Background:

Synovial sarcoma is a rare cancer with peak incidence in the young adult period. Despite poor outcomes of this aggressive cancer, there is little epidemiologic research addressing its etiology.

Methods:

We collected birth characteristic data on synovial sarcoma cases born during 1978–2015 and diagnosed during 1988–2015 in California (n = 244), and 12,200 controls frequency-matched on year of birth. We also constructed a dataset of cancer cases in siblings of sarcoma subjects to assess familial risk.

Results:

In multivariable logistic regression analyses, synovial sarcoma was more frequent in Hispanics compared with non-Hispanic whites [OR, 1.48; 95% confidence interval (CI), 1.06–2.08]. Higher birth weight was a risk factor in Hispanics; each 500 g increase in birth weight was associated with a 22% increase in disease risk (OR, 1.22; 95% CI, 1.00–1.48). Also, a strong role for birth order was suggested, with highest risk for the first born (second child compared with first: OR, 0.61; 95% CI, 0.44–0.84; third or later compared with first: OR, 0.53; 95% CI, 0.36–0.77). Siblings of patients with synovial sarcoma did not display elevated cancer incidence, suggesting the low likelihood that strong familial predisposition alleles play a significant role in this disease.

Conclusions:

The associations with birth weight and birth order suggest that nutritional, developmental, and environmental factors may play a role in the etiology of synovial sarcoma.

Impact:

Further epidemiologic research on synovial sarcoma should evaluate epigenetic and developmental mechanisms and the formation of the archetypical t(X;18) translocation that defines this disease.

This article is featured in Highlights of This Issue, p. 1079

Synovial sarcoma, the second most common soft tissue sarcoma in children and young adults, is typically an invasive and metastatic tumor with poor prognosis. It does not display synovial features despite commonly arising from positions proximal to developing joints; instead, the tumors have epithelial histology with spindle or biphasic morphology (1). A primary characterizing feature of synovial sarcoma is the t(X;18; p11;q11) translocation resulting in a fusion between the entire coding region of the SS18 gene and a portion of an SSX gene (SSX1, SSX2, or SSX4; ref. 2). The singular association of this fusion gene with synovial sarcoma suggests that the causes of its formation may be the key to understanding the genesis of this disease. The fusion oncogene participates in chromatin remodeling complexes, underscoring the possibility that epigenetic alterations may represent a major component of the transformation process (2). Epidemiologic data on etiology specific to synovial sarcoma are scant, which is likely attributable to its rarity. More common sarcomas have been included in prior studies, and such research has noted associations with parental age (3), gestational age (4, 5), birth weight (5–8), and birth order (9). We examined the potential etiologic role of such birth characteristics here in a large, population-based case–control study of synovial sarcoma in California, as well as any indication for familial cancer clustering (reflective of genetic predisposition) using high quality data from the California Cancer Registry (CCR) and statewide birth records.

We constructed two separate datasets by merging data from the CCR and California birth records to examine birth characteristics and familial cancer clustering, respectively. The study protocol was approved by the Institutional Review Boards at the California Health and Human Services Agency, University of Southern California (Los Angeles, CA), University of California at Berkeley (Berkeley, CA), and Yale University (New Haven, CT).

Birth characteristics

We identified a total of 244 synovial sarcoma cases who were born in California during 1978–2015, diagnosed with synovial sarcoma [International Classification of Diseases (ICD) for Oncology, 3rd edition, ICD-O-3 codes: 9040–9043] at the age of 0–35 years during 1988–2015, and reported to the CCR. Statewide birth records maintained by the California Department of Public Health were used to randomly select 50 times as many control subjects (n = 12,200) who were frequency matched to the cases by year of birth; none of the controls had been diagnosed with any type of cancer up to the age of 35 years based on CCR records.

For all cases and controls, data on the following variables were retrieved from their birth records: sex, race/ethnicity (white, black, Hispanic/Latino, Asian/Pacific Islander, and other), birth weight, gestational age, birth plurality, birth order, mode of delivery (vaginal or cesarean section), year of birth, maternal and paternal age, maternal education, mother's place of birth (United States/foreign), history of miscarriage/stillbirth (yes/no), complication during pregnancy (yes/no), and maternal history of cesarean section (yes/no). A multivariable unconditional logistic regression analysis was performed with case status as the outcome and all birth characteristics described above as independent variables. In addition, stratified analyses were performed for larger racial/ethnic groups (Hispanic and non-Hispanic whites), age at diagnosis, and histology subtype.

Familial cancer clustering

To assess potential familial aggregation that may reflect genetic predisposition to cancer, we captured the siblings of all young patients with synovial sarcoma (ages 0–19, 127 cases) from the statewide birth records and examined whether any of them had been diagnosed with any type of cancer per CCR record. We calculated standardized incidence ratios (SIR) for siblings' RR by dividing the observed number of cancer cases by the expected number of cases among siblings based on age-specific cancer incidence rates derived from the Surveillance Epidemiology and End Results (SEER) program (10). Similar analyses were performed for comparison on additional sarcomas including Ewing sarcoma (ICD-O-3 code 9260; 353 cases), osteosarcoma (ICD-O-3 codes: 9180–83, 9185–87, and 9192–95; 576 cases), and rhabdomyosarcoma (ICD-O-3 codes: 8900–02, 8910, 8912, 8920, and 8991; 719 cases).

All tests were two-sided with an alpha of 0.05 and were conducted using SAS version 9.4 (SAS Inc.).

Of the 244 cases with synovial sarcoma, 94 were recorded with ICD-O-3 code 9040 (synovial, not otherwise specified), 92 with ICD-O-3 code 9041 (spindle cell), one with ICD-O-3 code 9042 (epithelioid), and 57 with ICD-O-3 code 9043 (biphasic). A comparison with the number of California statewide cases reported to the SEER program during 2000–2015 suggested that we captured 77% of cases diagnosed under the age of 15 years and 56% of cases diagnosed at the age of 15–35 years; the noncaptured cases were presumably born outside of California. Of the 244 cases, 228 (93%) had their sarcoma located in connective tissue; among the 228 cases, the most commonly observed locations were the lower limb or hip (53%) and upper limb/shoulder (23%). Cases were 56.6% male which was not significantly higher than controls, and the largest group of cases and controls were Hispanic (Table 1). While both younger and older mothers and fathers were in excess among cases, parental ages were not significantly associated with case/control status. The only factor that appeared to significantly predict risk was “birth order,” where synovial sarcoma cases were more likely to be first born compared with controls in the bivariate χ2 analysis (P < 0.01; Table 1).

Table 1.

Characteristics of subjects with synovial sarcoma and controls, California, 1978–2015.

Synovial sarcomaControl group
n = 244n = 12,200
Number of participants%Number of participants%Pa
Overall 244  12,200   
Sex 
 Female 106 43.4 5,940 48.7 0.10 
 Male 138 56.6 6,260 51.3  
Race/ethnicity 
 White 93 38.1 4,834 39.6 0.45 
 Black 17 7.0 1,097 9.0  
 Hispanic 112 45.9 4,965 40.7  
 Asian 20 8.2 1,139 9.3  
 Other 0.8 165 1.4  
Birth weight (grams) 
 250–2,499 15 6.1 721 5.9 0.76 
 2,500–2,999 32 13.1 1,928 15.8  
 3,000–3,499 95 38.9 4,480 36.7  
 3,500–3,999 71 29.1 3,674 30.1  
 4,000+ 31 12.7 1,397 11.5  
Gestational age (weeks) 
 37–41 19 7.8 1,118 9.2 0.56 
 22–36 183 75.0 8,820 72.3  
 42–44 20 8.2 1,268 10.4  
 Unknown 22 9.0 994 8.1  
Birth plurality 
 Singleton 239 98.0 11,927 97.8 0.84 
 Multiple 2.0 273 2.2  
Birth order 
 1st 128 52.5 4,873 39.9 <0.01 
 2nd 66 27.0 3,932 32.2  
 3rd and higher 50 20.5 3,395 27.8  
Mode of delivery 
 Vaginal 193 79.1 9,630 78.9 0.95 
 Cesarean section 51 20.9 2,570 21.1  
Year of birth 
 1978–1982 47 19.3 2,350 19.3 1.00 
 1983–1987 66 27.0 3,300 27.0  
 1988–1992 66 27.0 3,300 27.0  
 1993–2013 65 26.6 3,250 26.6  
Maternal education 
 Up to 8 years 10 4.1 981 8.0 0.20 
 9–11 years 23 9.4 1,171 9.6  
 12 years 47 19.3 2,079 17.0  
 13–15 years 31 12.7 1,365 11.2  
 16 or more years 24 9.8 1,114 9.1  
 Unknown 109 44.7 5,490 45.0  
Mother's place of birth 
 United States 157 64.3 7,668 62.9 0.63 
 Foreign 87 35.7 4,532 37.1  
Miscarriage/stillbirth 
 Never 205 84.0 10,114 82.9 0.60 
 Ever 38 15.6 2,060 16.9  
 Unknown 0.4 26 0.2  
Maternal complication during pregnancy 
 Never 196 80.3 9,872 80.9 0.79 
 Ever 37 15.2 1,775 14.5  
 Unknown 11 4.5 553 4.5  
Previous cesarean section 
 Never 221 90.6 10,735 88.0 0.23 
 Ever 15 6.1 1,004 8.2  
 Unknown 3.3 461 3.8  
Maternal age (years) 
 <20 30 12.3 1,484 12.2 0.44 
 20–24 76 31.1 3,373 27.6  
 25–29 60 24.6 3,591 29.4  
 30–34 56 23.0 2,532 20.8  
 ≥35 22 9.0 1,220 10.0  
Paternal age (years) 
 <25 66 27.0 2,938 24.1 0.11 
 25–29 78 32.0 3,421 28.0  
 30–34 42 17.2 2,901 23.8  
 35–39 31 12.7 1,484 12.2  
 ≥40 22 9.0 865 7.1  
 Unknown 2.0 591 4.8  
Synovial sarcomaControl group
n = 244n = 12,200
Number of participants%Number of participants%Pa
Overall 244  12,200   
Sex 
 Female 106 43.4 5,940 48.7 0.10 
 Male 138 56.6 6,260 51.3  
Race/ethnicity 
 White 93 38.1 4,834 39.6 0.45 
 Black 17 7.0 1,097 9.0  
 Hispanic 112 45.9 4,965 40.7  
 Asian 20 8.2 1,139 9.3  
 Other 0.8 165 1.4  
Birth weight (grams) 
 250–2,499 15 6.1 721 5.9 0.76 
 2,500–2,999 32 13.1 1,928 15.8  
 3,000–3,499 95 38.9 4,480 36.7  
 3,500–3,999 71 29.1 3,674 30.1  
 4,000+ 31 12.7 1,397 11.5  
Gestational age (weeks) 
 37–41 19 7.8 1,118 9.2 0.56 
 22–36 183 75.0 8,820 72.3  
 42–44 20 8.2 1,268 10.4  
 Unknown 22 9.0 994 8.1  
Birth plurality 
 Singleton 239 98.0 11,927 97.8 0.84 
 Multiple 2.0 273 2.2  
Birth order 
 1st 128 52.5 4,873 39.9 <0.01 
 2nd 66 27.0 3,932 32.2  
 3rd and higher 50 20.5 3,395 27.8  
Mode of delivery 
 Vaginal 193 79.1 9,630 78.9 0.95 
 Cesarean section 51 20.9 2,570 21.1  
Year of birth 
 1978–1982 47 19.3 2,350 19.3 1.00 
 1983–1987 66 27.0 3,300 27.0  
 1988–1992 66 27.0 3,300 27.0  
 1993–2013 65 26.6 3,250 26.6  
Maternal education 
 Up to 8 years 10 4.1 981 8.0 0.20 
 9–11 years 23 9.4 1,171 9.6  
 12 years 47 19.3 2,079 17.0  
 13–15 years 31 12.7 1,365 11.2  
 16 or more years 24 9.8 1,114 9.1  
 Unknown 109 44.7 5,490 45.0  
Mother's place of birth 
 United States 157 64.3 7,668 62.9 0.63 
 Foreign 87 35.7 4,532 37.1  
Miscarriage/stillbirth 
 Never 205 84.0 10,114 82.9 0.60 
 Ever 38 15.6 2,060 16.9  
 Unknown 0.4 26 0.2  
Maternal complication during pregnancy 
 Never 196 80.3 9,872 80.9 0.79 
 Ever 37 15.2 1,775 14.5  
 Unknown 11 4.5 553 4.5  
Previous cesarean section 
 Never 221 90.6 10,735 88.0 0.23 
 Ever 15 6.1 1,004 8.2  
 Unknown 3.3 461 3.8  
Maternal age (years) 
 <20 30 12.3 1,484 12.2 0.44 
 20–24 76 31.1 3,373 27.6  
 25–29 60 24.6 3,591 29.4  
 30–34 56 23.0 2,532 20.8  
 ≥35 22 9.0 1,220 10.0  
Paternal age (years) 
 <25 66 27.0 2,938 24.1 0.11 
 25–29 78 32.0 3,421 28.0  
 30–34 42 17.2 2,901 23.8  
 35–39 31 12.7 1,484 12.2  
 ≥40 22 9.0 865 7.1  
 Unknown 2.0 591 4.8  

aP values were derived from χ2 tests.

Hispanics appeared to have a higher risk of synovial sarcoma than non-Hispanic whites [OR, 1.48; 95% confidence interval (CI), 1.06–2.08; Table 2), especially for synovial sarcoma diagnosed in young adults (age 19–35 years; OR, 1.94; 95% CI, 1.12–3.35; Supplementary Table S1) and spindle cell sarcoma (OR, 2.30; 95% CI, 1.27–4.17; Supplementary Table S1). Higher birth weight was a risk factor for synovial sarcoma in Hispanics, each 500 g increase in birth weight was associated with a 22% increase in disease risk (OR, 1.22; 95% CI, 1.00–1.48; Table 2). No relationship with birth weight was demonstrated among non-Hispanic whites. A similar association was observed between higher birth weight and the risk of pediatric synovial sarcoma (age 0–19 years; OR, 1.21; 95% CI, 1.03–1.42 for each 500 g increase in birth weight; Supplementary Table S1), with no association noted for those patients with synovial sarcoma diagnosed over 19 years.

Table 2.

ORs for the risk for synovial sarcoma according to race/ethnicity, birth weight and birth order, and parental age, California 1978–2015.

non-Hispanic White onlyHispanic only
(n = 93 cases(n = 112 cases
Entire population (n = 244 cases and 12,200 controls)1,939 controls)2,345 controls)
CaseControlUnadjustedAdjustedaAdjustedaAdjusteda
n (%)n (%)OR (95% CI)OR (95% CI)OR (95% CI)OR (95% CI)
Sex 
 Female 106 (43.3) 5,940 (48.7) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 
 Male 138 (56.6) 6,260 (51.3) 1.24 (0.96–1.60) 1.21 (0.94–1.57) 1.44 (0.93–2.24) 0.96 (0.65–1.43) 
Race/ethnicity 
 White 93 (38.1) 4,834 (39.6) 1.00 (Reference) 1.00 (Reference)   
 Black 17 (7.0) 1,097 (9.0) 0.81 (0.48–1.36) 0.90 (0.53–1.54)   
 Hispanic 112 (45.9) 4,965 (40.7) 1.17 (0.89–1.55) 1.48 (1.06–2.08)   
 Asian 20 (8.2) 1,139 (9.3) 0.91 (0.56–1.49) 1.03 (0.60–1.78)   
 Other 2 (0.8) 165 (1.4) 0.63 (0.15–2.58) 0.67 (0.16–2.78)   
Birth weight (grams) 
 250–2,499 15 (6.1) 721 (5.9) 0.98 (0.57–1.70) 1.06 (0.57–1.98) 0.94 (0.29–3.12) 0.78 (0.27–2.23) 
 2,500–2,999 32 (13.1) 1,928 (15.8) 0.78 (0.52–1.17) 0.78 (0.52–1.18) 0.71 (0.34–1.48) 0.80 (0.43–1.52) 
 3,000–3,499 95 (38.9) 4,480 (36.7) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 
 3,500–3,999 71 (29.1) 3,674 (30.1) 0.91 (0.67–1.24) 0.90 (0.66–1.23) 0.70 (0.42–1.19) 1.15 (0.72–1.82) 
 4,000+ 31 (12.7) 1,397 (11.5) 1.05 (0.69–1.58) 1.09 (0.72–1.66) 1.00 (0.52–1.90) 1.26 (0.65–2.47) 
 Every 500 g  1.08 (0.95–1.23) 1.07 (0.87–1.33) 1.22 (1.00–1.48) 
Birth order 
 1st 128 (52.5) 4,873 (39.9) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 
 2nd 66 (27.0) 3,932 (32.2) 0.64 (0.47–0.86) 0.61 (0.44–0.84) 0.58 (0.35–0.98) 0.54 (0.32–0.89) 
 3rd and higher 50 (20.5) 3,395 (27.8) 0.56 (0.40–0.78) 0.53 (0.36–0.77) 0.46 (0.23–0.90) 0.53 (0.30–0.95) 
Maternal age (years) 
 <20 30 (12.3) 1,484 (12.2) 1.21 (0.78–1.88) 0.92 (0.53–1.59) 0.79 (0.24–2.63) 0.97 (0.46–2.07) 
 20–24 76 (31.1) 3,373 (27.6) 1.35 (0.96–1.90) 1.13 (0.78–1.66) 1.16 (0.59–2.30) 1.23 (0.70–2.16) 
 25–29 60 (24.6) 3,591 (29.4) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 
 30–34 56 (23.0) 2,532 (20.8) 1.32 (0.92–1.91) 1.49 (1.00–2.23) 2.61 (1.39–4.88) 0.68 (0.31–1.50) 
 ≥35 22 (9.0) 1,220 (10.0) 1.08 (0.66–1.77) 1.06 (0.59–1.91) 1.08 (0.41–2.88) 1.81 (0.77–4.30) 
Paternal age (years) 
 <25 66 (27.0) 2,938 (24.1) 1.55 (1.05–2.29) 1.41 (0.86–2.32) 1.35 (0.56–3.25) 1.72 (0.82–3.62) 
 25–29 78 (32.0) 3,421 (28.0) 1.57 (1.08–2.30) 1.60 (1.06–2.41) 1.87 (0.97–3.60) 1.59 (0.80–3.16) 
 30–34 42 (17.2) 2,901 (23.8) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 
 35–39 31 (12.7) 1,484 (12.2) 1.44 (0.90–2.30) 1.49 (0.92–2.41) 1.60 (0.80–3.20) 2.07 (0.87–4.90) 
 ≥40 22 (9.0) 865 (7.1) 1.76 (1.04–2.96) 1.90 (1.07–3.35) 2.39 (1.05–5.44) 1.32 (0.45–3.90) 
 Unknown 5 (2.0) 591 (4.8) 0.58 (0.23–1.48) 0.63 (0.24–1.64)  0.54 (0.12–2.50) 
non-Hispanic White onlyHispanic only
(n = 93 cases(n = 112 cases
Entire population (n = 244 cases and 12,200 controls)1,939 controls)2,345 controls)
CaseControlUnadjustedAdjustedaAdjustedaAdjusteda
n (%)n (%)OR (95% CI)OR (95% CI)OR (95% CI)OR (95% CI)
Sex 
 Female 106 (43.3) 5,940 (48.7) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 
 Male 138 (56.6) 6,260 (51.3) 1.24 (0.96–1.60) 1.21 (0.94–1.57) 1.44 (0.93–2.24) 0.96 (0.65–1.43) 
Race/ethnicity 
 White 93 (38.1) 4,834 (39.6) 1.00 (Reference) 1.00 (Reference)   
 Black 17 (7.0) 1,097 (9.0) 0.81 (0.48–1.36) 0.90 (0.53–1.54)   
 Hispanic 112 (45.9) 4,965 (40.7) 1.17 (0.89–1.55) 1.48 (1.06–2.08)   
 Asian 20 (8.2) 1,139 (9.3) 0.91 (0.56–1.49) 1.03 (0.60–1.78)   
 Other 2 (0.8) 165 (1.4) 0.63 (0.15–2.58) 0.67 (0.16–2.78)   
Birth weight (grams) 
 250–2,499 15 (6.1) 721 (5.9) 0.98 (0.57–1.70) 1.06 (0.57–1.98) 0.94 (0.29–3.12) 0.78 (0.27–2.23) 
 2,500–2,999 32 (13.1) 1,928 (15.8) 0.78 (0.52–1.17) 0.78 (0.52–1.18) 0.71 (0.34–1.48) 0.80 (0.43–1.52) 
 3,000–3,499 95 (38.9) 4,480 (36.7) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 
 3,500–3,999 71 (29.1) 3,674 (30.1) 0.91 (0.67–1.24) 0.90 (0.66–1.23) 0.70 (0.42–1.19) 1.15 (0.72–1.82) 
 4,000+ 31 (12.7) 1,397 (11.5) 1.05 (0.69–1.58) 1.09 (0.72–1.66) 1.00 (0.52–1.90) 1.26 (0.65–2.47) 
 Every 500 g  1.08 (0.95–1.23) 1.07 (0.87–1.33) 1.22 (1.00–1.48) 
Birth order 
 1st 128 (52.5) 4,873 (39.9) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 
 2nd 66 (27.0) 3,932 (32.2) 0.64 (0.47–0.86) 0.61 (0.44–0.84) 0.58 (0.35–0.98) 0.54 (0.32–0.89) 
 3rd and higher 50 (20.5) 3,395 (27.8) 0.56 (0.40–0.78) 0.53 (0.36–0.77) 0.46 (0.23–0.90) 0.53 (0.30–0.95) 
Maternal age (years) 
 <20 30 (12.3) 1,484 (12.2) 1.21 (0.78–1.88) 0.92 (0.53–1.59) 0.79 (0.24–2.63) 0.97 (0.46–2.07) 
 20–24 76 (31.1) 3,373 (27.6) 1.35 (0.96–1.90) 1.13 (0.78–1.66) 1.16 (0.59–2.30) 1.23 (0.70–2.16) 
 25–29 60 (24.6) 3,591 (29.4) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 
 30–34 56 (23.0) 2,532 (20.8) 1.32 (0.92–1.91) 1.49 (1.00–2.23) 2.61 (1.39–4.88) 0.68 (0.31–1.50) 
 ≥35 22 (9.0) 1,220 (10.0) 1.08 (0.66–1.77) 1.06 (0.59–1.91) 1.08 (0.41–2.88) 1.81 (0.77–4.30) 
Paternal age (years) 
 <25 66 (27.0) 2,938 (24.1) 1.55 (1.05–2.29) 1.41 (0.86–2.32) 1.35 (0.56–3.25) 1.72 (0.82–3.62) 
 25–29 78 (32.0) 3,421 (28.0) 1.57 (1.08–2.30) 1.60 (1.06–2.41) 1.87 (0.97–3.60) 1.59 (0.80–3.16) 
 30–34 42 (17.2) 2,901 (23.8) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 1.00 (Reference) 
 35–39 31 (12.7) 1,484 (12.2) 1.44 (0.90–2.30) 1.49 (0.92–2.41) 1.60 (0.80–3.20) 2.07 (0.87–4.90) 
 ≥40 22 (9.0) 865 (7.1) 1.76 (1.04–2.96) 1.90 (1.07–3.35) 2.39 (1.05–5.44) 1.32 (0.45–3.90) 
 Unknown 5 (2.0) 591 (4.8) 0.58 (0.23–1.48) 0.63 (0.24–1.64)  0.54 (0.12–2.50) 

aThe adjusted ORs were derived from multivariable logistic regression models in which all the variables listed in the table were included simultaneously. Three separate models were fit for the overall study population, Hispanics, and non-Hispanic Whites, respectively.

Compared with first borns, those whose birth order was second or higher had a significantly lower risk of synovial sarcoma (second: OR, 0.61; 95% CI, 0.44–0.84; third or higher: OR, 0.53; 95% CI, 0.36–0.77; Table 2). This association was consistently observed across major racial/ethnic groups (Hispanics and non-Hispanic whites), both the pediatric and young adult populations, as well as histopathology and tumor site (Supplementary Table S1). Another familial characteristic, parental age, also demonstrated some marginal, but potentially notable associations. Both maternal and paternal age was divided into five groups, using the central group as the reference. In unadjusted bivariate analyses, both younger and older parental ages appeared to carry some risk, particularly for paternal age (Table 2). When adjusting for each other (and other birth characteristics), some of these relationships sustained, suggesting that increased or decreased maternal and paternal ages independently influence risk of synovial sarcoma. For instance, paternal age over 40 years was associated with an OR of 1.90 (95% CI, 1.07–3.35; Table 2) when adjusting for maternal age and other covariates for synovial sarcoma overall. The other factors assessed here, including gestational age, birth plurality, mode of delivery, maternal history of fetal loss, and pregnancy complications, were not significantly associated with synovial sarcoma risk.

Our family-based analysis included a total of 127 families with a case of pediatric synovial sarcoma (diagnosed at the age of 0–19 years). Of all siblings identified (n = 153), none had been diagnosed with any types of cancer per CCR record, rendering an inability to estimate SIR. For other sarcomas of childhood, we identified cancer diagnoses among siblings of patients with pediatric Ewing sarcoma (SIR, 1.79; 95% CI, 0.47–6.85), osteosarcoma (SIR, 3.83; 95% CI, 1.44–10.20), and rhabdomyosarcoma (SIR, 6.47; 95% CI, 3.01–13.95). On comparing father's birthdates, we found that 14% of siblings in this analysis with a shared mother had a nonshared father. These siblings will only share the genetic complement from their mothers and therefore our familial clustering analysis will be slightly biased toward the null.

Synovial sarcoma is the second most common soft tissue sarcoma across all ages (after rhabdomyosarcoma), and typically presents as a high-grade disease with local invasion and a high propensity to metastasize. While a variety of single institution and multicenter survival and prognostic studies on synovial sarcoma populate the literature, we believe this is the first study examining the role of birth characteristics in the etiology of synovial sarcoma.

Our analysis highlights several unknown or underappreciated aspects of the epidemiology of synovial sarcoma. First, Hispanics appear to have an increased risk of the disease, particularly the spindle cell form. Higher birth weight was a risk factor for synovial sarcoma in Hispanics but not the other racial/ethnic groups. Hispanics are known to exhibit faster growth rates than non-Hispanic whites in mid-gestation (11). When stratifying by age, the birth weight association was only observed for those who were diagnosed at younger ages (0–19 years) and null for young adults (>19 years). It should be reiterated that the association of synovial sarcoma with Hispanic ethnicity was independent of birth weight, and therefore the birth weight association here exclusively among Hispanics suggests that cultural or nutritional differences among Hispanics may drive the association possibly interacting with ethnic-specific genetic factors. Embryogenesis and adolescence are two points in life where the fastest growth rates occur for the human organism, which may implicate hormonal or nutritional influences on synovial sarcoma growth or genetic propensity for growth. These periods of rapid growth correspond to periods of sensitivity to environmental agents, which could impact synovial sarcoma risk, underscoring the need for further investigation particularly among Hispanics.

Our most intriguing finding is the strong association between birth order and synovial sarcoma risk, which is consistently observed across different racial/ethnic groups and different ages of diagnosis. Birth order is known to impact risk of childhood cancer overall (OR for subsequent births compared with the first = 0.87; 95% CI, 0.81–0.93 in a multistate analysis including California; ref. 9). While it is not necessarily a causal factor in itself, birth order is associated with a variety of developmental and environmental factors. First, later-born children are presumably exposed to more infectious agents earlier in life than their first-born siblings, potentially affecting immune system development. Having a higher birth order is thought to reduce risk of diseases such as atopy (12), leukemia (13–15), and type-1 diabetes (16) and increase risk of non-Hodgkin lymphoma (17), to name a few. Second, first-born children experience differential nutrition during pregnancy compared with their later-born siblings; specifically, placentation is less efficient during first pregnancies, resulting in a degree of nutrient restriction in utero compared with later borns (18). The first born tends to be taller, thinner, have greater IGF-1 concentrations at birth, and exhibit a reduction in insulin sensitivity (19). Third, estrogens are higher in first births compared with later births (20, 21). Birth order has not been consistently linked to the risk of breast cancer but has been suggested to play a role in the etiology of testicular cancer and adult glioma (22, 23). Fourth and finally, birth order is related to fetal microchimerism, in which circulating cells from a former pregnancy may be passed through the mother to subsequent pregnancies. This phenomenon has implications for immunologic diseases such as scleroderma (24) with uncertain effects on cancer. Given the association with birth weight seen here, it is likely that nutritional and growth-related attributes related to first-born birth status may influence synovial sarcoma incidence, but the other mechanisms listed above cannot be ruled out.

We did observe an association with parental age, particularly for paternal age, which appeared to be independent of maternal age. Paternal age is associated with epigenetic changes, DNA mutations, and aneuploidy, which are all characteristics of cancer cells and as such may contribute to risk should such features be contributed to the germline.

Our familial clustering analysis did not provide evidence for any increased risk of cancer among siblings of patients with pediatric synovial sarcoma, suggesting that strong cancer inherited predisposition alleles are not likely to play a major role in the incidence of this disease, at least in pediatric (0–19 years) synovial sarcoma examined here. In a prior study in the Utah Population Database, a 2-fold excess of BCL2-related malignancies (exclusively hematopoietic) were found in first- and second-degree relatives (25). This study presumably tracked adults which we did not consider here. In contrast to synovial sarcoma in our study, significant familial risks were demonstrated for pediatric osteosarcoma and rhabdomyosarcoma, which is expected given prior evidence for these cancers to be associated with cancer predisposition syndromes such as Li Fraumeni (26). We cannot exclude a significant role for low and intermediate penetrance susceptibility alleles for early-onset synovial sarcoma based on our data because our sample size was small.

This analysis, essentially a nested case–control study within the California birth cohort, has several strengths. First, cases were ascertained from the population-based CCR, and controls were randomly selected from the statewide birth records. None of the cases and controls had to be contacted for inclusion in this study, minimizing selection bias. Second, all data on birth characteristics were abstracted from preexisting birth records, reducing information bias in general and recall bias in particular. In addition, the diversity of the California population allowed us to include a large number of Hispanic subjects in this study. While the number of synovial sarcoma cases was relatively small due to the rarity of this disease, we included 50 times as many controls to help improve statistical power. To be included, cases must have been born and diagnosed in California, and controls only were required to have been born in the state. It is possible that some controls might have moved out of California, developed synovial sarcoma elsewhere, and therefore were not captured by CCR. However, we expect this to have very small, if any impact on our analysis, given the rarity of synovial sarcoma. In an extremely unlikely scenario where all of the 122,000 controls in our study moved out of California, we would only expect 1.98 cases, given SEER-based age-specific incidence rates of synovial sarcoma (10).

In summary, our findings regarding birth order and birth weight provide support that developmental and nutritional factors may contribute to the risk of synovial sarcoma. We found no evidence for strong familial genetic predisposition, apart from a disparity of risk by Hispanic status, which may be attributable to ethnic differences in genetic profiles or cultural differences related to diet or environmental exposures. Further studies into the etiology of synovial sarcoma will likely benefit from the evaluation of nutritional, environmental, and epigenetic factors, along with the mechanisms of formation of the archetypical t(X;18) translocations that define the disease.

E. Rubin reports receiving speakers bureau honoraria from Foundation Medicine. No potential conflicts of interest were disclosed by the other authors.

The ideas and opinions expressed herein are those of the author(s), and no endorsement by the California Department of Public Health, the NCI, or the Centers for Disease Control and Prevention or their contractors and subcontractors is intended or should be inferred.

Conception and design: J.L. Wiemels, X. Ma

Development of methodology: J.L. Wiemels

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): E. Rubin, L.M. Morimoto, C. Metayer, X. Ma

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): J.L. Wiemels, R. Wang, Q. Feng, J.F. Amatruda, A.C. Yee, C. Metayer, X. Ma

Writing, review, and/or revision of the manuscript: J.L. Wiemels, R. Wang, C.J. Clark, J.F. Amatruda, E. Rubin, A.C. Yee, L.M. Morimoto, C. Metayer, X. Ma

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J.L. Wiemels, A.C. Yee, L.M. Morimoto

Study supervision: J.L. Wiemels, X. Ma

We acknowledge the support of the NCI (R01CA175737, to J.L. Wiemels and X. Ma). The collection of cancer incidence data used in this study was supported by the California Department of Public Health as part of the statewide cancer reporting program mandated by California Health and Safety Code Section 103885; the NCI's Surveillance, Epidemiology, and End Results Program under contract HHSN261201000140C awarded to the Cancer Prevention Institute of California, contract HHSN261201000035C awarded to the University of Southern California, and contract HHSN261201000034C awarded to the Public Health Institute; and the Centers for Disease Control and Prevention's National Program of Cancer Registries, under agreement U58DP003862-01 awarded to the California Department of Public Health.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Kerouanton
A
,
Jimenez
I
,
Cellier
C
,
Laurence
V
,
Helfre
S
,
Pannier
S
, et al
Synovial sarcoma in children and adolescents
.
J Pediatr Hematol Oncol
2014
;
36
:
257
62
.
2.
Riggi
N
,
Cironi
L
,
Stamenkovic
I
. 
Synovial sarcoma: when epigenetic changes dictate tumour development
.
Swiss Med Wkly
2018
;
148
:
w14667
.
3.
Johnson
KJ
,
Carozza
SE
,
Chow
EJ
,
Fox
EE
,
Horel
S
,
McLaughlin
CC
, et al
Parental age and risk of childhood cancer: a pooled analysis
.
Epidemiology
2009
;
20
:
475
83
.
4.
Operskalski
EA
,
Preston-Martin
S
,
Henderson
BE
,
Visscher
BR
. 
A case-control study of osteosarcoma in young persons
.
Am J Epidemiol
1987
;
126
:
118
26
.
5.
Spector
LG
,
Puumala
SE
,
Carozza
SE
,
Chow
EJ
,
Fox
EE
,
Horel
S
, et al
Cancer risk among children with very low birth weights
.
Pediatrics
2009
;
124
:
96
104
.
6.
Hartley
AL
,
Birch
JM
,
McKinney
PA
,
Teare
MD
,
Blair
V
,
Carrette
J
, et al
The inter-regional epidemiological study of childhood cancer (IRESCC): case control study of children with bone and soft tissue sarcomas
.
Br J Cancer
1988
;
58
:
838
42
.
7.
Mirabello
L
,
Pfeiffer
R
,
Murphy
G
,
Daw
NC
,
Patino-Garcia
A
,
Troisi
RJ
, et al
Height at diagnosis and birth-weight as risk factors for osteosarcoma
.
Cancer Causes Control
2011
;
22
:
899
908
.
8.
Ognjanovic
S
,
Carozza
SE
,
Chow
EJ
,
Fox
EE
,
Horel
S
,
McLaughlin
CC
, et al
Birth characteristics and the risk of childhood rhabdomyosarcoma based on histological subtype
.
Br J Cancer
2010
;
102
:
227
31
.
9.
Von Behren
J
,
Spector
LG
,
Mueller
BA
,
Carozza
SE
,
Chow
EJ
,
Fox
EE
, et al
Birth order and risk of childhood cancer: a pooled analysis from five US States
.
Int J Cancer
2011
;
128
:
2709
16
.
10.
Surveillance, Epidemiology, End Results (SEER) Program
. 
SEER*Stat database: incidence - SEER 9 regs research data (1975–2016)
. 
2018
.
Available from:
https://seer.cancer.gov/data-software/documentation/seerstat/nov2018/.
11.
Overpeck
MD
,
Hediger
ML
,
Zhang
J
,
Trumble
AC
,
Klebanoff
MA
. 
Birth weight for gestational age of Mexican American infants born in the United States
.
Obstet Gynecol
1999
;
93
:
943
7
.
12.
Upchurch
S
,
Harris
JM
,
Cullinan
P
. 
Temporal changes in UK birth order and the prevalence of atopy
.
Allergy
2010
;
65
:
1039
41
.
13.
Dockerty
JD
,
Draper
G
,
Vincent
T
,
Rowan
SD
,
Bunch
KJ
. 
Case-control study of parental age, parity and socioeconomic level in relation to childhood cancers
.
Int J Epidemiol
2001
;
30
:
1428
37
.
14.
Hjalgrim
LL
,
Rostgaard
K
,
Hjalgrim
H
,
Westergaard
T
,
Thomassen
H
,
Forestier
E
, et al
Birth weight and risk for childhood leukemia in Denmark, Sweden, Norway, and Iceland
.
J Natl Cancer Inst
2004
;
96
:
1549
56
.
15.
Westergaard
T
,
Andersen
PK
,
Pedersen
JB
,
Olsen
JH
,
Frisch
M
,
Sorensen
HT
, et al
Birth characteristics, sibling patterns, and acute leukemia risk in childhood: a population-based cohort study
.
J Natl Cancer Inst
1997
;
89
:
939
47
.
16.
Cardwell
CR
,
Stene
LC
,
Joner
G
,
Bulsara
MK
,
Cinek
O
,
Rosenbauer
J
, et al
Birth order and childhood type 1 diabetes risk: a pooled analysis of 31 observational studies
.
Int J Epidemiol
2011
;
40
:
363
74
.
17.
Grulich
AE
,
Vajdic
CM
,
Kaldor
JM
,
Hughes
AM
,
Kricker
A
,
Fritschi
L
, et al
Birth order, atopy, and risk of non-Hodgkin lymphoma
.
J Natl Cancer Inst
2005
;
97
:
587
94
.
18.
Gluckman
PD
,
Hanson
MA
. 
Maternal constraint of fetal growth and its consequences
.
Semin Fetal Neonatal Med
2004
;
9
:
419
25
.
19.
Ayyavoo
A
,
Savage
T
,
Derraik
JG
,
Hofman
PL
,
Cutfield
WS
. 
First-born children have reduced insulin sensitivity and higher daytime blood pressure compared to later-born children
.
J Clin Endocrinol Metab
2013
;
98
:
1248
53
.
20.
Bernstein
L
,
Depue
RH
,
Ross
RK
,
Judd
HL
,
Pike
MC
,
Henderson
BE
. 
Higher maternal levels of free estradiol in first compared to second pregnancy: early gestational differences
.
J Natl Cancer Inst
1986
;
76
:
1035
9
.
21.
Panagiotopoulou
K
,
Katsouyanni
K
,
Petridou
E
,
Garas
Y
,
Tzonou
A
,
Trichopoulos
D
. 
Maternal age, parity, and pregnancy estrogens
.
Cancer Causes Control
1990
;
1
:
119
24
.
22.
Amirian
E
,
Scheurer
ME
,
Bondy
ML
. 
The association between birth order, sibship size and glioma development in adulthood
.
Int J Cancer
2010
;
126
:
2752
6
.
23.
Richiardi
L
,
Akre
O
,
Lambe
M
,
Granath
F
,
Montgomery
SM
,
Ekbom
A
. 
Birth order, sibship size, and risk for germ-cell testicular cancer
.
Epidemiology
2004
;
15
:
323
9
.
24.
Cockrill
T
,
del Junco
DJ
,
Arnett
FC
,
Assassi
S
,
Tan
FK
,
McNearney
T
, et al
Separate influences of birth order and gravidity/parity on the development of systemic sclerosis
.
Arthritis Care Res
2010
;
62
:
418
24
.
25.
Barrott
JJ
,
Zhu
JF
,
Smith-Fry
K
,
Susko
AM
,
Nollner
D
,
Burrell
LD
, et al
The influential role of BCL2 family members in synovial sarcomagenesis
.
Mol Cancer Res
2017
;
15
:
1733
40
.
26.
Schaefer
IM
,
Cote
GM
,
Hornick
JL
. 
Contemporary sarcoma diagnosis, genetics, and genomics
.
J Clin Oncol
2018
;
36
:
101
10
.