Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Oct 2.
Published in final edited form as: Pediatr Blood Cancer. 2026 May 22;73(8):e70415. doi: 10.1002/1545-5017.70415

Increased risk of sarcomas in children with congenital anomalies: Findings from the Genetic Overlap Between Anomalies and Cancer in Kids (GOBACK) Registry Linkage Study

Russ Wolters 1,2,*, Ji Yun Tark 3,*, Tiffany M Chambers 1, Tania A Desrosiers 4, Michael E Scheurer 5,6, Charles Shumate 7, Wendy N Nembhard 8, Mahsa M Yazdy 9, Eirini Nestoridi 9, Amanda E Janitz 10, Jean Paul Tanner 11, Russell S Kirby 11, Jason L Salemi 11, Yao Yu 12, Chad D Huff 12, Sharon E Plon 1, Philip J Lupo 5,6, Jeremy M Schraw 1
PMCID: PMC13629186  NIHMSID: NIHMS2211470  PMID: 42170858

Abstract

Background:

Pediatric sarcomas are a heterogeneous group of tumors that contribute disproportionately to cancer mortality in children. Although congenital anomalies are among the strongest known risk factors for childhood cancer, the risk of specific sarcoma subtypes among affected individuals has not yet been thoroughly evaluated.

Procedure:

We obtained data on maternal and perinatal characteristics, congenital anomalies, and pediatric sarcoma diagnoses for all live births in nine states. We used Cox proportional hazards regression to estimate the hazard ratio (HR) and 95% confidence interval (CI) of sarcoma (overall and by subtype) among children with non-syndromic congenital anomalies. We considered all non-syndromic anomalies collectively, and when sample size allowed, we also evaluated specific anomaly-sarcoma associations.

Results:

We evaluated 21,933,884 children, including 641,770 (2.9%) with major non-syndromic congenital anomalies. Compared to children without a congenital anomaly, children with a non-syndromic anomaly had a two-fold higher hazard for any soft tissue sarcoma (95% CI: 1.7–2.5), including rhabdomyosarcoma (HR: 2.2, 95% CI: 1.7–3.0) and embryonal rhabdomyosarcoma (HR: 2.5, 95% CI: 1.8–3.6), as well as non-rhabdomyosarcoma soft tissue sarcoma (HR: 1.8, 95% CI, 1.3–2.5). The hazard of embryonal rhabdomyosarcoma was markedly increased in children with central nervous system anomalies (HR: 7.9, 95% CI: 3.9–15.9), obstructive genitourinary defects (HR: 4.6, 95% CI: 2.2–9.7), and limb reduction deformities (HR: 3.8, 95% CI: 1.6–9.3).

Conclusions:

Children with non-syndromic congenital anomalies are at increased risk of sarcomas, especially soft tissue sarcomas. Future studies should clarify shared developmental pathways and evaluate implications for sarcoma risk prediction and surveillance.

Keywords: congenital anomalies, birth defects, sarcoma, rhabdomyosarcoma, pediatrics, epidemiology

Introduction

Sarcomas represent a diverse group of cancers believed to originate from mesenchymal cells. Although these malignancies account for only 10% of all pediatric cancer diagnoses and 15–20% of pediatric solid tumors, they remain among the five leading causes of cancer-related death in children and adolescents.1–3 The World Health Organization (WHO) recognizes over 50 different sarcoma subtypes,4 broadly categorized into bone sarcomas and soft tissue sarcomas. Among bone sarcomas, osteosarcoma and Ewing sarcoma are most common, with Ewing sarcoma occurring in both bone and soft tissues. Soft tissue sarcomas are further classified into rhabdomyosarcoma (RMS), which represents nearly half of all cases, and non-RMS soft tissue sarcomas (NRSTS). Common NRSTS types in children include fibrohistiocytic tumors, desmoid tumors, fibroblastic tumors, undifferentiated soft tissue sarcomas, synovial sarcoma, and malignant peripheral nerve sheath tumors (MPNST). According to the 2021 Surveillance, Epidemiology, and End Results (SEER) data, the incidence of pediatric soft tissue sarcomas among females and males under 20 years is 1.1 and 1.0 per 100,000, respectively, while bone sarcoma rates were 0.8 and 1.2 per 100,000.5

Congenital anomalies, affecting approximately 1 in 33 U.S. children, represent one of the strongest known risk factors for pediatric cancer.6 Several rare cancer predisposition syndromes are associated with increased sarcoma risk, including neurofibromatosis type 1 (MPNST, RMS), Li-Fraumeni syndrome (osteosarcoma, soft tissue sarcomas), hereditary retinoblastoma (osteosarcoma, soft tissue sarcomas), Rothmund-Thompson syndrome (osteosarcoma),7 familial adenomatous polyposis (desmoid tumor), Costello syndrome (RMS), DICER1 syndrome (RMS, soft tissue sarcomas ), Werner syndrome (soft tissue sarcomas), rhabdoid tumor predisposition syndrome (malignant rhabdoid tumors), and Noonan syndrome (soft tissue sarcomas). However, most congenital anomalies diagnosed in infants and children are non-syndromic (i.e., they lack a defined molecular etiology). The risk of sarcoma in children with non-syndromic anomalies is not well studied. Prior studies have examined associations between specific (or overall) non-syndromic anomalies and sarcomas; however, findings have been inconsistent, with some studies reporting significant associations8–15 and others finding none16–21. Much of the literature has been limited by small sample sizes, necessitating broad grouping of heterogenous tumors and anomalies, which can yield imprecise and sometimes conflicting risk estimates. In our previous Genetic Overlap Between Anomalies and Cancer in Kids (GOBACK 1.0) Registry Linkage study of 10 million children, children with non-syndromic anomalies had higher risks of RMS (Hazard Ratio [HR] 1.8, 95% confidence interval [CI] 1.3–2.6) and NRSTS (HR 3.4, 95% CI 2.8–4.2), with a suggestive but imprecise association for bone tumors (HR 1.2, 95% CI 0.8–1.9), compared to children without anomalies.22

Understanding the relationships between specific anomalies and sarcomas, along with their molecular mechanisms, could reveal novel cancer predisposition syndromes and inform cancer surveillance strategies. To facilitate this, we analyzed data from the GOBACK 2.0 Registry Linkage study, which now includes over 22 million live births. This expanded dataset enabled us to generate more precise risk estimates, characterize specific anomaly-sarcoma associations, and evaluate sarcoma risk in children with multiple congenital anomalies.

Methods

This study was approved by the Institutional Review Board (IRB) at each participating institution. These IRBs approved the study with a waiver of informed consent because it was a secondary analysis of de-identified public health data.

Data Linkages

The GOBACK 1.0 registry linkage study included 10 million live births from 1992–2013 in four U.S. states: Texas, Arkansas, Michigan, and North Carolina.22 The GOBACK 2.0 Registry Linkage Study expanded this effort to include all live births from : Texas (1999–2017), Michigan (1992–2018), New Jersey (1990–2018) Florida (1998–2012), North Carolina (2003–2018), Massachusetts (2000–2017), Oklahoma (1997–2012), Arkansas (1995–2011), and South Carolina (2008–2018). In each state, data from population-based birth defects and cancer registries were linked to each other and to vital records to identify all live births, cases with birth defects, and cases with cancer. Participating birth defects registries provided data on all definite birth defect diagnoses made within the first one to two years of life (International Classification of Diseases, Ninth and Tenth Editions [ICD-9 and ICD-10] or modified British Paediatric Association [BPA] diagnostic codes). Registries in Texas, Arkansas, North Carolina, Massachusetts, Oklahoma, and South Carolina used active ascertainment methods to identify affected children; registries in Michigan, New Jersey, and Florida used passive methods. Participating cancer registries provided data (International Classification of Diseases for Oncology, 3rd Edition [ICD-O-3] site, histology, and behavior codes) for all malignant neoplasms (and benign brain tumors) diagnosed through 18 years of age or the end of the study period (which varied by state and were the same as the birth years listed above). Child’s sex, birthweight, gestational age, and plurality (singleton or multiple) were ascertained from birth records, as were maternal age and race/ethnicity. The linked data were then de-identified and transferred to the investigators for cleaning, harmonization, and analysis.

Classification of Congenital Anomalies

This study focused on approximately 50 major congenital anomalies commonly ascertained by population-based birth defects registries.6 We also evaluated associations between the number of major congenital anomalies (none, one, two or more) and pediatric sarcoma, because we previously demonstrated that being born with multiple congenital anomalies is associated with an even higher risk of developing childhood cancer.22 We evaluated major anomalies, those likely to affect a child’s physical/social functioning, overall health, and life expectancy, consistent with surveillance recommendations from the National Birth Defects Prevention Network (NBDPN).23. Children with chromosomal or genetic syndromes (N=49,620) were evaluated separately, as our primary objective was to characterize the associations between non-syndromic anomalies and pediatric sarcoma. These included children with Trisomy 13, 18, and 21, 13q or 22q deletion syndromes, other autosomal deletions, neurofibromatosis, and tuberous sclerosis complex.

Classification of Sarcomas

Data on sarcoma type and age at diagnosis were obtained from population-based cancer registries in each state. Sarcomas were classified according to the International Classification of Childhood Cancer, Third Edition. Osteosarcoma, Ewing sarcoma, and “other specified malignant bone tumors” were coded as malignant bone tumors within the GOBACK dataset. Soft tissue sarcomas included RMS and NRSTS. RMS type was coded as embryonal, alveolar, or other (which could include pleomorphic or unknown).

Statistical Analysis

Counts and percentages were used to summarize the number of children in the cohort, number with a chromosomal anomaly or single-gene syndrome, number with a major congenital anomaly, number with sarcoma and each sarcoma subtype, as well as to compare the distributions of categorical maternal and perinatal characteristics between subjects with and without congenital anomalies. Means and standard deviations were computed to summarize the distributions of continuous covariates among individuals with and without congenital anomalies. The prevalence of each major organ system anomaly was compared between children with and without sarcomas using chi-square or Fisher’s exact tests.

To evaluate the risk of sarcoma among children with congenital anomalies, two separate analyses were performed: one analysis for children with known chromosomal or single-gene syndromes; and another for children with non-syndromic anomalies. The reference group for both analyses was children without a congenital anomaly. We used the Kaplan-Meier method and log-rank test to estimate the cumulative incidence of sarcomas and evaluate differences between groups. We used Cox proportional hazards models to estimate hazard ratios (HR) and corresponding 95% confidence intervals for congenital anomaly-sarcoma associations. Cox models were adjusted for child’s sex, state of birth, maternal age, and birth year. We did not adjust for potential mediators (e.g., birth weight and gestational age) because doing so could introduce overadjustment bias and estimate a controlled direct effect rather than the total effect of congenital anomaly-sarcoma associations.24 We evaluated cancer risk through 18 years of age and right-censored observations at this point. Children were followed up until the date of their first sarcoma diagnosis; children who were not linked to a record from the cancer registry in their state of birth were assumed to have survived without an event through 18 years of age or until the end of the study period, whichever occurred sooner. The analysis of hypospadias was limited to male children. All statistical tests were two-sided. We performed a 5% the Benjamini-Hochberg false discovery rate25 correction to account for multiple testing. Statistical analyses were performed in R studio Version 4.3.1 (R Core Team, Vienna, Austria). The ‘survival’ and ‘survminer’ packages were used to perform survival analyses.26 To protect privacy and confidentiality, only associations for which there were five or more co-occurring cases were evaluated.

Results

The study included 21,933,884 children, of whom 641,770 (2.9%) had one or more major non-syndromic congenital anomalies and 49,620 (0.23%) had a chromosomal anomaly or single-gene syndrome (Table 1). As compared to children without congenital anomalies, children with major non-syndromic anomalies were more likely to be male (58.3% vs 50.8%) or part of a multiple gestation pregnancy (6.7% vs 3.2%), had lower mean birthweight (2960g vs 3281g), and were delivered at a younger mean gestational age (37.0 vs 38.6 weeks). Children with chromosomal or genetic syndromes also had earlier mean gestational age (37.2 vs. 38.6 weeks) and lower mean birthweight (2,793 vs. 3,281 g) and were born to older mothers than children without congenital anomalies. The distribution of maternal race/ethnicity was similar between groups. Children with sarcoma had a higher prevalence of major circulatory system anomalies than children without sarcoma (2.8% vs. 1.6%; X2 p= 1.4 × 10−6). Higher prevalences were also observed for major integument (1.4% vs. 0.7%; X2 p=3.7× 10−5), musculoskeletal (1.1% vs. 0.6%; X2 p=4.2× 10−4), genitourinary (1.2% vs. 0.6%; X2 p=7.0× 10−4), and respiratory anomalies (0.10% vs. 0.04%; Fisher’s exact p=0.03), whereas digestive, central nervous system, and ear/face/neck anomalies did not differ significantly (all p>0.05).

Table 1.

Child and maternal characteristics among subjects without congenital anomalies, subjects with major non-syndromic anomalies, and subjects with chromosomal or genetic syndromes, Genetic Overlap Between Anomalies and Cancer in Kids (GOBACK) Registry Linkage Cohort.

Without anomalies Major non-syndromic congenital anomalies Chromosomal or genetic syndromes
Total in cohort 21,242,494 641,770 49,620
Age at sarcoma diagnosis (yrs, median [Q1, Q3]) 7.5 (3.0, 12.1) 4.2 (1.2, 8.4) 4.5 (2.1, 10.7)
 Soft tissue sarcomas 4.9 (2.0, 10.3) 2.5 (0.9. 5.5) 4.3 (2.1, 10.6)
 Bone tumors 10.5 (7.0, 13.4) 7.9 (5.2, 12.2) 9.7 (5.9, 11.8)
Child’s sex, N (%)
 Male 10,782,668 (50.8) 374,459 (58.3) 25,138 (50.7)
 Female 10,459,826 (49.2) 267,311 (41.7) 24,482 (49.3)
Birthweight, mean (SD), grams 3,281.4 (576.5) 2,960.3 (898.0) 2,793.5 (763.7)
Gestational age, mean (SD), weeks 38.6 (2.0) 37.0 (3.8) 37.2 (2.9)
Maternal race/ethnicity, N (%)
 American Indian/Alaskan Native 157,998 (0.7) 4,812 (0.7) 374 (0.8)
 Hispanic 5,033,467 (23.7) 149,708 (23.3) 13,325 (26.9)
 Non-Hispanic Asian 861,278 (4.1) 19,938 (3.1) 1,562 (3.1)
 Non-Hispanic Black 3,337,143 (15.7) 107,376 (16.7) 7,214 (14.5)
 Non-Hispanic White 11,403,848 (53.7) 348,390 (54.3) 26,277 (53.0)
 Additional non-Hispanic groups 269,228 (1.3) 8,317 (1.3) 558 (1.1)
 Missing 179,532 (0.8) 3,229 (0.5) 310 (0.6)
Maternal age (yrs)
 <20 2,128,856 (10.0) 64,334 (10.0) 3,399 (6.9)
 20–24 5,072,520 (23.9) 150,492 (23.4) 8,192 (16.5)
 25–29 5,786,380 (27.2) 170,558 (26.6) 9,499 (19.1)
 30–34 5,101,839 (24.0) 152,508 (23.8) 10,616 (21.4)
 35–39 2,459,931 (11.6) 80,102 (12.5) 10,724 (21.6)
 ≥40 564,486 (2.7) 20,502 (3.2) 6,982 (14.1)
 Missing 128,482 (0.6) 3,274 (0.5) 208 (0.4)
Plurality, N (%)
 Singleton 20,532,787 (96.7) 597,750 (93.1) 47,722 (96.2)
 Multiple 686,607 (3.2) 42,705 (6.7) 1,789 (3.6)
 Missing 23,100 (0.1) 1,315 (0.2) 109 (0.2)
Open in a new tab

During the study period, 2,921 children were diagnosed with any sarcoma: 2,665 in the group without anomalies, 135 among children with major, non-syndromic anomalies, and 21 among those with syndromes. The corresponding incidence rates (95% CIs) were 12.7 (12.2–13.2), 22.4 (18.8–26.5), and 46.8 (29.0–71.6) per million person-years, respectively. The median age at sarcoma diagnosis was 7.5 (interquartile range 3.0–12.1) among children without anomalies, 4.2 (1.2–8.4) among those with major, non-syndromic anomalies, and 4.5 (2.1–10.7) among those with syndromes (Table 1). Children with bone tumors were diagnosed at older ages than those with soft tissue tumors. Table 2 presents information on sarcoma diagnoses by group: the proportion of soft tissue sarcomas, particularly embryonal RMS, was higher among children with major, non-syndromic anomalies (26.7%) than among those without anomalies (18.0%), whereas the prevalence of bone tumors was lower (28.1% vs. 39.5%) (Supplementary Figure 1). For details regarding the group with NRSTS, refer to Supplementary Table 1.

Table 2.

Distribution of Sarcoma diagnoses among subjects without congenital anomalies, subjects with major non-syndromic anomalies, and subjects with chromosomal or genetic syndromes, Genetic Overlap Between Anomalies and Cancer in Kids (GOBACK) Registry Linkage Cohort.

Diagnoses, N (%) Without anomalies Major non-syndromic congenital anomalies Chromosomal or genetic syndromes
Sarcoma 2,665 135 21
 Bone tumorsa 1,062 (39.5) 38 (28.1) 3 (14.3)
  Osteosarcoma 500 (18.6) 18 (13.3) 1 (4.8)
  Ewing sarcomab 452 (16.8) 14 (10.4) 2 (9.5)
  Other malignant bone tumors 112 (4.2) 6 (4.4) 0 (0.0)
 Soft tissue sarcomasc 1,637 (60.9) 97 (71.9) 18 (85.7)
  Rhabdomyosarcoma 836 (31.1) 54 (40.0) 6 (28.6)
   Alveolar rhabdomyosarcoma 206 (7.7) 8 (5.9) 0 (0.0)
   Embryonal rhabdomyosarcoma 484 (18.0) 36 (26.7) 5 (23.8)
   Other rhabdomyosarcomad 148 (5.5) 10 (7.4) 1 (4.8)
  NRSTS 804 (29.9) 43 (31.9) 12 (57.1)
Open in a new tab

Abbreviations: NRSTS=non-rhabdomyosarcoma soft tissue sarcoma.

a

Bone tumor includes osteosarcoma, Ewing Sarcoma, and other malignant bone tumors.

b

Ewing sarcoma includes extra-skeletal cases. PNET of soft tissue and bone is categorized as Ewing sarcoma.

c

Soft tissue sarcoma includes rhabdomyosarcoma and non-rhabdomyosarcoma soft tissue sarcomas.

d

Other rhabdomyosarcoma includes not rhabdomyosarcoma not otherwise specified (NOS) and pleomorphic rhabdomyosarcoma.

Risk of sarcoma among children with major, non-syndromic congenital anomalies

We observed stronger and more consistent associations with soft tissue sarcomas than with bone tumors (Tables 3–4). There were numerous associations between major, non-syndromic congenital anomalies and soft tissue sarcomas (Table 3). Specifically, there were increases in the hazard of any soft tissue sarcoma (HR 2.04, 95% CI 1.66–2.52), any RMS (HR 2.24, 95% CI 1.70–2.95), embryonal RMS (HR 2.54, 95% CI 1.81–3.57), and NRSTS (HR 1.83, 95% CI 1.34–2.51). Specific and notable congenital anomaly-sarcoma associations were identified: the hazard of embryonal RMS was increased in children with central nervous system anomalies (HR 7.90, 95% CI: 3.93–15.91), obstructive genitourinary defects (HR: 4.59, 95% CI: 2.17–9.69), or limb reduction deformities (HR: 3.84, 95% CI: 1.59–9.29). Children with hypospadias were 2.97 (95% CI, 1.32–6.65) times more likely to be diagnosed with NRSTS. Conversely, no statistically significant associations were found between congenital anomalies and alveolar RMS. The results were largely consistent after controlling for multiple testing (Supplementary Table 2 and 3).

Table 3.

Hazard Ratios and 95% Confidence Invervals for Diagnosis of Soft Tissue Sarcomas Among Children with Non-syndromic Congenital Anomalies, Genetic Overlap Between Anomalies and Cancer in Kids (GOBACK) Registry Linkage Cohort.

Soft tissue sarcoma Rhabdomyosarcoma Embryonal Rhabdomyosarcoma NRSTSa
Any major congenital anomaly 2.04 (1.66–2.52) 2.24 (1.70–2.95) 2.54 (1.81–3.57) 1.83 (1.34–2.51)
Congenital anomalies of the central nervous system 2.91 (1.55–5.44) 4.91 (2.53–9.56) 7.90 (3.93–15.91) NT
Congenital anomalies of the cardiovascular system 2.08 (1.58–2.74) 2.05 (1.39–3.02) 2.12 (1.29–3.49) 2.11 (1.42–3.13)
 Congenital anomaly of the pulmonary artery 2.39 (1.14–5.03) NTb NT NT
 Ventricular septal defect 2.31 (1.20–4.47) 2.19 (1.13–4.27) NT 2.36 (1.21–4.59)
 Atrial septal defect or Patent foramen ovale 2.51 (1.81–3.46) 2.49 (1.59–3.89) 2.78 (1.60–4.83) 2.53 (1.58–4.03)
Congenital anomalies of the digestive system 1.53 (0.72–3.25) 2.12 (0.87–5.17) NT NT
Congenital anomalies of the genitourinary system 2.53 (1.71–3.75) 2.33 (1.34–4.06) 3.19 (1.70–5.98) 2.76 (1.59–4.78)
 Obstructive genitourinary defects 3.25 (1.98–5.33) 3.73 (1.99–7.00) 4.59 (2.17–9.69) 2.67 (1.19–5.96)
 Hypospadiasc 2.09 (1.08–4.02) NT NT 2.97 (1.32–6.65)
Congenital anomalies of the musculoskeletal system 2.19 (1.42–3.65) 2.36 (1.33–4.20) 2.80 (1.39–5.63) 1.98 (1.03–3.83)
 Limb reduction deformities 3.02 (1.75–5.22) 3.11 (1.47–6.55) 3.84 (1.59–9.29) 2.92 (1.31–6.53)
Congenital anomalies of the integumentary system 2.45 (1.64–3.65) 1.92 (1.03–3.61) 2.15 (1.01–4.57) 2.99 (1.78–5.02)
Open in a new tab

Only co-occurring anomaly-soft tissue sarcoma pairs with count ≥5 are shown. Models were adjusted for covariates of sex, maternal age, state, and birth year.

a

NRSTS = Non-rhabdomyosarcoma soft tissue sarcoma

b

NT = not tested for phenotypes with <5 exposed cases.

c

Analyses were conducted among male children.

Table 4.

Hazard Ratios and 95% Confidence IntervaIs for Diagnosis of Bone Tumors Among Children with Non-syndromic Congenital Anomalies, Genetic Overlap Between Anomalies and Cancer in Kids (GOBACK) Registry Linkage Cohort.

Any bone tumor Osteosarcoma Ewing Sarcomaa
Any major congenital anomaly 1.31 (0.95–1.82) 1.35 (0.84–2.16) 1.09 (0.63–1.89)
Congenital anomalies of the cardiovascular system 1.59 (1.06–2.40) 2.04 (1.20–3.48) 1.10 (0.52–2.33)
 Atrial septal defect or patent foramen ovale 1.44 (0.82–2.56) 1.63 (0.73–3.66) 1.40 (0.58–3.39)
Congenital anomalies of the genitourinary system 0.97 (0.43–2.18) NTb NT
Congenital anomalies of the musculoskeletal system 1.49 (0.77–2.89) 1.79 (0.74–4.33) NT
Congenital anomalies of the integumentary system 2.20 (1.21–4.00) 2.64 (1.17–5.94) NT
Open in a new tab

Only co-occurring anomaly-bone tumor pairs with count ≥5 are shown. Models were adjusted for covariates of sex, maternal age, state, birth year.

a

Ewing sarcoma includes extra-skeletal cases.

b

NT = not tested for phenotypes with <5 exposed cases.

Table 4 presents the HRs and 95% CIs for the diagnosis of bone sarcomas in children with major non-syndromic anomalies compared to children without anomalies. Children with any major congenital anomaly did not have clear evidence of increased hazards of any bone tumors (HR 1.31, 95% CI 0.95–1.82), osteosarcoma (HR 1.35, 95% CI 0.84–2.16), and Ewing sarcoma (HR 1.09, 95% CI 0.63–1.89). However, certain anomaly subgroups were significantly associated with increased risk: cardiovascular anomalies were associated with any bone tumor (HR 1.59, 95% CI: 1.06–2.40) and osteosarcoma (HR 2.04, 95% CI: 1.20–3.48), and integumentary anomalies were associated with any bone tumor (HR 2.20, 95% CI: 1.21–4.00) and osteosarcoma (HR 2.64, 95% CI: 1.17–5.94). Because bone tumors are more likely to peak during adolescence,27,28 we conducted sensitivity analyses restricting the cohort to children with ≥10 years of follow-up (n=10,766,074) and to children born between 1990 and 1999 (n=3,254,655), among whom 79.4% had complete follow-up through age 18 years. Results were consistent with the primary analysis (Supplementary Table 4 and 5).

Risk of sarcoma among children with multiple congenital anomalies

Increased cumulative incidence of any sarcoma, any soft tissue sarcoma, embryonal RMS, and NRSTS was observed in children with major anomalies compared to children without a congenital anomaly, most notably in the group with multiple major anomalies (Figure 1A–D). For NRSTS, the cumulative incidence curves for children with one versus multiple major anomalies were similar until approximately 8 years of age, after which incidence appeared to increase more markedly among children with multiple major anomalies. The cumulative incidence of bone tumors did not differ between children with and without congenital anomalies (Supplementary Figure 2).

Figure 1.

Figure 1.

Open in a new tab

Cumulative incidence of any sarcoma and soft tissue sarcoma according to number of major congenital anomalies, Genetic Overlap Between Anomalies and Cancer in Kids (GOBACK) Registry Linkage Cohort.

Risk of sarcoma among children with chromosomal anomalies or genetic syndromes

Twenty-one children with chromosomal anomalies or genetic syndromes developed sarcoma. Of these cases, eleven were nerve sheath tumors, all occurring in children with neurofibromatosis. Thus, the risk of NRSTS was dramatically higher in children with neurofibromatosis, who were 189.07 times more likely to develop these tumors compared to children without congenital anomalies (95% CI, 103.23–346.27). No other genetic or chromosomal syndromes were associated with sarcoma.

Discussion

This large, population-based registry linkage study demonstrates an increased risk of sarcomas in children with major congenital anomalies and provides critical details regarding specific anomaly-sarcoma associations. Approximately 5% of sarcoma cases had a recorded major, non-syndromic congenital anomaly, compared to 2.9% of our cohort. Leveraging our expanded cohort, we were able to evaluate the risk of specific sarcoma subtypes (embryonal RMS, alveolar RMS, NRSTS), which has been difficult in previous studies with smaller samples but is critically needed because these diseases are biologically distinct and have different clinical behavior. There were clear differences when considering sarcomas of soft tissue versus bone. Whereas many anomalies were associated with soft tissue sarcomas, especially embryonal RMS, few associations were evident for bone tumors and hazard ratios tended to be closer to unity. Together, these findings highlight the importance of studying these diseases individually and not as one broad group when seeking to understand their etiologies and risk factors. Our results were largely consistent with our prior GOBACK 1.0 study22 and extend the literature by showing increasing risk of soft tissue sarcomas with a greater number of anomalies.

With respect to RMS, the largest increase in the hazard was observed among children with central nervous system anomalies. Ruymann et al. first suggested a potential association in an autopsy study of 115 cases; congenital anomalies were present in 32% of patients, including nine with central nervous system anomalies, but results were not evaluated by histologic subtype.29 To our knowledge, no subsequent studies reported associations involving central nervous system anomalies. Given sparse co-occurrence, we were unable to estimate risks for specific central nervous system anomaly subtypes. Our finding of the association between embryonal RMS and obstructive genitourinary defects is consistent with the 1998 registry case-control study by Altmann et al.12 However, that estimate was based on only one child with both a genitourinary anomaly and embryonal RMS, yielding an OR of 18.2 with very wide CIs (2.1–157). The associations of central nervous system and genitourinary anomalies with embryonal RMS (n=8) may point to important shared disrupted developmental pathways. For example, Hedgehog signaling is essential for embryonic development, particularly neural tube formation and patterning.30 Both reduced and excessive Sonic Hedgehog (Shh) signaling have been associated with central nervous system anomalies, including holoprosencephaly, exencephaly, and spina bifida.30,31 Shh signaling also plays an important role in urogenital organ development.32,33 Given evidence supporting an oncogenic role of Hedgehog signaling in RMS, including increased pathway activation and associations with poorer prognosis,34–36 this pathway may warrant further investigation as a potential shared mechanism linking central nervous system and genitourinary anomalies and embryonal RMS.

Also noteworthy were the associations of atrial septal defects and limb reduction deformities, each with embryonal RMS. There are no previous reports of cardiac septal defects being associated with an increased risk of RMS. This is notable because established RMS predisposition syndromes, including Li-Fraumeni syndrome, neurofibromatosis, Costello syndrome, and Beckwith-Wiedemann syndrome, do not classically include the anomalies that we observed to be associated with embryonal RMS (cardiac defects are common in children with Costello syndrome, though these are typically classified as cardiomyopathy, while septal defects have been less frequently reported). Likewise, our finding that limb reduction deformities were associated with embryonal RMS is novel, and these are not features of RMS predisposition syndromes. Children with Beckwith-Wiedemann syndrome can have hemihypertrophy, but the limb reduction deformities observed in this cohort are not classical features of this syndrome. Our findings suggest the possibility of undiagnosed or previously unrecognized phenotypes characterized by the co-occurrence of congenital anomalies and embryonal RMS.

Consistent with prior literature,37,38 children with neurofibromatosis had a markedly higher risk of NRSTS than children without anomalies, driven largely by nerve sheath tumors. However, evidence on associations between non-syndromic anomalies and NRSTS remains limited. In our study, children with major non-syndromic congenital anomalies were nearly twice as likely to be diagnosed with a NRSTS, and about three-fold increases in the hazard of NRSTS were seen for hypospadias and anomalies of the integument. The most common diagnoses within the NRSTS group were unspecified soft tissue sarcomas, fibrohistiocytic tumors, and fibroblastic and myofibroblastic tumors. The category of NRSTS remains very broad and includes numerous diagnoses with very different clinical behavior. Because of this heterogeneity, it is challenging to make inferences about these diseases and associated risk factors. Further studies focused on specific NRSTS types would be of benefit.

Children with cardiovascular anomalies were twice as likely to be diagnosed with osteosarcoma. Co-occurring cases were diagnosed with patent ductus arteriosus, atrial septal defect or patent foramen ovale, transposition of the great vessels/double-outlet right ventricle, ventricular septal defect, coarctation of the aorta, anomaly of the pulmonary artery, and pulmonary valve atresia, indicating that increased risk was not specific to a certain cardiac anomaly. There are three rare cancer predisposition syndromes associated with an increased risk of osteosarcoma (Rothmund-Thompson syndrome, hereditary retinoblastoma, Li-Fraumeni syndrome), none of which are known to be associated with cardiac anomalies. This finding suggests the need for future exploration of the relationship between cardiac anomalies and osteosarcoma. We also observed increased osteosarcoma risk in children with congenital anomalies of the integumentary system, an association that was not reported in our own prior study or in an earlier systematic review of congenital anomaly-cancer associations.22,39,40

Strengths and limitations

The present study was performed in a large sample, from states that represent approximately 25% of U.S. births, which enabled the most comprehensive assessment of congenital anomaly-sarcoma associations to date. Confirmed diagnoses of both congenital anomalies and sarcomas were ascertained from population-based registries with essentially complete case ascertainment, and most participating cancer registries achieved the highest level of certification from the North American Association of Central Cancer Registries (NAACCR) during the study period. Applying this systematic case definition across states reduces heterogeneity in outcome classification and eliminates recall bias associated with parental or self-report. In addition, sensitivity analyses addressing the potential for underascertainment of bone tumors among more recent birth cohorts showed that the results were robust. Despite these strengths, this study has important limitations. Congenital anomaly ascertainment may differ by state surveillance system (e.g., active vs. passive), which could lead to misclassification; however, our prior study showed consistent results between states.22 To help address potential differences in case ascertainment methodology, all multivariable models were adjusted for state and year of birth. In addition, children with congenital anomalies may have more frequent clinical encounters and diagnostic imaging, which could lead to earlier detection of sarcomas. We were unable to account for residential mobility, meaning that outcomes among children who left the state of their birth before diagnosis with a congenital anomaly and/or cancer are unobserved in our data. Studies of early-life residential mobility in children who developed cancer suggest that while changing residential addresses is common, most moves occur within the state of birth and over a relatively short distance, which would not result in loss to follow-up in our study.41,42 Furthermore, our previous study of residential mobility and environmental exposure assignment among children with central nervous system tumors in Texas suggested that any resulting outcome misclassification would likely be non-differential and bias the results toward the null.43 Finally, the study was likely underpowered to evaluate associations between congenital anomalies and specific NRSTS types and associations between specific central nervous system anomalies and soft tissue sarcomas.

Conclusions

This analysis of a cohort of over 22 million births demonstrated that children with non-syndromic congenital anomalies are at increased risk of sarcomas, especially embryonal RMS and NRSTS. Few associations were evident for alveolar RMS or bone tumors. Our study also identified greater incidence of sarcoma and embryonal RMS among children with multiple major congenital anomalies as compared to one major anomaly or no anomalies. Finally, we reported several novel associations, such as between congenital anomalies of the central nervous system and embryonal RMS. The congenital anomaly-sarcoma associations observed in this study may be indicative of etiologic heterogeneity among pediatric sarcomas, and unrecognized or undiscovered cancer predisposition syndromes. Future work should investigate shared developmental pathways between these conditions by identifying pleiotropic genetic variants associated with both conditions and evaluate the utility of these findings for risk prediction and surveillance strategies.

Supplementary Material

Supplementary tables 1-5
Supplementary figures 1 and 2

Acknowledgements

The authors extend their gratitude to the many investigators and staff who made essential contributions to the GOBACK Registry Linkage Cohort. This research was supported by the Rally Foundation for Childhood Cancer Research (Career Development Award 23CND05), the National Cancer Institute (1R01CA284531), and the Department of Defense (Impact Award CA190214). This research was also supported in part by a grant from the Cancer Prevention and Research Institute of Texas (CPRIT) for the Systems Epidemiology of Cancer Training (SECT) Program (RP210037; PI: Thrift A).

Abbreviations key

RMS

Rhabdomyosarcoma

NRSTS

Non-rhabdomyosarcoma soft tissue sarcoma

ES

Ewing sarcoma

OS

Osteosarcoma

HR

Hazard ratio

CI

Confidence interval

MPNST

Malignant peripheral nerve sheath tumor

SEER

Surveillance, Epidemiology, and End Results

CDC

Center for Disease Control

GOBACK

Genetic Overlap Between Anomalies and Cancer in Kids

Footnotes

Conflict of Interest statement

The authors have no conflicts of interest to declare.

References

  • 1.Kaatsch P. Epidemiology of childhood cancer. Cancer treatment reviews. 2010;36(4):277–285. [DOI] [PubMed] [Google Scholar]
  • 2.Smith MA, Altekruse SF, Adamson PC, Reaman GH, Seibel NL. Declining childhood and adolescent cancer mortality. Cancer. 2014;120(16):2497–2506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Siegel RL, Miller KD, Jemal A. Cancer statistics, 2018. CA: a cancer journal for clinicians. 2018;68(1):7–30. [DOI] [PubMed] [Google Scholar]
  • 4.Sbaraglia M, Bellan E, Dei Tos AP. The 2020 WHO classification of soft tissue tumours: news and perspectives. Pathologica. 2020;113(2):70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.SEER*Explorer. An interactive website for SEER cancer statistics 2023. https://seer.cancer.gov/statistics-network/explorer/
  • 6.Stallings EB, Isenburg JL, Rutkowski RE, et al. National population-based estimates for major birth defects, 2016–2020. Birth defects research. 2024;116(1):e2301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Martin-Giacalone BA, Rideau T-T, Scheurer ME, Lupo PJ, Wang LL. Cancer risk among RECQL4 heterozygotes. Cancer genetics. 2022;262:107–110. [DOI] [PubMed] [Google Scholar]
  • 8.Lupo PJ, Luna-Gierke RE, Chambers TM, et al. Perinatal and familial risk factors for soft tissue sarcomas in childhood through young adulthood: A population-based assessment in 4 million live births. International Journal of Cancer. 2020;146(3):791–802. [DOI] [PubMed] [Google Scholar]
  • 9.Berbel-Tornero O, Ortega-Garcia J, Ferris-Tortajada J. Congenital abnormalities and childhood cancer: A cohort record-linkage study. Cancer: Interdisciplinary International Journal of the American Cancer Society. 2006;106(6):1418–1419. [DOI] [PubMed] [Google Scholar]
  • 10.Norwood MS, Lupo PJ, Chow EJ, et al. Childhood cancer risk in those with chromosomal and non-chromosomal congenital anomalies in Washington State: 1984–2013. PloS one. 2017;12(6):e0179006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Schumacher R, Mai A, Gutjahr P. Association of rib anomalies and malignancy in childhood. European Journal of Pediatrics. 1992;151(6):432–434. [DOI] [PubMed] [Google Scholar]
  • 12.Altmann AE, Halliday J, Giles G. Associations between congenital malformations and childhood cancer. A register-based case-control study. British journal of cancer. 1998;78(9):1244–1249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Agha MM, Williams JI, Marrett L, To T, Zipursky A, Dodds L. Congenital abnormalities and childhood cancer: a cohort record-linkage study. Cancer. 2005;103(9):1939–1948. [DOI] [PubMed] [Google Scholar]
  • 14.Carozza SE, Langlois PH, Miller EA, Canfield M. Are children with birth defects at higher risk of childhood cancers? American journal of epidemiology. 2012;175(12):1217–1224. [DOI] [PubMed] [Google Scholar]
  • 15.Janitz AE, Neas BR, Campbell JE, et al. Childhood cancer in children with congenital anomalies in Oklahoma, 1997 to 2009. Birth Defects Research Part A: Clinical and Molecular Teratology. 2016;106(7):633–642. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Botto LD, Flood T, Little J, et al. Cancer risk in children and adolescents with birth defects: a population-based cohort study. PloS one. 2013;8(7):e69077. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Zierhut H, Murati M, Holm T, Hoggard E, Spector LG. Association of rib anomalies and childhood cancers. British journal of cancer. 2011;105(9):1392–1395. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Mann J, Dodd H, Draper G, et al. Congenital abnormalities in children with cancer and their relatives: results from a case-control study (IRESCC). British journal of cancer. 1993;68(2):357–363. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Gelberg KH, Fitzgerald EF, Hwang S, Dubrow R. Growth and development and other risk factors for osteosarcoma in children and young adults. International journal of epidemiology. 1997;26(2):272–278. [DOI] [PubMed] [Google Scholar]
  • 20.Dawson S, Charles AK, Bower C, de Klerk NH, Milne E. Risk of cancer among children with birth defects: a novel approach. Birth Defects Research Part A: Clinical and Molecular Teratology. 2015;103(4):284–291. [DOI] [PubMed] [Google Scholar]
  • 21.Savitz DA, Ananth CV. Birth characteristics of childhood cancer cases, controls, and their siblings. Pediatric hematology and oncology. 1994;11(6):587–599. [DOI] [PubMed] [Google Scholar]
  • 22.Lupo PJ, Schraw JM, Desrosiers TA, et al. Association between birth defects and cancer risk among children and adolescents in a population-based assessment of 10 million live births. JAMA oncology. 2019;5(8):1150–1158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Mai CT, Cassell CH, Meyer RE, et al. Birth defects data from population-based birth defects surveillance programs in the United States, 2007 to 2011: highlighting orofacial clefts. Birth defects research Part A, Clinical and molecular teratology. 2014;100(11):895. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Richiardi L, Bellocco R, Zugna D. Mediation analysis in epidemiology: methods, interpretation and bias. International journal of epidemiology. 2013;42(5):1511–1519. [DOI] [PubMed] [Google Scholar]
  • 25.Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal statistical society: series B (Methodological). 1995;57(1):289–300. [Google Scholar]
  • 26.Kassambara A, Kosinski M, Biecek P. survminer: Drawing Survival Curves using’ggplot2’. CRAN: Contributed Packages. 2016; [Google Scholar]
  • 27.Kar E, Ammanamanchi A, Yousif M, et al. From bimodal to unimodal: The transformed incidence of osteosarcoma in the United States. Journal of Bone Oncology. 2024;47:100613. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Hosseini H, Heydari S, Hushmandi K, Daneshi S, Raesi R. Bone tumors: a systematic review of prevalence, risk determinants, and survival patterns. BMC cancer. 2025;25(1):321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Ruymann FB, Maddux HR, Ragab A, et al. Congenital anomalies associated with rhabdomyosarcoma: An autopsy study of 115 cases. A report from the intergroup rhabdomyosarcoma study committee (representing the children’s cancer study group, the pediatric oncology group, the United Kingdom children’s cancer study group, and the pediatric intergroup statistical center). Medical and pediatric oncology. 1988;16(1):33–39. [DOI] [PubMed] [Google Scholar]
  • 30.Murdoch JN, Copp AJ. The relationship between sonic Hedgehog signaling, cilia, and neural tube defects. Birth Defects Research Part A: Clinical and Molecular Teratology. 2010;88(8):633–652. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Marigo V, Tabin CJ. Regulation of patched by sonic hedgehog in the developing neural tube. Proceedings of the National Academy of Sciences. 1996;93(18):9346–9351. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Haraguchi R, Motoyama J, Sasaki H, et al. Molecular analysis of coordinated bladder and urogenital organ formation by Hedgehog signaling. 2007; [DOI] [PubMed]
  • 33.Perriton CL, Powles N, Chiang C, Maconochie MK, Cohn MJ. Sonic hedgehog signaling from the urethral epithelium controls external genital development. Developmental biology. 2002;247(1):26–46. [DOI] [PubMed] [Google Scholar]
  • 34.Zarzosa P, Garcia-Gilabert L, Hladun R, et al. Targeting the hedgehog pathway in rhabdomyosarcoma. Cancers. 2023;15(3):727. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Zibat A, Missiaglia E, Rosenberger A, et al. Activation of the hedgehog pathway confers a poor prognosis in embryonal and fusion gene-negative alveolar rhabdomyosarcoma. Oncogene. 2010;29(48):6323–6330. [DOI] [PubMed] [Google Scholar]
  • 36.Pressey JG, Anderson JR, Crossman DK, Lynch JC, Barr FG. Hedgehog pathway activity in pediatric embryonal rhabdomyosarcoma and undifferentiated sarcoma: a report from the Children’s Oncology Group. Pediatric blood & cancer. 2011;57(6):930–938. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Evans DGR, Baser ME, McGaughran J, Sharif S, Howard E, Moran A. Malignant peripheral nerve sheath tumours in neurofibromatosis 1. Journal of medical genetics. 2002;39(5):311–314. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Ferner RE, Gutmann DH. International consensus statement on malignant peripheral nerve sheath tumors in neurofibromatosis. Cancer Res. 2002. Mar 1;62(5):1573–7. [PubMed] [Google Scholar]
  • 39.Schraw JM, Desrosiers TA, Nembhard WN, et al. Cancer diagnostic profile in children with structural birth defects: An assessment in 15,000 childhood cancer cases. Cancer. 2020;126(15):3483–3492. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Johnson KJ, Lee JM, Ahsan K, et al. Pediatric cancer risk in association with birth defects: A systematic review. PloS one. 2017;12(7):e0181246. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Clark CJ, Warren JL, Saiers JE, Ma X, Bell ML, Deziel NC. Predictors of early life residential mobility in urban and rural Pennsylvania children with acute lymphoblastic leukemia and implications for environmental exposure assessment. Journal of exposure science & environmental epidemiology. 2024;34(6):990–999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Urayama KY, Von Behren J, Reynolds P, Hertz A, Does M, Buffler PA. Factors associated with residential mobility in children with leukemia: implications for assigning exposures. Annals of epidemiology. 2009;19(11):834–840. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Danysh HE, Mitchell LE, Zhang K, Scheurer ME, Lupo PJ. Differences in environmental exposure assignment due to residential mobility among children with a central nervous system tumor: Texas, 1995–2009. Journal of Exposure Science & Environmental Epidemiology. 2017;27(1):41–46. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary tables 1-5
NIHMS2211470-supplement-Supplementary_tables_1-5.xlsx (23.2KB, xlsx)
Supplementary figures 1 and 2
NIHMS2211470-supplement-Supplementary_figures_1_and_2.docx (78.4KB, docx)

RESOURCES