 45.7 Total 17 37 48 81 183
Age
<5 y old 72 40.7
5–9 y old 30 16.9 mortality cancers when seen as second malignancies. A
10–14 y old 45 25.4 similar effect was also noted for CNS tumors. The only
15–19 y old 30 16.9 exception was malignant bone tumor, which has no dif-
Race ference in mortality whether presenting as a primary
White 143 80.8
African American 16 9.0 or second malignancy.
Other 18 10.2 Comparison of survival rates performed between
Ethnicity patients who developed a subsequent second malig-
Hispanic 30 16.9
nancy within 5 y of primary tumor and those after 5 y
non-Hispanic 147 83.1
Tumor Stage demonstrated that a latent period of less than 5 y was
Local 48 27.1 associated with significantly lower 5- and 10-y survival
Regional 35 19.8 rates (36% versus 100%, and 26% versus 86%, respec-
Distant 43 24.3
Unknown 51 28.8
tively, P ¼ 0.0001).
Survival
Alive 76 42.9 DISCUSSION
Deceased 101 57.1

With just over 6,100,000 incident cancer cases, the

Kaplan-Meier method (observed) and compared with the SEER Program of the National Cancer Institute (NCI)
survival of the similar histologic subtypes from the is the largest registry source of information on cancer
SEER database for all childhood cancers (expected) incidence and survival in the United States. SEER cur-
(Table 6). For most tumor subtypes, there was a statisti- rently collects and publishes cancer incidence and sur-
cally significant decrease in 5- and 10-y survival rates vival data from 17 population-based cancer registries
once the patient developed a second malignancy. An es- encompassing approximately 26% of the U.S. popula-
pecially significant drop in survival was noted for hema- tion. The SEER program is the only comprehensive
tologic malignancies (19.7% versus 74% at 5 y and 16.6% source of population-based information in the United
versus 69.3% at 10 y, P < 0.001), indicating that leuke- States that includes stage of cancer at the time of diag-
mia and lymphoma, which are potentially curable child- nosis and actively collected patient survival data. The
hood cancers with high survival rates, behave as high data collected provide insight into tumor behavior and
allows investigators to examine outcomes from current
TABLE 2 treatment strategies. In addition, SEER has been previ-
Histological Subtypes of Primary Solid
ously used to report outcomes for a wide variety of pedi-
Tumors by Age atric malignancies [3–6].
Pediatric solid malignancies represent about 60% of all
Histologic type >1 y 1–4 y 5–9 y 10–14 y 15–19 y Total (%) childhood malignancies. In the United States, second
malignancies account for 6% to 10% of all cancer diagno-
Carcinoma 1 1 3 5 7 17 (9.6)
Sarcoma 2 6 8 7 5 28 (15.8) sis [7]. We studied a cohort of 177 patients with solid
Bone tumors 0 2 5 12 4 23 (13.0) malignancies who went on to develop a subsequent new
CNS tumors 1 14 6 13 6 40 (22.6) malignancy before 20 y of age. Because of the popula-
Retinoblastoma 16 9 0 0 0 25 (14.1)
Neuroblastoma 6 5 1 0 0 12 (6.8)
tion-based design and large size of SEER, the results pre-
Wilms tumor 0 5 6 2 0 13 (7.3) sented here are presumed to be broadly representative of
Gonadal tumors 1 0 1 3 4 9 (5.1) the survivors of childhood cancer in the U.S. as a whole.
Other 0 3 0 3 4 10 (5.6) We specifically included children less than 20 y of age to
Total 27 45 30 45 30 177
study the pattern of second tumors during childhood.
 VASUDEVAN ET AL.: PEDIATRIC SOLID TUMORS AND SECOND MALIGNANCIES 187

TABLE 4
Histology Subtypes of Primary and Second Malignancy

Primary tumor Histology of second malignancy

Histology Carcinoma Sarcoma Bone CNS Hematologic Skin Other

Carcinoma 0 3 6 3 3 0 2
Sarcoma 4 3 4 3 13 1 1
Bone tumor 1 1 2 1 16 1 1
CNS 9 5 5 3 16 3 2
Retinoblastoma 0 4 11 3 4 1 3
Neuroblastoma 5 3 1 1 2 0 0
Wilms tumor 2 4 3 1 4 0 0

Certain types of primary tumors are shown to have used is the main contributor [7]. Furthermore, genetic
a higher incidence of subsequent second tumors. In syndromes like NF-1 and Li-Fraumeni syndrome are
this study, CNS tumors were found to be the most com- known to play a prominent part in the development of
mon primary solid malignancy associated with the a second malignancy in patients with sarcoma [12].
development of a second malignancy (22.6%), followed Leukemia, both myeloid and lymphoid, formed the
by soft tissue sarcoma (16%), and retinoblastoma single largest histologic subtype of second malignancies
(14%). Although CNS tumors are the most common in this cohort, followed by bone and soft tissue sarcomas
solid malignancies in children, they have not been (18% and 15%, respectively). The 2007 report on the
noted in previous studies to be associated with a partic- SEER dataset reported an observed to expected (O/E)
ularly high risk for second tumors. A further detailed ratio of acute nonlymphocytic leukemia to be 17:1 in
analysis of pediatric brain tumors is warranted in order childhood cancer survivors [9]. In comparison, many
to study the risk of second malignancies with primary large studies report carcinomas of the breast and thy-
tumors at this site. A review of the literature reveals roid, bone sarcomas, and brain tumors to be the most
Hodgkin’s lymphoma, retinoblastoma, soft tissue common types of second malignancies [7–9, 13]. One pos-
sarcoma, and bone tumors are primary cancers associ- sible explanation for this difference could be that we only
ated with a significantly higher risk for second tumors included second malignancies in patients less than 20 y
[7–10]. Hereditary retinoblastoma has especially been of age in our analysis, suggesting leukemia to be the
associated with a higher cumulative incidence of second most common second malignancy encountered in the
tumors as opposed to patients with no hereditary pre- pediatric population, while on follow-up into adulthood,
disposition (36% at 50 y versus 5.7%) [11]. The high rel- other tumors like sarcoma and carcinoma become more
ative risk of a second cancer in patients with common. Also, many of these studies only concentrated
retinoblastoma reflects the combination of genetic pre- on radiation-associated solid malignancies, most com-
disposition to multiple cancers and sensitivity to radio- monly seen after treatment for Hodgkin’s lymphoma.
genic cancer in very young patients. In addition, soft Female gender is associated with increased risk of
tissue sarcomas are also associated with an increased second cancers, but this is primarily due to the excess
risk for second tumors. It is not completely clear if sar- numbers of breast cancers in adult female survivors,
coma is an independent risk factor for the development typically after chest irradiation for Hodgkin’s lym-
of a second cancer or whether the treatment modality phoma at a median age of about 35 y [7, 8, 11]. In the
pediatric population, both male and female patients
TABLE 5 had a similar incidence of second malignancies, further
Second Tumor Latency and Histological Subtypes supporting the above association.
The median interval from the diagnosis of primary to
Histologic type <1 y 1–4 y 5–9 y 10–17 y Mean LP
the development of a second malignancy was signifi-
Carcinoma 4 5 5 9 12.3 cantly different for solid versus non-solid tumors. About
Sarcoma 2 11 9 5 10.7 85% of the hematologic malignancies were seen within
Bone tumors 1 12 14 7 12.3 5 y of diagnosis of primary tumors, while bone and soft
CNS tumors 5 3 5 4 10.5
Hematologic 6 49 6 4 3.1
tissue sarcomas were seen on average 10–12 y after the
Skin 3 4 0 0 1.7 primary, and this finding is consistent with the results
Other 4 4 2 0 3.6 from most previous studies. Nonhematopoietic second
Total 25 88 41 29 malignancies seem to have a longer latency with the
LP ¼ latency period in years. risk continuing for two or more decades [7, 11].
188 JOURNAL OF SURGICAL RESEARCH: VOL. 160, NO. 2, MAY 15, 2010
TABLE 6
Observed versus Expected Second Malignancy Survival by Histology
5-y survival (%)
Observed Expected
Carcinoma 67.4 82.2
Sarcoma 55.6 69.5
Bone tumors 53.3 59.8
CNS tumors 40.7 65.2
Hematological tumors 19.7 74
Skin tumors 63.6 91.1
The above finding may be related to the treatment
modality used for the primary tumors. Significantly
elevated risks for acute myelocytic leukemia have
been noted in patients with childhood soft tissue sar-
coma treated with chemotherapy, especially alkylating
agents and topoisomerase II inhibitors (etoposide, or
teniposide) [7, 14]. Solid tumors like breast, thyroid,
and brain tumors, on the other hand, are common after
local radiotherapy, and typically have a longer latency
[8, 10, 11]. We found radiotherapy to have a signifi-
cantly higher association with tumors developing after
5 y of the diagnosis of primary malignancy. Exposure to
radiation therapy has been noted to be the most impor-
tant risk factor for the development of subsequent brain
tumors [15]. A higher risk for glioma in children irradi-
ated at a very young age may reflect greater susceptibil-
ity of the developing brain to radiation [15]. In addition,
the risk of osteosarcoma is reported to be a linear func-
tion of the cumulative absorbed dose of radiation with
an excess relative risk per gray of 1.8%, although
a lower risk is noted with orthovoltage radiotherapy
that delivers more superficial bone doses [16]. Simi-
larly, soft tissue sarcomas have been shown to occur
after very high doses of radiation (>10 Gy exposure),
suggesting a quadratic dose response relationship [17,
18]. These results clearly should guide us towards
a more judicial use of radiotherapy in childhood cancer
in order to minimize the devastating long-term conse-
quence of second malignancies.
To our knowledge, very few studies have looked into
the actual survival data after the development of sec-
ond malignancies. Overall, for all histologic tumor sub-
types, survival was significantly lower after the
development of a second malignancy compared with
the general pediatric cancer population. In addition,
significantly lower survival was noted when second
malignancies developed within 5 y of the primary
malignancy. This may be related to genetic susceptibil-
ity to cancer or to mutagenic effects of therapy or both.
The SEER cancer registry is an excellent resource to
study large numbers of patients in a population based
setting; however several limitations of this study
10-y survival (%)
P Observed Expected P
<0.001 49 80.1 <0.001
0.003 55.6 66.9 0.021
NS 47.7 54.1 NS
< 0.001 13.6 60.7 <0.001
<0.001 16.6 69.3 <0.001
<0.001 63.6 88.2 <0.001
should be noted. Since the database provides passive
follow-up for its registered cases, many new cancers
may have migrated outside of the SEER catchment
area, hence our numbers are likely to be conservative.
Furthermore, the data on family history and genetic
associations of cancer is unknown. This study is also
limited by the lack of complete, detailed treatment in-
formation on individual patients. Also, as with all
such studies, treatment strategies for childhood cancer
have changed considerably over the last 30 y, which
could affect the overall risk estimates.
In conclusion, second malignancies are a devastating
long-term complication of childhood solid cancer survi-
vors with significant morbidity and mortality. CNS
tumors, retinoblastoma, and soft tissue sarcomas in
children are the most common solid primary tumors
with increased risk of second malignancy. Leukemia
is the most common second malignancy seen in child-
hood. Young children are particularly susceptible to
the mutagenic effects of both chemotherapy and radio-
therapy, as proven by several studies. Although effec-
tive treatment for the initial cancer is of paramount
importance, future therapeutic strategies for childhood
cancer should continuously evolve to minimize this
potential problem in the long run.
REFERENCES
1. Mertens AC. Causes of mortality in survivors of childhood can-
cer. Pediatr Blood Cancer 2007;48:723.
2. Neglia JP, Freiedman DL, Yasui Y, et al. Second malignant neo-
plasms in five year survivors of childhood cancer: Childhood can-
cer survivor study. J Natl Cancer Inst 2001;93:618.
3. Gutierrez JC, Fischer AC, Sola JE, et al. Markedly improving
survival of neuroblastoma: A 30-year analysis of 1646 patients.
Pediatr Surg Int 2007;23:637. Epub May 3, 2007.
4. Gutierrez JC, Housri N, Koniaris LG, et al. Malignant breast
cancer in children: A review of 75 patients. J Surg Res 2008;
147:182. Epub Apr 14, 2008.
5. Perez EA, Gutierrez JC, Koniaris LG, et al. Malignant pancre-
atic tumors in children: Incidence and outcomes in 58 patients.
J Pediatr Surg 2009;44:197.
6. Yang R, Cheung MC, Zhuge Y, et al. Primary solid tumors of the
colon and rectum in the pediatric patient: A review of 270 cases.
J Surg Res Epub Dec 30, 2008.
 VASUDEVAN ET AL.: PEDIATRIC SOLID TUMORS AND SECOND MALIGNANCIES
7. Bhatia S, Sklar C. Second cancers in survivors of childhood can-
cer. Nat Rev Cancer 2002;2:124.
8. Robison LL. Treatment-associated subsequent neoplasms among
long-term survivors of childhood cancer: The experience of the
childhood cancer survivor study. Pediatr Radiol 2009;39:32.
9. Inskip PD, Curtis RE. New malignancies following childhood can-
cer in the United States, 1973-2002. Int J Cancer 2007;121:2233.
10. Green DM, Zevon MA, Reese PA, et al. Second malignant tumors
following treatment during childhood and adolescence for can-
cer. Med Pediatr Oncol 1994;22:1.
11. Guerin S, Hawkins M, Shamsaldin A, et al. Treatment–adjusted
predisposition to second malignant neoplasms after a solid can-
cer in childhood: A case-control study. J Clin Oncol 2007;25:2833.
12. Cohen RJ, Curtis RE, Inskip PD, et al. The risk of developing sec-
ond cancers among survivors of childhood soft tissue sarcoma.
Cancer 2005;103:2391.
13. MacArthur AC, Spinelli JJ, Rogers PC, et al. Risk of a second
malignant neoplasm among 5-year survivors of cancer in child-
189
hood and adolescence in British Columbia, Canada. Pediatr
Blood Cancer 2007;48:453.
14. Hijaya N, Ness KK, Ribiero RC, et al. Acute leukemia as a sec-
ondary malignancy in children and adolescents. Cancer 2009;
115:23.
15. Neglia JP, Robison LL, Stovall M, et al. New primary neoplasms
of the central nervous system in survivors of childhood cancer:
A report from the childhood cancer survivor study. J Natl Cancer
Inst 2006;98:1528.
16. Vu BL, Vathaire FD, Shamsaldin A, et al. Radiation dose, che-
motherapy and risk of osteosarcoma after solid tumors during
childhood. Int J Cancer 1998;77:370.
17. Menu-Branthomme A, Rubino C, Shamsaldin A, et al. Radiation
dose, chemotherapy and risk of soft tissue sarcoma after solid
tumors during childhood. Int J Cancer 2004;110:87.
18. Bisogno G, Sotti G, Nowicki Y, et al. Soft tissue sarcoma as a sec-
ond malignant neoplasm in the pediatric age group. Cancer
2004;100:1758.