Advertisement

African ancestry is associated with aggressive endometrial cancer

Published:August 06, 2022DOI:https://doi.org/10.1016/j.ajog.2022.07.040

      Objective

      Significant racial disparities exist in endometrial cancer (EC). Relative to White women, Black women have a much higher incidence of aggressive histologic (serous)
      • Cote M.L.
      • Ruterbusch J.J.
      • Olson S.H.
      • Lu K.
      • Ali-Fehmi R.
      The growing burden of endometrial cancer: a major racial disparity affecting black women.
      and molecularly classified (copy number high [CNH]) tumors.
      • Dubil E.A.
      • Tian C.
      • Wang G.
      • et al.
      Racial disparities in molecular subtypes of endometrial cancer.
      Further, the incidence of endometrial cancer has increased, and Black women continue to have worse outcomes.
      • Lawrence W.R.
      • McGee-Avila J.K.
      • Vo J.B.
      • et al.
      Trends in cancer mortality among black individuals in the US From 1999 to 2019.
      • Giaquinto A.N.
      • Miller K.D.
      • Tossas K.Y.
      • Winn R.A.
      • Jemal A.
      • Siegel R.L.
      Cancer statistics for African American/Black people 2022.
      • Giaquinto A.N.
      • Broaddus R.R.
      • Jemal A.
      • Siegel R.L.
      The changing landscape of gynecologic cancer mortality in the United States.
      Variation in EC risk exists among geographic subpopulations of admixed Black women,
      • Pinheiro P.S.
      • Medina H.N.
      • Koru-Sengul T.
      • et al.
      Endometrial cancer type 2 incidence and survival disparities within subsets of the US Black population.
      ,
      • Schlumbrecht M.
      • Baeker Bispo J.A.
      • Balise R.R.
      • Huang M.
      • Slomovitz B.
      • Kobetz E.
      Variation in type II endometrial cancer risk by Hispanic subpopulation: an exploratory analysis.
      suggesting that ancestry may drive disease pathogenesis. Our aim was to determine the extent to which self-identified race and genetic ancestry are associated with the probability of developing high-grade tumors among women with EC, as determined by both histologic and molecular classifications.

      Study Design

      This study analyzed secondary data and was therefore institutional review board-exempt. We reviewed tumor genomic data from publicly available next-generation sequencing in The Cancer Genome Atlas (TCGA)
      • Kandoth C.
      • Schultz N.
      • et al.
      Cancer Genome Atlas Research Network
      Integrated genomic characterization of endometrial carcinoma.
      and The Cancer Genetic Ancestry Atlas (TCGAA)
      • Yuan J.
      • Hu Z.
      • Mahal B.A.
      • et al.
      Integrated analysis of genetic ancestry and genomic alterations across cancers.
      under the project Uterine Corpus Endometrial Carcinoma. We merged both datasets by matching the patient ID and extracted the age of diagnosis, histology, molecular classification, self-identified race, EIGENSTRAT-assigned genomic ancestry, and STRUCTURE-assigned genomic-based ancestry percentage (Supplement S1). Demographic data were described with mean and standard deviations for continuous variables and counts and percentages for categorical variables. An independent t test or analysis of variance (ANOVA) was used compare continuous variables, and chi-square test was used to compare categorical variables. Controlling for age, we used multivariate logistic regression to determine the odds of diagnosis with high-grade cancers.

      Results

      We identified 568 women, including 115 (20%) self-identified Black, 417 (73%) White, 23 (4%) Asian, and 13 (3%) others (Table). Self-identified race and assigned genomic-based ancestry did not match in 24 (4.2%) patients. Percent African ancestry varied among women who were genomically identified as Black (Supplement S2). Patients with serous histology had a higher mean percentage of African ancestry (26.57%) than those with low-grade (LG) endometrioid histology (15.57%), high-grade (HG) endometrioid histology (14.62%), and carcinosarcoma (12.96%) (P=.0078) (Supplement S3). Percent African ancestry was higher in women with CNH tumors (27.28%) relative to other genomic classifications (Supplement S4). Relative to self-identified White women, Black women had a higher probability of serous histology (odds ratio [OR], 2.163; confidence interval [CI], 1.317–3.552; P=.0023) and CNH tumors (OR, 2.846; CI, 1.759–4.604; P<.0001) when compared with all other histologies or subtypes combined (Figure). Relative to women genomically characterized as White, those characterized as Black had a greater odds of serous (OR, 2.310; CI, 1.414–3.773; P=.0008) and CNH (OR, 2.839; CI, 1.760–4.579; P<.0001) tumors (Figure) with similar comparator groups.
      TableDemographics and clinical and genomic characteristics overall in 568 participants in The Cancer Genome Atlas research network endometrial carcinoma cohort


      Variable
      Self-identified race
      White

      N=417 (%)
      Black or African American

      N=115 (%)
      Asian/other

      N=36 (%)
      Age at diagnosis (mean+standard deviation)65.02 (10.78)65.41 (9.82)56.81 (12.91)
      EIGENSTRAT-assigned genomic-based ancestry
      African American5 (1.20)113 (98.26)1 (2.78)
      European American403 (96.64)2 (1.74)1 (2.78)
      East Asian American1 (0.24)0 (0)29 (80.56)
      Native American8 (1.92)0 (0)3 (8.33)
      Other American0 (0)0 (0)2 (5.56)
      STRUCTURE-assigned genomic-based ancestry
      Median (min–max)
      % African ancestry0.37 (0.12–90.98)78.65 (0.31–94.4)1.74 (0.15–34.88)
      % European ancestry98.13 (6.86–99.41)19.11 (1.55–98.35)6.83 (0.17–91.32)
      % Asian ancestry0.73 (0.19–89.28)0.92 (0.23–17.94)90.93 (0.42–99.28)
      % Native American Ancestry0.56 (0.08–50.63)0.49 (0.15–2.94)0.50 (0.25–45.93)
      Grade
      Grade 1+2163 (39.09)39 (33.91)15 (41.67)
      Grade 3254 (60.91)76 (66.09)21 (58.33)
      Stage
      Stage I+II296 (70.98)72 (62.61)26 (72.22)
      Stage III+IV121 (29.02)43 (37.39)10 (27.78)
      Histology
      Low-grade endometrioid161 (38.61)39 (33.91)15 (41.67)
      High-grade endometrioid130 (31.18)30 (26.09)13 (36.11)
      Serous64 (15.35)32 (27.83)5 (13.89)
      Uterine carcinosarcoma44 (10.55)8 (6.96)3 (8.33)
      Mixed epithelial carcinoma18 (4.32)5 (4.35)0 (0)
      Clear cell0 (0)1 (0.87)0 (0)
      Genomic classification (n=480)
      POLE26 (7.49)8 (8)7 (21.21)
      Microsatellite instability108 (31.12)23 (23)10 (30.30)
      Copy number low116 (33.43)18 (18)10 (30.30)
      Copy number high97 (27.95)51 (51)6 (18.18)
      Sanchez-Covarrubias. African ancestry is associated with aggressive endometrial cancer. Am J Obstet Gynecol 2022.
      Figure thumbnail gr1
      FigureAssociations between age, self-reported race, genomically-assigned race, percent African ancestry, and serous (A) and CNH high (B) EC
      CI, confidence interval; EC, endometrial cancer; CNH, copy number high.
      Sanchez-Covarrubias. African ancestry is associated with aggressive endometrial cancer. Am J Obstet Gynecol 2022.
      For every 1% increase in African ancestry, there was a 1.1% increase in the probability of diagnosis with serous EC (CI, 1.004–1.017; P=.0010) (Figure, A) and a 1.4% increase in the likelihood of diagnosis with a CNH tumor (CI, 1.008–1.020; P<.0001) (Figure, B). Pairwise histotype analyses confirmed differences in serous EC when compared with LG endometrioid, HG endometrioid, and carcinosarcoma (Supplement S5). Pairwise analysis using molecular classification was significant when comparing CNH vs microsatellite instability (MSI) and CNH vs copy number low (CNL); CNH vs POLE comparison was not significant (Supplement S6). Other nonsignificant comparisons by stage, histology, and molecular classification are included in Supplement S7.

      Conclusion

      Increasing African ancestry is associated with a concurrent increase in the odds of both serous and CNH tumors relative to all other histologies and molecular subtypes of EC, respectively. There was variation in the proportions of women who self-identified as Black vs those genomically-assigned Black in this population (4.2%), more than the <1% variation reported elsewhere
      • Fang H.
      • Hui Q.
      • Lynch J.
      • et al.
      Harmonizing genetic ancestry and self-identified race/ethnicity in genome-wide association studies.
      ; this highlights the subjective assignment of race. Race as a social construct groups large numbers of women together into a single monolith without recognizing the diversity in genetic phenotypes, cultural norms, behaviors, history, and structural racism that may contribute to disease.
      • Fuller K.E.
      Health disparities: reframing the problem.
      How to best use race and/or ancestry in the context of cancer risk and outcomes remains a topic of debate. Some authors have suggested that genomically-assigned race may be a better predictor for outcomes in cancer,
      • Burchard E.G.
      • Ziv E.
      • Coyle N.
      • et al.
      The importance of race and ethnic background in biomedical research and clinical practice.
      ,
      • Agyemang C.
      • Bhopal R.
      • Bruijnzeels M.
      Negro, Black, Black African, African Caribbean, African American or what? Labelling African origin populations in the health arena in the 21st century.
      whereas others suggest that an integrated approach recognizing both race and ancestry is more appropriate.
      • Fang H.
      • Hui Q.
      • Lynch J.
      • et al.
      Harmonizing genetic ancestry and self-identified race/ethnicity in genome-wide association studies.
      In this investigation, serous and CNH were the only subtypes to have a strong association with genetic ancestry; this was not demonstrated with other aggressive subtypes such as carcinosarcoma, which is known to be strongly associated with Black race.
      • Pinheiro P.S.
      • Medina H.N.
      • Koru-Sengul T.
      • et al.
      Endometrial cancer type 2 incidence and survival disparities within subsets of the US Black population.
      This suggests that for the serous or CNH subtypes specifically, a biological driver of disease pathogenesis in addition to social factors may be contributing to the disproportionate incidence of these tumors in women of African ancestry.
      Our study is limited by the lack of patient-level data, as these are not reported by the TCGA, and our analyses are adjusted by age alone. This restricts multiple statistical comparisons and inclusion of factors such a zip code from which inferences about social determinants of health including redlining, food insecurity, and immigrant status could be made. It is likely that there are multiple factors that contribute to disease risk in Black women, but we show here that an underlying biological or genetic cause for high-grade EC exists. Understanding the elements driving cancer pathogenesis, how they may interact with ancestral genetics and the resulting cellular biology, and how the experience of race further contributes to disease etiology is required to develop strategies for risk reduction, earlier detection, and treatment optimization in Black women.

      Supplement S1 Methods

      Data sources

      We reviewed tumor genomic data obtained from next generation sequencing and publicly available in The Cancer Genome Atlas (TCGA) under the project Uterine Corpus Endometrial Carcinoma
      • Kandoth C.
      • Schultz N.
      • et al.
      Cancer Genome Atlas Research Network
      Integrated genomic characterization of endometrial carcinoma.
      . Variables extracted from this project included Tumor ID, age at diagnosis, self-identified race, stage at diagnosis, histology, neoplasm histologic grade and genomic classification. In addition, we used The Cancer Genetic Ancestry Atlas (TCGAA) under the study Uterine Corpus Endometrial Carcinoma
      • Yuan J.
      • Hu Z.
      • Mahal B.A.
      • et al.
      Integrated analysis of genetic ancestry and genomic alterations across cancers.
      to extract the variables EIGENSTRAT assigned genomic – based ancestry, and STRUCTURE assigned genomic – based ancestry. Both datasets were merged matching the Patient ID. EIGENSTRAT
      • Price A.L.
      • Patterson N.J.
      • Plenge R.M.
      • Weinblatt M.E.
      • Shadick N.A.
      • Reich D.
      Principal components analysis corrects for stratification in genome-wide association studies.
      is a method that models ancestry differences between cases and controls while detecting and correcting for population stratification, in the TCGAA it was used to genomically assigned one of four races (African American, European American, Asian American and Native American) to the ovarian cancer dataset. STRUCTURE
      • Pritchard J.K.
      • Stephens M.
      • Donnelly P.
      Inference of population structure using multilocus genotype data.
      ,
      • Porras-Hurtado L.
      • Ruiz Y.
      • Santos C.
      • Phillips C.
      • Carracedo A.
      • Lareu M.V.
      An overview of STRUCTURE: applications, parameter settings, and supporting software.
      is an algorithm that estimates variant frequencies and assign samples to groups whose members share similar patterns of genetic variation. Subsequently, STRUCTURE can determine the percentage of contribution for each ancestral population. In the TCGAA, STRUCTURE was used in K=4 ancestral populations (African, European, Asian, Native American). The merged dataset is available to download (Supplement S8)

      Statistical analysis

      we used SAS software 9.4 (SAS Institute Inc., Cary, NC) and R 4.1.0 to perform the statistical analysis. Demographic data were described with mean and standard deviations for continuous variables and counts and percentages for categorical variables. An independent t-test or one – way ANOVA was used to determine differences in means for (normally distributed) continuous variables with two or more categories, when appropriate. Controlling for age at diagnosis, we performed logistic regression comparing self-identified race, EIGENSTRAT assigned genomic – based ancestry, and STRUCTURE assigned genomic – based ancestry using the histologic type or molecular classification. We also included models comparing these variables in Stage I II vs Stage III IV tumors, and for each histologic type or molecular classification in a pair way fashion (pairwise comparison). The histologic type of mixed epithelial carcinoma and clear cell carcinoma were excluded from pair comparison because of low numbers of patients with these diseases which were felt to prohibit meaningful interpretation of data. All p values reported were 2-sided, and p<0.05 was considered statistically significant. The R package ggplot2 was used to create figures representing the ancestral percentages for each group.
      Figure thumbnail fx1
      Supplement S2Distribution of ancestry (in percent) among women genomically classified as Black
      Sanchez-Covarrubias. African ancestry is associated with aggressive endometrial cancer. Am J Obstet Gynecol 2022.
      Supplement S3Cohort characteristics by histology
      VariableLow grade Endometrioid

      N=215
      High grade endometrioid

      N=173
      Serous

      N=101
      Carcinosarcoma

      N= 55
      p value
      Age (Mean, SD)63.23 (11.52)62.07 (10.95)68.65 (8.44)70.09 (9.35)<0.0001
      Self-identified Race0.0791
      Black or African American39 (18.14)30 (17.34)32 (31.68)8 (14.55)
      White161 (74.88)130 (75.14)64 (63.37)44 (80)
      Asian/Other14 (6.98)13 (7.51)5 (4.95)3 (5.45)
      EIGENSTRAT – Assigned genomic – based ancestry0.0323
      African American40 (18.60)31 (17.92)34 (33.66)8 (14.55)
      European American158 (73.49)126 (72.83)62 (61.39)43 (78.18)
      EAA/NA/OA17 (7.91)16 (9.25)5 (4.95)4 (7.27)
      STRUCTURE –Assigned genomic – based ancestry
      %African Ancestry15.57 (30.60)14.62 (29.40)26.57 (36.45)12.96 (26.69)0.0078
      %European Ancestry76.76 (35.38)76.51 (35.12)68.36 (37.93)79.56 (33.13)0.1596
      %Native American Ancestry1.24 (4.35)1.14 (3.44)1.42 (5.94)0.78 (1.43)0.8350
      %Asian Ancestry6.43 (20.87)7.72 (22.793.65 (14.04)6.7 (22.33)0.4732
      Stage<0.0001
      Stage I+II184 (85.58)125 (72.25)44 (43.56)26 (47.27)
      Stage III+IV31 (14.42)48 (27.75)57 (56.44)29 (53.73)
      Sanchez-Covarrubias. African ancestry is associated with aggressive endometrial cancer. Am J Obstet Gynecol 2022.
      Supplement S4Cohort characteristics by molecular classification
      VariablePOLE

      N= 41 (8.5%)
      MSI

      N=141 (29.4%)
      CN+ Low

      N= 144 (30%)
      CN+ High

      N= 154 (32.1%)
      p value
      Age (Mean, SD)57.32 (11.55)62.99 (9.90)61.68 (11.76)68.53 (8.95)<0.0001
      Self-identified Race<0.0001
      Black or African American8 (19.5)23 (16.3)18 (12.5)51 (33.1)
      White26 (63.4)108 (76.6)116 (80.6)97 (63)
      Asian/Other7 (17.1)10 (7.1)10 (6.9)6 (3.9)
      EIGENSTRAT – Assigned Genomic – based ancestry<0.0001
      African American8 (19.51)25 (17.73)19 (13.19)52 (33.77)
      European American25 (60.98)106 (75.18)112 (77.78)94 (61.04)
      EAA/NA/OA8 (19.51)10 (7.09)13 (9.03)8 (5.19)
      STRUCTURE – Assigned Genomic – based ancestry
      %African Ancestry16.79 (31.49)14.53 (29.33)10.96 (25.98)27.28 (37.15)<0.0001
      %European Ancestry65.46 (40.43)77.35 (34.92)81.25 (32.32)67.90 (38.13)0.0033
      %NA Ancestry1.38 (4.95)0.91 (2.50)1.55 (5.27)1.53 (6.01)0.6652
      %Asian Ancestry16.39 (33.31)7.20 (22.14)6.24 (20.57)3.28 (12.49)0.0036
      Stage<0.0001
      Stage I+II29 (70.7)113 (80.1)120 (83.3)80 (52)
      Stage III+IV12 (29.3)28 (19.9)24 (16.7)74 (48)
      CN+, + copy number; MSI, microsatellite instability
      Sanchez-Covarrubias. African ancestry is associated with aggressive endometrial cancer. Am J Obstet Gynecol 2022.
      Figure thumbnail fx2
      Supplement S5Pairwise comparison showing associations between age, self-reported race, genomically assigned race, percent African ancestry and serous endometrial cancer (EC) vs low-grade (LG) endometrioid (A), high-grade (HG) endometrioid (B), and carcinosarcoma (C)
      Sanchez-Covarrubias. African ancestry is associated with aggressive endometrial cancer. Am J Obstet Gynecol 2022.
      Figure thumbnail fx3
      Supplement S6Pairwise comparison showing associations between age, self-reported race, genomically assigned race, percent African ancestry and copy number high (CNH) vs copy number low (CNL) (A), microsatellite instability (MSI) (B) and POLE (C) subtypes
      Sanchez-Covarrubias. African ancestry is associated with aggressive endometrial cancer. Am J Obstet Gynecol 2022.
      Figure thumbnail fx4ad
      Supplement S7Pairwise comparison showing associations between age, self-reported race, genomically assigned ancestry, percent African ancestry and Stage I II vs Stage III IV endometrial cancer (EC)(A); low-grade (LG) endometrioid EC vs all others (B); high-grade (HG) endometrioid EC vs all others (C); carcinosarcoma vs all others (D); LG endometrioid vs HG endometrioid EC (E); LG endometrioid vs carcinosarcoma EC (F); HG endometrioid EC vs carcinosarcoma (G); POLE EC (H); microsatellite instability (MSI) EC (I); copy number (CNL) EC (J); POLE vs MSI EC (K); POLE vs CNL EC (L) and MSI vs CNL EC (M)
      Sanchez-Covarrubias. African ancestry is associated with aggressive endometrial cancer. Am J Obstet Gynecol 2022.
      Figure thumbnail fx4eg
      Supplement S7Pairwise comparison showing associations between age, self-reported race, genomically assigned ancestry, percent African ancestry and Stage I II vs Stage III IV endometrial cancer (EC)(A); low-grade (LG) endometrioid EC vs all others (B); high-grade (HG) endometrioid EC vs all others (C); carcinosarcoma vs all others (D); LG endometrioid vs HG endometrioid EC (E); LG endometrioid vs carcinosarcoma EC (F); HG endometrioid EC vs carcinosarcoma (G); POLE EC (H); microsatellite instability (MSI) EC (I); copy number (CNL) EC (J); POLE vs MSI EC (K); POLE vs CNL EC (L) and MSI vs CNL EC (M)
      Sanchez-Covarrubias. African ancestry is associated with aggressive endometrial cancer. Am J Obstet Gynecol 2022.
      Figure thumbnail fx4hj
      Supplement S7Pairwise comparison showing associations between age, self-reported race, genomically assigned ancestry, percent African ancestry and Stage I II vs Stage III IV endometrial cancer (EC)(A); low-grade (LG) endometrioid EC vs all others (B); high-grade (HG) endometrioid EC vs all others (C); carcinosarcoma vs all others (D); LG endometrioid vs HG endometrioid EC (E); LG endometrioid vs carcinosarcoma EC (F); HG endometrioid EC vs carcinosarcoma (G); POLE EC (H); microsatellite instability (MSI) EC (I); copy number (CNL) EC (J); POLE vs MSI EC (K); POLE vs CNL EC (L) and MSI vs CNL EC (M)
      Sanchez-Covarrubias. African ancestry is associated with aggressive endometrial cancer. Am J Obstet Gynecol 2022.
      Figure thumbnail fx4km
      Supplement S7Pairwise comparison showing associations between age, self-reported race, genomically assigned ancestry, percent African ancestry and Stage I II vs Stage III IV endometrial cancer (EC)(A); low-grade (LG) endometrioid EC vs all others (B); high-grade (HG) endometrioid EC vs all others (C); carcinosarcoma vs all others (D); LG endometrioid vs HG endometrioid EC (E); LG endometrioid vs carcinosarcoma EC (F); HG endometrioid EC vs carcinosarcoma (G); POLE EC (H); microsatellite instability (MSI) EC (I); copy number (CNL) EC (J); POLE vs MSI EC (K); POLE vs CNL EC (L) and MSI vs CNL EC (M)
      Sanchez-Covarrubias. African ancestry is associated with aggressive endometrial cancer. Am J Obstet Gynecol 2022.

      Supplementary Data

      References

        • Cote M.L.
        • Ruterbusch J.J.
        • Olson S.H.
        • Lu K.
        • Ali-Fehmi R.
        The growing burden of endometrial cancer: a major racial disparity affecting black women.
        Cancer Epidemiol Biomarkers Prev. 2015; 24: 1407-1415
        • Dubil E.A.
        • Tian C.
        • Wang G.
        • et al.
        Racial disparities in molecular subtypes of endometrial cancer.
        Gynecol Oncol. 2018; 149: 106-116
        • Lawrence W.R.
        • McGee-Avila J.K.
        • Vo J.B.
        • et al.
        Trends in cancer mortality among black individuals in the US From 1999 to 2019.
        JAMA Oncol. 2022; ([Epub ahead of print])
        • Giaquinto A.N.
        • Miller K.D.
        • Tossas K.Y.
        • Winn R.A.
        • Jemal A.
        • Siegel R.L.
        Cancer statistics for African American/Black people 2022.
        CA Cancer J Clin. 2022; 72: 202-229
        • Giaquinto A.N.
        • Broaddus R.R.
        • Jemal A.
        • Siegel R.L.
        The changing landscape of gynecologic cancer mortality in the United States.
        Obstet Gynecol. 2022; 139: 440-442
        • Pinheiro P.S.
        • Medina H.N.
        • Koru-Sengul T.
        • et al.
        Endometrial cancer type 2 incidence and survival disparities within subsets of the US Black population.
        Front Oncol. 2021; 11699577
        • Schlumbrecht M.
        • Baeker Bispo J.A.
        • Balise R.R.
        • Huang M.
        • Slomovitz B.
        • Kobetz E.
        Variation in type II endometrial cancer risk by Hispanic subpopulation: an exploratory analysis.
        Gynecol Oncol. 2017; 147: 329-333
        • Kandoth C.
        • Schultz N.
        • et al.
        • Cancer Genome Atlas Research Network
        Integrated genomic characterization of endometrial carcinoma.
        Nature. 2013; 497: 67-73
        • Yuan J.
        • Hu Z.
        • Mahal B.A.
        • et al.
        Integrated analysis of genetic ancestry and genomic alterations across cancers.
        Cancer Cell. 2018; 34 (e9): 549-560
        • Fang H.
        • Hui Q.
        • Lynch J.
        • et al.
        Harmonizing genetic ancestry and self-identified race/ethnicity in genome-wide association studies.
        Am J Hum Genet. 2019; 105: 763-772
        • Fuller K.E.
        Health disparities: reframing the problem.
        Med Sci Monit. 2003; 9: SR9-SR15
        • Burchard E.G.
        • Ziv E.
        • Coyle N.
        • et al.
        The importance of race and ethnic background in biomedical research and clinical practice.
        N Engl J Med. 2003; 348: 1170-1175
        • Agyemang C.
        • Bhopal R.
        • Bruijnzeels M.
        Negro, Black, Black African, African Caribbean, African American or what? Labelling African origin populations in the health arena in the 21st century.
        J Epidemiol Community Health. 2005; 59: 1014-1018

      References

        • Kandoth C.
        • Schultz N.
        • et al.
        • Cancer Genome Atlas Research Network
        Integrated genomic characterization of endometrial carcinoma.
        Nature. 2013; 497: 67-73
        • Yuan J.
        • Hu Z.
        • Mahal B.A.
        • et al.
        Integrated analysis of genetic ancestry and genomic alterations across cancers.
        Cancer Cell. 2018; 34 (e9): 549-560
        • Price A.L.
        • Patterson N.J.
        • Plenge R.M.
        • Weinblatt M.E.
        • Shadick N.A.
        • Reich D.
        Principal components analysis corrects for stratification in genome-wide association studies.
        Nat Genet. 2006; 38: 904-909
        • Pritchard J.K.
        • Stephens M.
        • Donnelly P.
        Inference of population structure using multilocus genotype data.
        Genetics. 2000; 155: 945-959
        • Porras-Hurtado L.
        • Ruiz Y.
        • Santos C.
        • Phillips C.
        • Carracedo A.
        • Lareu M.V.
        An overview of STRUCTURE: applications, parameter settings, and supporting software.
        Front Genet. 2013; 4: 98