{"paper_id":"d55626e4-2b12-40ee-bd41-f3c5c82ffb73","body_text":"UC Irvine\nUC Irvine Previously Published Works\nTitle\nA framework for assessing interactions for risk stratification models: the example of \novarian cancer\nPermalink\nhttps://escholarship.org/uc/item/8f56n701\nJournal\nJournal of the National Cancer Institute, 115(11)\nISSN\n0027-8874\nAuthors\nPhung, Minh Tung\nLee, Alice W\nMcLean, Karen\net al.\nPublication Date\n2023-11-08\nDOI\n10.1093/jnci/djad137\n \nPeer reviewed\neScholarship.org Powered by the California Digital Library\nUniversity of California\n\nA framework for assessing interactions for risk\nstratification models: the example of ovarian cancer\nMinh Tung Phung , PhD, MPH, 1,* Alice W. Lee, PhD, MPH, 2 Karen McLean, MD, PhD, 3 Hoda Anton-Culver , PhD,4\nElisa V. Bandera, MD, PhD, 5 Michael E. Carney, MD, 6 Jenny Chang-Claude, PhD, 7,8 Daniel W. Cramer, MD, ScD, 9,10\nJennifer Anne Doherty, MS, PhD, 11 Renee T. Fortner, PhD, 7,12 Marc T. Goodman, PhD, 13,14 Holly R. Harris, ScD, MPH, 15,16\nAllan Jensen , PhD,17 Francesmary Modugno, PhD, MPH, 18,19,20 Kirsten B. Moysich, MS, PhD, 21 Paul D. P. Pharoah, PhD, 22\nBo Qin, PhD, 5 Kathryn L. Terry, ScD, 9,10 Linda J. Titus, PhD, 23 Penelope M. Webb, PhD, 24; on behalf of the Australian Ovarian Cancer\nStudy Group, Anna H. Wu, PhD, 25 Nur Zeinomar , PhD,5 Argyrios Ziogas , PhD,4 Andrew Berchuck, MD, 26 Kathleen R. Cho, MD, 27\nGillian E. Hanley, PhD, 28 Rafael Meza, PhD, 1,29 Bhramar Mukherjee , PhD,1,30 Malcolm C. Pike, PhD, 25,31\nCeleste Leigh Pearce, PhD, MPH, 1,‡ Britton Trabert, PhD, 32,33,‡; on behalf of the Ovarian Cancer Association Consortium\n1Department of Epidemiology, University of Michigan School of Public Health, Ann Arbor, MI, USA\n2Department of Public Health, California State University, Fullerton, Fullerton, CA, USA\n3Department of Gynecologic Oncology and Department of Pharmacology & Therapeutics, Elm & Carlton Streets, Roswell Park Comprehensive Cancer Center,\nBuffalo, NY, USA\n4Department of Medicine, University of California, Irvine, Irvine, CA, USA\n5Cancer Epidemiology and Health Outcomes, Rutgers Cancer Institute of New Jersey, New Brunswick, NJ, USA\n6Department of Obstetrics and Gynecology, John A. Burns School of Medicine, University of Hawaii, Honolulu, HI, USA\n7Division of Cancer Epidemiology, German Cancer Research Center (DKFZ), Heidelberg, Germany\n8Cancer Epidemiology Group, University Cancer Center Hamburg, University Medical Center Hamburg-Eppendorf, Hamburg, Germany\n9Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA, USA\n10Department of Obstetrics and Gynecology, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA, USA\n11Huntsman Cancer Institute, Department of Population Health Sciences, University of Utah, Salt Lake City, UT, USA\n12Department of Research, Cancer Registry of Norway, Oslo, Norway\n13Samuel Oschin Comprehensive Cancer Institute, Cancer Prevention and Genetics Program, Cedars-Sinai Medical Center, Los Angeles, CA, USA\n14Community and Population Health Research Institute, Department of Biomedical Sciences, Cedars-Sinai Medical Center, Los Angeles, CA, USA\n15Program in Epidemiology, Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA, USA\n16Department of Epidemiology, University of Washington School of Public Health, Seattle, WA, USA\n17Department of Lifestyle, Reproduction and Cancer, Danish Cancer Society Research Center, Copenhagen, Denmark\n18Women’s Cancer Research Center, Magee-Women’s Research Institute and Hillman Cancer Center, Pittsburgh, PA, USA\n19Division of Gynecologic Oncology, Department of Obstetrics, Gynecology and Reproductive Sciences, University of Pittsburgh School of Medicine,\nPittsburgh, PA, USA\n20Department of Epidemiology, University of Pittsburgh Graduate School of Public Health, Pittsburg, PA, USA\n21Division of Cancer Prevention and Control, Roswell Park Comprehensive Cancer Center, Buffalo, NY, USA\n22Department of Computational Biomedicine, Cedars-Sinai Medical Centre, Los Angeles, CA, USA\n23Public Health, Muskie School of Public Service, University of Southern Maine, Portland, ME, USA\n24Population Health Program, QIMR Berghofer Medical Research Institute, Brisbane, QLD, Australia\n25Department of Population and Public Health Sciences, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA\n26Division of Gynecologic Oncology, Duke University School of Medicine, Durham, NC, USA\n27Department of Pathology, University of Michigan Medical School, Ann Arbor, MI, USA\n28Department of Obstetrics & Gynecology, University of British Columbia Faculty of Medicine, Vancouver, BC, Canada\n29Department of Integrative Oncology, BC Cancer Research Institute, Vancouver, BC, Canada\n30Department of Biostatistics, University of Michigan School of Public Health, Ann Arbor, MI, USA\n31Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY, USA\n32Department of Obstetrics and Gynecology, University of Utah, Salt Lake City, UT, USA\n33Cancer Control and Populations Sciences Program, Huntsman Cancer Institute at the University of Utah, Salt Lake City, UT, USA\n*Correspondence to: Minh Tung Phung, PhD, MPH, Department of Epidemiology, University of Michigan School of Public Health, 4642 SPH Tower, 1415 Washington\nHeights, Ann Arbor, MI, USA 48109-2029 (e-mail: phungmt@umich.edu).\n‡These authors contributed equally to this work.\nAbstract\nGenerally, risk stratification models for cancer use effect estimates from risk/protective factor analyses that have not assessed potential\ninteractions between these exposures. We have developed a 4-criterion framework for assessing interactions that includes statistical,\nqualitative, biological, and practical approaches. We present the application of this framework in an ovarian cancer setting because this is\nan important step in developing more accurate risk stratification models. Using data from 9 case-control studies in the Ovarian Cancer\nAssociation Consortium, we conducted a comprehensive analysis of interactions among 15 unequivocal risk and protective factors for\novarian cancer (including 14 non-genetic factors and a 36-variant polygenic score) with age and menopausal status. Pairwise interactions\nReceived: March 23, 2023. Revised: June 08, 2023. Accepted: June 30, 2023\nVC The Author(s) 2023. Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com\nJNCI: Journal of the National Cancer Institute , 2023, 115(11), 1420–1426\nhttps://doi.org/10.1093/jnci/djad137\nAdvance Access Publication Date: July 12, 2023\nBrief Communications\n\nbetween the risk/protective factors were also assessed. We found that menopausal status modifies the association among endometriosis,\nfirst-degree family history of ovarian cancer, breastfeeding, and depot-medroxyprogesterone acetate use and disease risk, highlighting\nthe importance of understanding multiplicative interactions when developing risk prediction models.\nThe development of risk stratification approaches to identify\nindividuals who would most benefit from primary prevention\nstrategies has become increasingly important. Risk stratification\nmodels use the effect estimates for the risk/protective factors\nconsidered to be unequivocal in their association with the disease\nunder study. Generally, the effect estimates come from analyses\nin which multiplicative relationships were assumed among risk\nand protective factors. Using invasive epithelial ovarian cancer\n(ovarian cancer), we offer a strategy for the initial steps needed to\ndevelop accurate risk stratification models, including a 4-crite-\nrion framework for assessing whether potential interactions\nshould be included. Interaction analyses are notoriously under-\npowered, so using this framework ensures that important differ-\nences that may indicate departures from multiplicativity are not\nmissed.\n• Criterion A (statistical approach): A likelihood ratio test com-\nparing a logistic model with the interaction term vs the same\nmodel without the interaction term (a 2-sided P < .05 for inter-\naction was considered statistically significant was used here,\nbut other statistical approaches could be used);\n• Criterion B (qualitative approach): Comparing the consis-\ntency and magnitude of the odds ratios (ORs) of a factor\nacross the levels of the other factor (visualization from strati-\nfied analysis);\n• Criterion C (biological approach): Considering biological plau-\nsibility; and\n• Criterion D (practical approach): Assessing the prevalence of\nthe risk/protective factors to determine whether an interaction\nwould have a meaningful impact on the risk stratification\nmodel.\nOvarian cancer is an ideal example for refining risk stratification\napproaches because primary prevention strategies are available for\nw o m e na tb o t ha v e r a g ea n dh i g hr i s k ,i n c l u d i n gr i s k - r e d u c i n gs a l -\npingo-oophorectomy, opportunistic salpingectomy, tubal ligation,\nand possibly hormonal contraceptives ( 1-4). Unequivocal ovarian\ncancer risk and protective factors include 14 non-genetic factors (4-\n14) and a 36-variant polygenic score for ovarian cancer (15)( 1 5f a c -\ntors are shown in Supplementary Table 1 , available online).\nImportantly for ovarian and many other cancers affecting women,\nthe effects of age and menopausal status on the risk/protective fac-\ntors must first be disentangled to determine whether one, both, or\nneither modifies the associations (16).\nWe applied the framework to questionnaire data from 9\nOvarian Cancer Association Consortium case-control studies\nfrom Australia ( 17), Germany (18), and the United States ( 19-25).\nInstitutional review board approval was obtained by the original\nstudies, and all participants had provided written informed con-\nsent. To determine whether there was an age interaction, a men-\nopausal status interaction, or both, the initial ovarian cancer and\nrisk and protective factor analyses were conducted among partic-\nipants in the following strata (Table 1):\n• Stratum 1: Younger than 45 years of age and premenopausal\n• Stratum 2: Aged 45 to 54 years and premenopausal\n• Stratum 3: Aged 45 to 54 years and postmenopausal\n• Stratum 4: Aged 55 to 64 years and postmenopausal\n• Stratum 5: Aged 65 to 84 years and postmenopausal\nWe found differences in the associations between the risk/pro-\ntective factors for ovarian cancer by menopausal status but not by\nage (particularly informed by comparing results between strata 2\nand 3; Supplementary Table 2, A-D, available online) based on the\n4-criterion interaction evaluation framework described earlier.\nMenopausal status appeared to modify the associations\nbetween ovarian cancer risk and endometriosis, first-degree fam-\nily history of ovarian cancer, breastfeeding, and depot-medroxy-\nprogesterone acetate use ( Table 2). For example, a self-reported\nhistory of endometriosis was associated with a greater increase\nin risk of ovarian cancer among premenopausal women than\namong postmenopausal participants ( P ¼ .04 for interaction; cri-\nterion A). Moreover, although no standardized definitions exist\non how different the 2 stratum-specific associations should be\nfor a factor to be an effect modifier, it is widely accepted that an\nOR less than 1.5 is considered a small effect size, while an\nOR between 1.5-2.0 is considered medium ( 26). Thus, the magni-\ntude of the difference in the endometriosis association between\npremenopausal (OR ¼ 1.94) and postmenopausal (OR ¼ 1.33)\nwomen is qualitatively meaningful (criterion B). Further, the\nendometriosis-menopausal status interaction is biologically\nplausible (criterion C) because during the premenopausal period,\nendometriosis is active (ovulatory proinflammatory and prolifer-\native processes) ( 27-30), whereas endometriosis is generally qui-\nescent in the postmenopausal period ( 31). Finally, endometriosis\nis estimated to have a prevalence of up to 10% in the general pop-\nulation (32); thus, it is sufficiently common to warrant fitting sep-\narate risk stratification models for pre- and postmenopausal\nwomen to be able to incorporate different effect estimates for\nendometriosis (criterion D).\nGiven that 4 risk and protective factors suggest an interaction\nwith menopausal status based on our framework, including one\nthat met all 4 criteria, we further evaluated pairwise interactions\nbetween the risk and protective factors separately for pre- and\npostmenopausal women. Ultimately, our application of the\nframework led to the decision that there were no meaningful\ninteractions among the 14 environmental factors or the polygenic\nscore within the pre- or postmenopausal groups. As an example,\namong premenopausal women, the pairwise interaction between\nfamily history and parity was statistically significant ( P ¼ .022 for\ninteraction; criterion A; Supplementary Table 2, O , available\nonline). Parity also appeared to be more protective among women\nwith a family history of ovarian cancer (OR ¼ 0.25 for 3 þ parity\ncompared with nulliparity) vs women without a family history\n(OR ¼ 0.52) (criterion B); this interaction may also be biologically\nplausible (criterion C). Elevated progesterone levels during preg-\nnancy may clear genetically abnormal cells in the Fallopian tube\nfimbriae (33), which may preferentially benefit genetically driven\novarian cancers (34). Although this potential pairwise interaction\nmay be useful for individual-level precision prevention, it would\nhave minor impacts on ovarian cancer risk stratification because\nof the low proportion of people with a positive family history of\novarian cancer [approximately 2% ( 35)] as well as the low abso-\nlute risk of ovarian cancer among premenopausal women ( 36)\n(criterion D). Thus, we concluded that it is not necessary to\nM. T. Phung et al. | 1421\n\nTable 1. Characteristics of participants with (cases) and without (controls) ovarian cancer included in the analysis, by age and menopausal status group\nPremenopausal women\nyounger than 45 years\nPremenopausal\nwomen aged 45-54 years\nPostmenopausal\nwomen aged 45-54 years\nPostmenopausal\nwomen aged 55-64 years\nPostmenopausal\nwomen aged 65-84 years\nCase partici-\npants\nControl partici-\npants\nCase partici-\npants\nControl partici-\npants\nCase partici-\npants\nControl partici-\npants\nCase partici-\npants\nControl partici-\npantss\nCase partici-\npants\nControl partici-\npants\n(n ¼ 965) (n ¼ 2111) (n ¼ 1269) (n ¼ 2109) (n ¼ 903) (n ¼ 1214) (n ¼ 2493) (n ¼ 3502) (n ¼ 2226) (n ¼ 3148)\nOvarian Cancer Association Consortium study, No. (%)\nAUS 2001-2005Australia 114 (11.8) 266 (12.6) 183 (14.4) 226 (10.7) 134 (14.8) 146 (12.0) 479 (19.2) 454 (13.0) 435 (19.5) 390 (12.4)\nDOV 2002-2009Washington, USA 116 (12.0) 182 (8.6) 209 (16.5) 311 (14.7) 99 (11.0) 122 (10.0) 425 (17.0) 660 (18.8) 224 (10.1) 414 (13.2)\nGER 1993-1998Germany 26 (2.7) 90 (4.3) 25 (2.0) 66 (3.1) 15 (1.7) 53 (4.4) 72 (2.9) 175 (5.0) 42 (1.9) 125 (4.0)\nHAW1993-2008Hawaii, USA 105 (10.9) 246 (11.7) 89 (7.0) 174 (8.3) 111 (12.3) 127 (10.5) 175 (7.0) 240 (6.9) 203 (9.1) 290 (9.2)\nHOP 2003-2009Western Pennsylvania,\nnortheast Ohio, western\nNew York, USA\n64 (6.6) 176 (8.3) 120 (9.5) 354 (16.8) 34 (3.8) 125 (10.3) 208 (8.3) 489 (14.0) 252 (11.3) 534 (17.0)\nNEC 1992-2008New Hampshire,\neastern Massachusetts,\nUSA\n235 (24.4) 496 (23.5) 249 (19.6) 354 (16.8) 187 (20.7) 214 (17.6) 422 (16.9) 542 (15.5) 347 (15.6) 452 (14.4)\nNJO 2002-2009New Jersey, USA 22 (2.3) 19 (0.9) 50 (3.9) 43 (2.0) 20 (2.2) 21 (1.7) 77 (3.1) 154 (4.4) 45 (2.0) 205 (6.5)\nUCI 1994-2005Southern California, USA 41 (4.2) 132 (6.3) 74 (5.8) 99 (4.7) 33 (3.7) 74 (6.1) 101 (4.1) 150 (4.3) 117 (5.3) 140 (4.4)\nUSC 1993-2010Los Angeles, California,\nUSA\n242 (25.1) 504 (23.9) 270 (21.3) 482 (22.9) 270 (29.9) 332 (27.3) 534 (21.4) 638 (18.2) 561 (25.2) 598 (19.0)\nAge at diagnosis for cases/reference age for controls, year\nMean (SD) 38.3 (5.28) 36.9 (5.92) 48.9 (2.63) 48.7 (2.63) 51.4 (2.31) 51.4 (2.34) 59.6 (2.77) 59.5 (2.78) 70.8 (4.46) 70.9 (4.49)\nMedian (Min, Max) 40.0 (20.0, 44.0) 38.0 (18.0, 44.0) 49.0 (45.0, 54.0) 49.0 (45.0, 54.0) 52.0 (45.0, 54.0) 52.0 (45.0, 54.0) 60.0 (55.0, 64.0) 59.0 (55 .0, 64.0) 70.0 (65.0, 84.0) 70.0 (65.0, 84.0)\nRace/ethnicity, No. (%)\nAsian 111 (11.5) 148 (7.0) 108 (8.5) 121 (5.7) 69 (7.6) 61 (5.0) 100 (4.0) 99 (2.8) 130 (5.8) 158 (5.0)\nBlack 30 (3.1) 54 (2.6) 24 (1.9) 45 (2.1) 28 (3.1) 24 (2.0) 57 (2.3) 53 (1.5) 35 (1.6) 52 (1.7)\nHispanic White 67 (6.9) 135 (6.4) 49 (3.9) 92 (4.4) 59 (6.5) 67 (5.5) 120 (4.8) 110 (3.1) 70 (3.1) 56 (1.8)\nNon-Hispanic White 691 (71.6) 1603 (75.9) 1044 (82.3) 1749 (82.9) 688 (76.2) 1005 (82.8) 2121 (85.1) 3116 (89.0) 1927 (86.6) 2778 (88.2)\nOther\na 62 (6.4) 157 (7.4) 41 (3.2) 97 (4.6) 54 (6.0) 55 (4.5) 84 (3.4) 119 (3.4) 58 (2.6) 99 (3.1)\nMissing 4 (0.4) 14 (0.7) 3 (0.2) 5 (0.2) 5 (0.6) 2 (0.2) 11 (0.4) 5 (0.1) 6 (0.3) 5 (0.2)\nEducation level, No. (%)\nLess than high school 55 (5.7) 95 (4.5) 87 (6.9) 105 (5.0) 86 (9.5) 89 (7.3) 352 (14.1) 348 (9.9) 467 (21.0) 439 (13.9)\nHigh school 205 (21.2) 389 (18.4) 261 (20.6) 374 (17.7) 172 (19.0) 251 (20.7) 585 (23.5) 805 (23.0) 634 (28.5) 932 (29.6)\nSome college 297 (30.8) 605 (28.7) 376 (29.6) 644 (30.5) 275 (30.5) 374 (30.8) 742 (29.8) 1021 (29.2) 593 (26.6) 869 (27.6)\nCollege graduate or above 396 (41.0) 959 (45.4) 529 (41.7) 935 (44.3) 346 (38.3) 461 (38.0) 742 (29.8) 1237 (35.3) 441 (19.8) 810 (25.7)\nMissing 12 (1.2) 63 (3.0) 16 (1.3) 51 (2.4) 24 (2.7) 39 (3.2) 72 (2.9) 91 (2.6) 91 (4.1) 98 (3.1)\na Other includes mixed race and those that do not belong in one of the specified racial/ethnic groups. AUS¼ Australian Ovarian Cancer Study; DOV ¼ Diseases of the Ovary and their Evaluation; GER¼ German Ovarian\nCancer Study; HAW ¼ Hawaii Ovarian Cancer Case-Control Study; HOP ¼ Hormones and Ovarian Cancer Prediction; NEC ¼ New England Case Control Study; NJO ¼ New Jersey Ovarian Cancer Study; SD ¼ Standard\ndeviation; USA ¼ United States of America; UCI ¼ University California Irvine Ovarian Study; USC ¼ Study of Lifestyle and Women’s Health.\n1422 | JNCI: Journal of the National Cancer Institute , 2023, Vol. 115, No. 11\n\ninclude an interaction term for family history and parity in a risk\nstratification model.\nOur proposed framework has some level of subjectivity. The\nrisk associations for 3 of the 4 risk factors that drove our conclu-\nsion that associations differ by menopausal status were not stat-\nistically significantly different in the 2 strata (criterion A) but met\nthe other 3 criteria used for evaluation. Some investigators, how-\never, may want to prioritize statistical significance (either using\nthe interaction test presented here or using the Bayes false-\npositive probability) over the other 3 criteria and only use criteria\nB through D to decide against there being an interaction.\nOperationally, we decided that criterion A or B must be met\nbefore criteria C and D are considered. When criteria conflict\nwith each other, however, we considered all criteria to inform our\ndecision-making process (see the examples earlier). Another\nexample is the age-parity interaction among postmenopausal\nwomen. The interaction was statistically significant ( P ¼ .009 for\ninteraction; criterion A) and the prevalence of ever having given\nbirth [85% (37)] is sufficient for this potential interaction to have\na meaningful impact on risk stratification (criterion D). There\nwas no pattern in the odds ratios for parity across the age groups\n(Supplementary Table 2, C, available online; criterion B), suggest-\ning that this is a chance finding. We therefore determined, based\non applying our framework, that this was not an interaction that\nshould be incorporated into a risk stratification model.\nIn conclusion, the application of our 4-criterion interaction\nevaluation framework ( Supplementary Tables 2, A-F , available\nonline) demonstrates that menopausal status modifies the asso-\nciation of at least one ovarian cancer risk/protective factor and\nthe disease risk, supporting the use of separate models by meno-\npausal status in risk stratification. The menopausal status–risk\nfactors interactions are likely not influenced by histotype\nbecause the distributions are similar between pre- and postme-\nnopausal women aged 45 to 54 years ( Supplementary Table 3 ,\navailable online). The finding of no age–risk factor interactions\ncould in part be due to the differences in histotype distributions\nacross age groups. Interaction analyses stratified by histotype,\nhowever, would not be meaningful because of the small sample\nsize of the rare histotypes. Additional research in prospective\ncohorts is needed to estimate absolute risk incorporating interac-\ntions to assess their impact on risk stratification.\nTo develop meaningful risk stratification models, it is critical\nfirst to comprehensively assess interactions using statistical,\nqualitative, biological, and practical approaches (criteria A-D).\nMany published cancer risk stratification models either do not\nconsider interactions or are based solely on P values (criterion A)\nto assess interactions ( 38-44). This approach has limitations\nbecause P values vary according to sample size, and there are\nissues related to multiple comparison. As such, we propose a\nframework that co-emphasizes the statistical (criterion A) and\nqualitative (criterion B) approaches and also includes the biologi-\ncal approach (criterion C) and practical approach (criterion D).\nComprehensive interaction analysis for risk stratification can\nmost effectively be done within consortia with large sample sizes.\nContinued collaboration in the field is necessary, and using the\ndata fully must be a priority to move closer to realizing the goals\nof precision cancer prevention.\nData availability\nThe data generated in this study are not publicly available\nbecause of limitations imposed by the original studies in which\nthese data were collected. The corresponding author will\nTable 2. Associations between risk/protective factors and ovarian cancer that differed by menopausal status among women aged 45-54 years\nRisk/protective\nfactor\nAll women aged 45-54 years (pre- and postmenopausal\ncombined) Premenopausal women aged 45-54 years Postmenopausal women aged 45-54 years\nCase\nparticipants,a\nNo.\nControl\nparticipants,a\nNo.\nORb\n(95% CI)\nCase\nparticipants,a\nNo.\nControl\nparticipants,a\nNo.\nORb\n(95% CI)\nCase\nparticipants,a\nNo.\nControl\nparticipants,a\nNo.\nORb\n(95% CI)\nP-\ninteractionc\nInteraction\ncriteria\nmetd\nBreastfeeding\nNever 1212 1278 1.0 693 763 1.0 519 515 1.0\n<12 months 551 978 0.76 (0.64 to 0.89) 327 595 0.78 (0.62 to 0.98) 224 383 0.71 (0.55 to 0.92)\n/C21 12 months 397 998 0.59 (0.49 to 0.72) 240 706 0.53 (0.42 to 0.68) 157 292 0.69 (0.51 to 0.94) .064 (b), (c), (d)\nDepot-medroxyprogesterone acetate use\nNo 1886 2793 1.0 1109 1793 1.0 777 1000 1.0\nYes 21 76 0.61 (0.36 to 1.02) 10 52 0.51 (0.26 to 1.00) 11 24 0.80 (0.38 to 1.69) .35 (b), (c), (d)\nFirst-degree family history of ovarian cancer\nNo 1610 2548 1.0 943 1633 1.0 667 915 1.0\nYes 119 87 2.15 (1.56 to 2.97) 69 48 2.43 (1.58 to 3.73) 50 39 1.83 (1.15 to 2.91) .39 (b), (c), (d)\nEndometriosis\nNo 1889 3058 1.0 1128 1981 1.0 761 1077 1.0\nYes 269 255 1.60 (1.32 to 1.95) 135 123 1.94 (1.47 to 2.57) 134 132 1.33 (1.00 to 1.76) .041 (a), (b), (c), (d)\na Numbers may not sum to total because of missing values. CI¼ confidence interval; OR ¼ odds ratio.\nb Pooled estimates from logistic regression models in the 50 imputed datasets, adjusted for age at diagnosis for cases/reference age for controls (45-49 years vs 50-54 years), race/ethnicity, education level, and Ovarian\nCancer Association Consortium study.\nc P value for interaction between risk or protective factor and menopausal status using the likelihood ratio test.\nd Criteria to assess interactions: (a) P < .05 for interaction; (b) odds ratios of a factor across the levels of the other factor are consistent, and the differences in magnitude are large; (c) the interaction is biologically\nplausible; and (d) the prevalence of the risk factors is large enough so that the interaction would have a meaningful impact on the risk stratification model.\nM. T. Phung et al. | 1423\n\nfacilitate access through existing data request processes for the\nOvarian Cancer Association Consortium.\nAuthor contributions\nMinh Tung Phung, PhD, MPH (Conceptualization; Formal analy-\nsis; Methodology; Writing—original draft; Writing—review & edit-\ning), Malcolm C. Pike, PhD (Funding acquisition; Writing—review\n& editing), Bhramar Mukherjee, PhD (Writing—review & editing),\nRafael Meza, PhD (Writing—review & editing), Gillian E. Hanley,\nPhD (Writing—review & editing), Kathleen R. Cho, MD (Writing—\nreview & editing), Andrew Berchuck, MD (Funding acquisition;\nWriting—review & editing), Argyrios Ziogas, PhD (Funding acquis-\nition; Writing—review & editing), Nur Zeinomar, PhD (Funding\nacquisition; Writing—review & editing), Anna H. Wu, PhD\n(Funding acquisition; Writing—review & editing), Penelope M.\nWebb, PhD (Funding acquisition; Writing—review & editing),\nLinda J. Titus, PhD (Funding acquisition; Writing—review & edit-\ning), Kathryn L. Terry, ScD (Funding acquisition; Writing—review\n& editing), Bo Qin, PhD (Funding acquisition; Writing—review &\nediting), Celeste Leigh Pearce, PhD, MPH (Conceptualization;\nFunding acquisition; Methodology; Resources; Supervision;\nWriting—original draft; Writing—review & editing), Paul D. P.\nPharoah, PhD (Funding acquisition; Writing—review & editing),\nFrancesmary Modugno, PhD, MPH (Funding acquisition;\nWriting—review & editing), Allan Jensen, PhD (Funding acquisi-\ntion; Writing—review & editing), Holly R. Harris, ScD, MPH\n(Funding acquisition; Writing—review & editing), Marc T.\nGoodman, PhD (Funding acquisition; Writing—review & editing),\nRenee T. Fortner, PhD (Funding acquisition; Writing—review &\nediting), Jennifer Anne Doherty, MS, PhD (Funding acquisition;\nWriting—review & editing), Daniel W. Cramer, MD, ScD (Funding\nacquisition; Writing—review & editing), Jenny Chang-Claude,\nPhD (Funding acquisition; Writing—review & editing), Michael E.\nCarney, MD (Funding acquisition; Writing—review & editing),\nElisa V. Bandera, MD, PhD (Funding acquisition; Writing—review\n& editing), Hoda Anton-Culver, PhD (Funding acquisition;\nWriting—review & editing), Karen McLean, MD, PhD (Writing—\nreview & editing), Alice W. Lee, PhD, MPH (Writing—review &\nediting), Kirsten B. Moysich, MS, PhD (Funding acquisition;\nWriting—review & editing), Britton Trabert, PhD\n(Conceptualization; Funding acquisition; Methodology;\nSupervision; Writing—original draft; Writing—review & editing).\nFunding\nThe Ovarian Cancer Association Consortium is supported by a\ngrant from the Ovarian Cancer Research Fund thanks to dona-\ntions by the family and friends of Kathryn Sladek Smith (PPD/\nRPCI.07). The scientific development and funding for this project\nwere in part supported by the US National Cancer Institute\nGAME-ON (Genetic Associations and Mechanisms in Oncology)\nPost-GWAS Initiative (U19-CA148112). This study used data gen-\nerated by the Wellcome Trust Case Control consortium, which\nwas funded by the Wellcome Trust under award No. 076113. The\nresults published here are in part based on data generated by The\nCancer Genome Atlas Pilot Project established by the National\nCancer Institute and National Human Genome Research\nInstitute (dbGap accession No. phs000178.v8.p7).\nThe Ovarian Cancer Association Consortium OncoArray geno-\ntyping project was funded through grants from the US National\nInstitutes of Health (CA1X01HG007491-01, U19-CA148112, R01-\nCA149429, and R01-CA058598; Canadian Institutes of Health\nResearch (MOP-86727 and the Ovarian Cancer Research Fund\n(A.B.). The COGS project was funded through a European\nCommission’s Seventh Framework Programme grant (agreement\nNo. 223175 - HEALTH-F2-2009-223175).\nFunding for individual studies: AUS: The Australian Ovarian\nCancer Study was supported by the US Army Medical Research\nand Materiel Command (DAMD17-01-1-0729); National Health &\nMedical Research Council of Australia (199600, 400413, and\n400281); Cancer Councils of New South Wales, Victoria,\nQueensland, South Australia, and Tasmania; and Cancer\nFoundation of Western Australia (Multi-State Applications 191,\n211, and 182). The Australian Ovarian Cancer Study gratefully\nacknowledges additional support from Ovarian Cancer Australia\nand the Peter MacCallum Foundation; DOV: National Institutes\nof Health R01-CA112523 and R01-CA87538; GER: German Federal\nMinistry of Education and Research, Programme of Clinical\nBiomedical Research (01 GB 9401), and the German Cancer\nResearch Center (DKFZ); HAW: US National Institutes of Health\n(R01-CA58598, N01-CN-55424, and N01-PC-67001); HOP:\nUniversity of Pittsburgh School of Medicine Dean’s Faculty\nAdvancement Award (F. Modugno), US Department of Defense\n(DAMD17-02-1-0669), and National Cancer Institute (K07-\nCA080668, R01-CA95023, P50-CA159981, MO1-RR000056, R01-\nCA126841); NEC: R01-CA54419 and P50-CA105009 and US\nDepartment of Defense W81XWH-10-1-02802; NJO: National\nCancer Institute (NIH-K07 CA095666, R01-CA83918, NIH-K22-\nCA138563, and P30-CA072720) and the Rutgers Cancer Institute\nof New Jersey; UCI: Australian Ovarian Cancer Study R01-\nCA058860 and the Lon V. Smith Foundation grant LVS-39420;\nUSC: P01CA17054, P30CA14089, R01CA61132, N01PC67010,\nR03CA113148, R03CA115195, N01CN025403, and California\nCancer Research Program (00-01389 V-20170, 2II0200).\nFunding for individuals: B.T.: Cancer Center Support Grant,\nNational Cancer Institute, P30CA040214; P.M.W.: National Health\n& Medical Research Council of Australia GNT1173346; C.L.P.: the\nNational Institutes of Health/National Cancer Institute Support\nGrant P30 CA046592. M.C.P. was supported in part through the\nNational Institutes of Health/National Cancer Institute Support\nGrant P30 CA008748 (P.I. S.M. Vickers) to Memorial Sloan\nKettering Cancer Center.\nConflicts of interest\nP.M.W. has received a speaker’s fee and funding from\nAstraZeneca (2017-2020) for an unrelated study of ovarian can-\ncer. E.V.B. served in Pfizer’s Advisory Board to enhance participa-\ntion of people from minority groups in clinical trials. Other\nauthors do not have any conflicts of interest.\nAcknowledgements\nWe are grateful to the family and friends of Kathryn Sladek\nSmith for their generous support of the Ovarian Cancer\nAssociation Consortium through their donations to the Ovarian\nCancer Research Fund. The OncoArray and COGS genotyping\nprojects would not have been possible without the contributions\nof the following: Per Hall (COGS); Douglas F. Easton, Kyriaki\nMichailidou, Manjeet K. Bolla, Qin Wang (BCAC); Marjorie J.\nRiggan (OCAC); Rosalind A. Eeles, Douglas F. Easton, Ali Amin Al\nOlama, Zsofia Kote-Jarai, and Sara Benlloch (PRACTICAL);\nGeorgia Chenevix-Trench, Antonis Antoniou, Lesley McGuffog,\nFergus Couch, and Ken Offit (CIMBA); Joe Dennis, Jonathan P.\n1424 | JNCI: Journal of the National Cancer Institute , 2023, Vol. 115, No. 11\n\nTyrer, Siddhartha Kar, Alison M. Dunning, Andrew Lee, and Ed\nDicks; Craig Luccarini and the staff of the Centre for Genetic\nEpidemiology Laboratory; Javier Benitez, Anna Gonzalez-Neira,\nand the staff of the CNIO genotyping unit; Jacques Simard and\nDaniel C. Tessier; Francois Bacot, Daniel Vincent, Sylvie\nLaBoissie` re, Frederic Robidoux, and the staff of McGill University\nand G /C19enome Qu /C19ebec Innovation Centre; Stig E. Bojesen, Sune F.\nNielsen, Borge G. Nordestgaard, and the staff of the Copenhagen\nDNA laboratory; and Julie M. Cunningham, Sharon A.\nWindebank, Christopher A. Hilker, Jeffrey Meyer, and the staff of\nMayo Clinic Genotyping Core Facility. We pay special tribute to\nthe contribution of Brian Henderson to the GAME-ON consortium\nuntil he sadly passed away on June 20, 2015; to Olga M.\nSinilnikova for her contribution to CIMBA and for her part in the\ninitiation and coordination of GEMO until she sadly passed away\non June 30, 2014, and to Catherine M. Phelan for her contribution\nto OCAC and coordination of the OncoArray until she passed\naway on September 22, 2017. We thank the study participants,\ndoctors, nurses, clinical and scientific collaborators, health care\nproviders, and health information sources who have contributed\nto the many studies contributing to this manuscript.\nAcknowledgements for individual studies: AUS: The AOCS also\nacknowledges the cooperation of the participating institutions in\nAustralia and the contribution of the study nurses, research\nassistants, and all clinical and scientific collaborators. The com-\nplete AOCS Study Group can be found at www.aocstudy.org.W e\nwould like to thank all the women who participated in this\nresearch program; GER: The German Ovarian Cancer Study (GER)\nthank Ursula Eilber for competent technical assistance; NJO:W e\nthank Drs Sara Olson, Lisa Paddock, Lorna Rodriguez, and all par-\nticipants and research staff at the Rutgers Cancer Institute of\nNew Jersey, the New Jersey State Cancer Registry, and Memorial\nSloan Kettering Cancer Center.\nThe funder had no role in study design, data collection and\nanalysis, decision to publish, or preparation of the manuscript.\nEarlier results were presented at the American Association for\nCancer Research annual meeting in June 2020 as an e-poster.\nReferences\n1. Hanley GE, Pearce CL, Talhouk A, et al. Outcomes from opportun-\nistic salpingectomy for ovarian cancer prevention. JAMA Netw\nOpen. 2022;5(2):e2147343. doi: 10.1001/jamanetworkopen.2021.\n47343.\n2. Menon U, Karpinskyj C, Gentry-Maharaj A. Ovarian cancer pre-\nvention and screening. Obstet Gynecol . 2018;131(5):909-927.\ndoi:10.1097/AOG.0000000000002580.\n3. Beral V, Doll R, Hermon C, Peto R, Reeves G; Collaborative Group\non Epidemiological Studies of Ovarian Cancer. Ovarian cancer\nand oral contraceptives: collaborative reanalysis of data from\n45 epidemiological studies including 23,257 women with ovar-\nian cancer and 87,303 controls. Lancet. 2008;371(9609):303-314.\ndoi:10.1016/S0140-6736(08)60167-1.\n4. Pearce CL, Rossing MA, Lee AW, et al.; Ovarian Cancer\nAssociation Consortium. Combined and interactive effects of\nenvironmental and GWAS-identified risk factors in ovarian can-\ncer. Cancer Epidemiol Biomarkers Prev . 2013;22(5):880-890.\ndoi:10.1158/1055-9965.EPI-12-1030-T.\n5. Wentzensen N, Poole EM, Trabert B, et al. Ovarian cancer risk\nfactors by histologic subtype: an analysis from the Ovarian\nCancer Cohort Consortium. J Clin Oncol . 2016;34(24):2888-2898.\ndoi:10.1200/JCO.2016.66.8178.\n6. Lee AW, Rosenzweig S, Wiensch A, et al.; Australian Ovarian\nCancer Study Group. Expanding our understanding of ovarian\ncancer risk: the role of incomplete pregnancies.J Natl Cancer Inst.\n2021;113(3):301-308. doi:10.1093/jnci/djaa099.\n7. Babic A, Sasamoto N, Rosner BA, et al. Association between\nbreastfeeding and ovarian cancer risk. JAMA Oncol .\n2020;6(6):e200421. doi:10.1001/jamaoncol.2020.0421.\n8. Phung MT, Lee AW, Wu AH, et al.; Ovarian Cancer Association\nConsortium. Depot-medroxyprogesterone acetate use is associ-\nated with decreased risk of ovarian cancer: the mounting evidence\nof a protective role of progestins.Cancer Epidemiol Biomarkers Prev.\n2021;30(5):927-935. doi:10.1158/1055-9965.EPI-20-1355.\n9. Dixon-Suen SC, Nagle CM, Thrift AP, et al.; Ovarian Cancer\nAssociation Consortium. Adult height is associated with\nincreased risk of ovarian cancer: a Mendelian randomisation\nstudy. Br J Cancer . 2018;118(8):1123-1129. doi: 10.1038/s41416-\n018-0011-3\n10. Day FR, Thompson DJ, Helgason H, et al.; LifeLines Cohort\nStudy; InterAct Consortium; kConFab/AOCS Investigators;\nEndometrial Cancer Association Consortium; Ovarian Cancer\nAssociation Consortium; PRACTICAL consortium. Genomic\nanalyses identify hundreds of variants associated with age at\nmenarche and support a role for puberty timing in cancer risk.\nNat Genet. 2017;49(6):834-841. doi:10.1038/ng.3841.\n11. Wu AH, Pearce CL, Lee AW, et al. Timing of births and oral con-\ntraceptive use influences ovarian cancer risk. Int J Cancer .\n2017;141(12):2392-2399. doi:10.1002/ijc.30910.\n12. Lee AW, Ness RB, Roman LD, et al.; Ovarian Cancer Association\nConsortium. Association between menopausal estrogen-only\ntherapy and ovarian carcinoma risk. Obstet Gynecol .\n2016;127(5):828-836. doi:10.1097/AOG.0000000000001387.\n13. Lee AW, Wu AH, Wiensch A, et al.; Ovarian Cancer Association\nConsortium. Estrogen plus progestin hormone therapy and ovar-\nian cancer: a complicated relationship explored. Epidemiology.\n2020;31(3):402-408. doi:10.1097/EDE.0000000000001175.\n14. Olsen CM, Nagle CM, Whiteman DC, et al.; Ovarian Cancer\nAssociation Consortium. Obesity and risk of ovarian cancer sub-\ntypes: evidence from the Ovarian Cancer Association\nConsortium. Endocr Relat Cancer. 2013;20(2):251-262. doi:10.1530/\nERC-12-0395.\n15. Lee A, Yang X, Tyrer J, et al. A comprehensive epithelial tubo-\novarian cancer risk prediction model incorporating genetic and\nepidemiological risk factors. J Med Genet . 2020;59(7):632-643.\ndoi:10.1136/jmedgenet-2021-107904.\n16. Trentham-Dietz A, Sprague BL, Hampton JM, et al. Modification\nof breast cancer risk according to age and menopausal status: a\ncombined analysis of five population-based case-control stud-\nies. Breast Cancer Res Treat . 2014;145(1):165-175. doi: 10.1007/\ns10549-014-2905-y.\n17. Merritt MA, Green AC, Nagle CM, Webb PM; ACSO Cancer, AOCS\nGroup. Talcum powder, chronic pelvic inflammation and\nNSAIDs in relation to risk of epithelial ovarian cancer. Int J\nCancer. 2008;122(1):170-176. doi:10.1002/ijc.23017.\n18. Royar J, Becher H, Chang-Claude J. Low-dose oral contracep-\ntives: protective effect on ovarian cancer risk. Int J Cancer .\n2001;95(6):370-374. doi: 10.1002/1097-\n0215(20011120)95:6<370::aid-ijc1065>3.0.co;2-t.\n19. Bodelon C, Cushing-Haugen KL, Wicklund KG, Doherty JA,\nRossing MA. Sun exposure and risk of epithelial ovarian cancer.\nCancer Causes Control . 2012;23(12):1985-1994. doi: 10.1007/\ns10552-012-0076-x.\n20. Lurie G, Terry KL, Wilkens LR, et al. Pooled analysis of the asso-\nciation of PTGS2 rs5275 polymorphism and NSAID use with\nM. T. Phung et al. | 1425\n\ninvasive ovarian carcinoma risk. Cancer Causes Control .\n2010;21(10):1731-1741. doi:10.1007/s10552-010-9602-x.\n21. Ness RB, Dodge RC, Edwards RP, Baker JA, Moysich KB.\nContraception methods, beyond oral contraceptives and tubal\nligation, and risk of ovarian cancer. Ann Epidemiol .\n2011;21(3):188-196. doi:10.1016/j.annepidem.2010.10.002.\n22. Terry KL, De Vivo I, Titus-Ernstoff L, Shih MC, Cramer DW.\nAndrogen receptor cytosine, adenine, guanine repeats, and hap-\nlotypes in relation to ovarian cancer risk. Cancer Res .\n2005;65(13):5974-5981. doi:10.1158/0008-5472.CAN-04-3885.\n23. Bandera EV, King M, Chandran U, Paddock LE, Rodriguez-\nRodriguez L, Olson SH. Phytoestrogen consumption from foods\nand supplements and epithelial ovarian cancer risk: a\npopulation-based case control study. BMC Womens Health .\n2011;11:40. doi:10.1186/1472-6874-11-40\n24. Ziogas A, Gildea M, Cohen P, et al. Cancer risk estimates for fam-\nily members of a population-based family registry for breast\nand ovarian cancer. Cancer Epidemiol Biomarkers Prev .\n2000;9(1):103-111.\n25. Wu AH, Pearce CL, Tseng CC, Pike MC. African Americans and\nHispanics remain at lower risk of ovarian cancer than non-\nHispanic Whites after considering nongenetic risk factors and\noophorectomy rates. Cancer Epidemiol Biomarkers Prev .\n2015;24(7):1094-1100. doi:10.1158/1055-9965.EPI-15-0023.\n26. Sullivan GM, Feinn R. Using effect size-or why the P value is not\nenough. JG r a dM e dE d u c. 2012;4(3):279-282. doi:10.4300/JGME-D-\n12-00156.1.\n27. Salehi F, Dunfield L, Phillips KP, Krewski D, Vanderhyden BC.\nRisk factors for ovarian cancer: an overview with emphasis on\nhormonal factors. J Toxicol Environ Health B Crit Rev . 2008;11(3-\n4):301-321. doi:10.1080/10937400701876095.\n28. Vercellini P, Somigliana E, Vigan /C18o P, Abbiati A, Barbara G,\nCrosignani PG. Endometriosis: current therapies and new phar-\nmacological developments. Drugs. 2009;69(6):649-675.\ndoi:10.2165/00003495-200969060-00002.\n29. Macci /C18o A, Madeddu C. Inflammation and ovarian cancer.\nCytokine. 2012;58(2):133-147. doi:10.1016/j.cyto.2012.01.015.\n3 0 . D u f f yD M ,K oC ,J oM ,B r a n n s t r o mM ,C u r r yT E .O v u l a t i o n :p a r a l -\nlels with inflammatory processes. Endocr Rev. 2019;40(2):369-416.\ndoi:10.1210/er.2018-00075.\n31. Tan DA, Almaria MJG. Postmenopausal endometriosis: drawing\na clearer clinical picture. Climacteric. 2018;21(3):249-255.\ndoi:10.1080/13697137.2018.1450855.\n32. Eskenazi B, Warner ML. Epidemiology of endometriosis. Obstet\nGynecol Clin North Am . 1997;24(2):235-258. doi: 10.1016/s0889-\n8545(05)70302-8.\n33. Rodriguez GC, Kauderer J, Hunn J, et al. Phase II trial of chemo-\npreventive effects of levonorgestrel on ovarian and fallopian\ntube epithelium in women at high risk for ovarian cancer: an\nNRG Oncology Group/GOG study. Cancer Prev Res (Phila) .\n2019;12(6):401-412. doi:10.1158/1940-6207.CAPR-18-0383.\n34. Milne RL, Osorio A, Ram\n/C19on y Cajal T, et al. Parity and the risk of\nbreast and ovarian cancer in BRCA1 and BRCA2 mutation car-\nriers. Breast Cancer Res Treat . 2010;119(1):221-232. doi: 10.1007/\ns10549-009-0394-1.\n35. Ramsey SD, Yoon P, Moonesinghe R, Khoury MJ. Population-\nbased study of the prevalence of family history of cancer: impli-\ncations for cancer screening and prevention. Genet Med .\n2006;8(9):571-575. doi:10.1097/01.gim.0000237867.34011.12.\n36. Roett MA, Evans P. Ovarian cancer: an overview. Am Fam\nPhysician. 2009;80(6):609-616.\n37. Martinez GM, Daniels K, Febo-Vazquez I. Fertility of men and\nwomen aged 15-44 in the United States: national survey of fam-\nily growth, 2011-2015. Natl Health Stat Rep. 2018;(113):1-17.\n38. Colditz GA, Atwood KA, Emmons K, et al. Harvard report on\ncancer prevention volume 4: Harvard Cancer Risk Index. Risk\nIndex Working Group, Harvard Center for Cancer Prevention.\nCancer Causes Control . 2000;11(6):477-488. doi: 10.1023/\na:1008984432272 .\n39. Rosner BA, Colditz GA, Webb PM, Hankinson SE. Mathematical\nmodels of ovarian cancer incidence. Epidemiology.\n2005;16(4):508-515. doi:10.1097/01.ede.0000164557.81694.63.\n40. Vitonis AF, Titus-Ernstoff L, Cramer DW. Assessing ovarian cancer\nrisk when considering elective oophorectomy at the time of hys-\nterectomy. Obstet Gynecol . 2011;117(5):1042-1050. doi: 10.1097/\nAOG.0b013e318212fcb7.\n41. Pfeiffer RM, Park Y, Kreimer AR, et al. Risk prediction for breast,\nendometrial, and ovarian cancer in white women aged 50 y or\nolder: derivation and validation from population-based cohort\nstudies. PLoS Med . 2013;10(7):e1001492. doi: 10.1371/jour-\nnal.pmed.1001492.\n42. Pearce CL, Stram DO, Ness RB, et al. Population distribution of\nlifetime risk of ovarian cancer in the United States. Cancer\nEpidemiol Biomarkers Prev . 2015;24(4):671-676. doi: 10.1158/1055-\n9965.EPI-14-1128.\n43. Clyde MA, Palmieri Weber R, Iversen ES, et al.; on behalf of the\nOvarian Cancer Association Consortium. Risk prediction for epi-\nthelial ovarian cancer in 11 United States-based case-control\nstudies: incorporation of epidemiologic risk factors and 17 con-\nfirmed genetic loci. Am J Epidemiol . 2016;184(8):579-589.\ndoi:10.1093/aje/kww091.\n44. Lee A, Yang X, Tyrer J, et al. Comprehensive epithelial tubo-\novarian cancer risk prediction model incorporating genetic and\nepidemiological risk factors. J Med Genet . 2022;59(7):632-643.\ndoi:10.1136/jmedgenet-2021-107904.\n1426 | JNCI: Journal of the National Cancer Institute , 2023, Vol. 115, No. 11","source_license":"CC0","license_restricted":false}