F Purpose: We evaluate the possibility, in a MIC country such as Argentina, of having the data from the normally required clinical-pathological report of the BC, in order to apply 10 selected published algorithms that are offered as alternatives to ODX to predict the ODX RS specifically in patients with "early-stage breast cancer" cases. Methods: From the total of 1832 (stage I to IV) BC cases reported, between 2012 and 2014, to PROYCAM2012 (www.cancerdemama2012.org.ar) a consortium (still in force) created for the study of BC in Argentina a subset of 706 (38,5%) "early-stage breast cancer", was identified and analyzed. An online search and selection of published scientific research on SP, (2010 onwards), was carried out. Ten publications were selected and analyzed and the SP used in them identified. The presence of the SP shared by the different studies selected on -line was analyzed in our series, namely: age, tumor size, histology, grade, Estrogen/ Progesterone receptor, Her2 / neu status and Ki-67. Results: The subset of 706 (38,5%) showed the following characteristics: predominant in menopausal women (72%), average age of 61 years, 3.2% bilaterally and a personal (30,3%) and/or family history (9,3%) of BC. Pathologically, the average size of the tumors is between 1.8 -2.0, with predominance of infiltrating ductal lesions (67, 7%) grade2 (46,7%). From the total of 10 nomograms selected only in 1 our series can complete the required two SPs: HR (96, 6%) andHer2 / neu (100%). In a second nomogram our cases complete the three of the proposed SPs being: Grade (94 %), ER (100%) and PR (99.6%). In a third nomogram cases complete the required six SPs being these: tumor size (88.5%), patient age (100%), laterality (100%), ER receptor (100%), PR receptor (99.6%) and HER2 / neu (100%). In a fourth our cases nomogram completes the required six SPs, being these: size (88,5%), ER (100%), PR status (99,6%), HER2neu (100%), Nottingham (94%) and histomorphology (77,5%). In a fifth nomogram the cases complete the required five SPs: age (100%), tumor size (88.5%), grade (94), PR status (99,6) and histologic type (77,5). In the five remaining nomograms evaluated the percentage of cases that could complete the required SPs only reached 61.5% of the cases. # Introduction reast Cancer (BC) is the most commonly diagnosed cancer amongst women worldwide [1] and is a leading cause of death and disability among women in low-and middle-income countries (MICs)among which is Argentina [2]. Nowadays in BC, beyond the standard determination of cancer stage according to the classic anatomical criteria of the TNM classification, expression of estrogen and progesterone receptors and HER2/neu receptor are required. The study of genic profile (GP) of cancer has also been encouraged although, for now, not in mandatory form [3]Implementation of the multigene panels assay has led to a change in the manner in which chemotherapy is utilized mainly in patients with,early stage, estrogen receptor (ER)positive, Her2-neu negative, lymph node-negative[ER (+) / HER2 (-) / lymph node-negative], BC ensuring that patients at highest risk of recurrence are prescribed systemic treatment, while at the same time sparing lowrisk patients potential adverse events from therapy unlikely to influence their survival [4,5]. Multigene panels can provide better risk discrimination relative to clinic-pathological factors, which are significantly superior to traditional prognostic factors in predicting clinical outcome and identifying patients who can be spared chemotherapy safely. There are several tests available at the moment. The Oncotype DX (ODX) BC recurrence test (ODXRS) is the one recommended based on the experience accumulated since its implementation in 2010 [6] Unfortunately, all these tests are expensive. Oncotype DX (ODX) is expensive and is performed in only 1/3 of patients with BC positive for the estrogen receptor (ER) in developed countries [7,8]and are not affordable or available for the majority of the breast cancer patients globally [9]. The economic nonaccessibility and / or technical availability are also the main reasons why in a middle-income country (MIC), such as Argentina, the study is only carried out in very few cases (0,23%) who fulfilled the established guidelines for gene-expression profile study and in a sporadic way [10]. Given the potential savings in cost and resource utilization, several algorithms have been proposed to predict Oncotype DX recurrence score (ODX RS) using commonly acquired clinical and histopathologic variables designated as subrogate parameters (SP). These studies reached different conclusions regarding which model demonstrated the best statistical discrimination power, mainly due to differences in clinical and pathologic variables used [11][12][13][14][15][16][17][18][19][20]. In this study we evaluate the possibility, in a MIC country like Argentina, of having the necessary data from the normally required clinical-pathological report of the BC, in order to apply some selected published algorithms that are offered as alternatives to ODX to predict the ODX RSspecifically in patients with "earlystage breast cancer". # II. # Material and Methods # a) Data source We retrospectively reviewed the pathology reports from successive incident BC cases reported, between 2012 and 2014, to PROYCAM2012 (www.cancerdemama2012.org.ar) a consortium (still in force) created for the study of BC in Argentina [10]. The PROYCAM2012 recruited a total of 1832 (stage I to IV) BC until the end of the period under study. From the total of cases reported a subset of 706 (38,5%) "ER (+) / HER2 (-) / lymph node-negative" BC, was identified and analyzed. # b) Surrogate parameters An online search and selection of published scientific research on the same thematic from 2010 onwards, was carried out [11][12][13][14][15][16][17][18][19][20]. The presence of the SP shared by the different studies selected was analyzed in our series, namely: age, tumor size, histology, grade, Estrogen/Progesterone receptor, Her2 / neu status and Ki-67. # c) Ethical Approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards, the National Law 25326 of Habeas Data Personal Data Protection and the National Patient Rights Act 26529. For retrospective studies (applies to our study): "for this type of study formal consent is not required''. # III. # Results # a) Substitute parameters in selected publications The selected 10 publications were analyzed and the SP used in them identified as shown in table 1. # b) Clinical-pathological characteristics and/or family history (9,3%) of BC. Pathologically, the average size of the tumors is between 1.8 -2.0, cm according to the left and right side, with predominance of infiltrating ductal lesions (67, 7%) grade2 (46,7%). # c) Substitute parameters evaluated The frequency of the presence, in the 706 cases of "ER (+) / HER2 (-) / negative lymph node" BC, of the selected SPs is shown in Table 3. Of the total of 8 (12) 378/1063 (35,5%) 1990-1999 Australia, Canada, Hungary, Italy, New Zealand, Slovenia, South Africa, Spain, Sweden, and Switzerland. Age, Histotype, Tumor size, Grade, ER/PgR , Her2/neu status and Ki-67. Rouanet R et al.,2013 (13) 714/1271 (56,1%) HR and HER2 status.HER2-HR+; HER2-HR-; HER2+HR+; HER2+HR-Turner B et al; 2015 (14) 299 cases with available Oncotype DX recurrence scores 2009-2013 USA ER, PR, HER-2, and Ki-67, Nottingham score (NS) and tumor size. Gage MM et al.2015 (15) 221ODX-tested ER(+)/HER2(-)/lymph nodenegative 2006-2013 USA. -Low grade and positive progesterone receptor tumors (LG+PR). -High grade or low estrogen receptor (ER) (ER < 20%) tumors (HG/ LER). Özmen V et al., 2016 (16) 165 ODX-tested ER(+)/HER2(-) /lymph nodenegative # Turkey Age, LN Status, Grade, ER score ?10%, PR score ?20% and Ki67 score. Harowicz MR et al., 2017 (17) 305ODX-tested ER (+) /HER2(-)/lymph node-negative # 2000-2014 USA Estrogen receptor status and progesterone receptor status along with different combinations of grade, proliferation indices (Ki-67, mitotic rate), HER2 status, and tumor size. Farrugia DJ et al,2017 (17) 237 The subset of 706 (38,5%) "ER (+) / HER2 (-) / lymph node-negative" cases selected from a database of 1832 cases of BC (stages I-IV) showed the following clinic-pathological characteristics (Table 2). The main clinical characteristics that define this group are: predominant in menopausal women (72%), average age of 61 years, 3.2% bilaterally and a personal (30,3%) parameters, 4 (50%) (Age, ER / PR and Her2 / neu) are present between 96 and 100% of cases.The next most frequently found SP were grade (94%) and size (88.5%). Finally, the least frequent were histology 77.5%) and Ki67 expression (61.5%). # d) Coincidence between substitute parameters The percentages in which, in our series, each of the SPs required in the different nomograms is fulfilled and also according to the possibility of a complete application of all SP of each proposed nomogram are shown in table 4. From the total of 10 nomograms selected only in 1 [ 13] our series can, in the 99.6% of the cases, complete the required two SPs being them: HR (96, 6%) and the Her2 / neu (100%). In a second nomogram [15] at least 94% of our cases complete the three of the proposed SPs being these: Grade (94 %), ER (100%) and PR (99.6%). In a third nomogram [11] at least 88.5% of the cases complete the required six SPs being these: tumor size (88.5%), patient age (100%), laterality (100%), ER receptor (100%), PR receptor (99.6%) and HER2 / neu (100%). In a fourth nomogram [19] 77,5% of cases complete the required six SPs, being these: size (88,5%), ER (100%), PR status (99,6%), HER2neu (100%), Nottingham (94%) and histomorphology (77,5%). In a fifth nomogram [20] also a 77,5% of the cases complete the required five SPs: age (100%), tumor size (88.5%), grade (94), PR status (99,6) and histologic type (77,5).Finally, in the five remaining nomograms evaluated [12,14,[16][17][18] the percentage of cases that could complete the required SPs only reached 61.5% of the cases in each of them because this was the percentage of cases in which the Ki67 status study was conducted to evaluate the rate of tumor proliferation. # IV. # Discussion BC in Argentina, as we reported earlier, has an epidemiological pattern and an incidence rate typical of a "western" and "developed" country [10] without major variations of both qualifiersin the BCover the last 40 years [21]We also report that, with the resources currently available, the BC can be staged properly, according to the latest version of the TNM, in 75% of cases [22] That a subset of no more than 706 (38.5%) of "early" BC "ER (+) / HER2 (-) / negative nodes" come from a database of 1832 new cases of BC, (stages I-IV) recorded between 2012 and 2014, presents its logic. The highest incidence of the whole "early stages" BC in developed countries, due to a massive and continuous use of mammogram screening in these populations [23][24][25] entails a high frequency of "ER (+) / HER2 (-) / negative nodes" BC. On the contrary, in our population, although the mammography is known and applied, but not in a massive and systematic way the frequency of this kind of BC ("ER (+) / HER2 (-) / negative nodes") is related to a lowest diagnosis of the whole types of "early stages BC" [10,22]. Genomic platform tests are now considered "standard of care" to maintain treatment decisionmaking for patients "ER (+) / HER2 (-) / negative lymph node" BC. Previous analyses of genomic platforms testing have assessed hypothetical cohorts under ideal conditions and concluded that testing had low costs relative to its benefits; the application of gene panels in clinical practice avoids overtreatment, with its possible adverse effects, in the short term and toxic in the longer term [26][27][28] as well as reducing treatment cost [29][30][31] MIC countries, such as Argentina, where gross national income per capita is between US$9,950 and $12,055 per year [32] are mainly characterized by fragmented and poorly coordinated medical care, moderate or high levels of poverty and disparities to access a basic standard of care not only for cancer but also for other complex diseases beyond of being covered by law. Patients in the public environment cannot pay for targeted therapy, so there are currently no hospital laboratories offering genomic platforms. OncotypeDx® is expensive [the current estimated cost is U$4000 [9]. The cost of the study is the main reason for the almost no realization in the past (2012-13) and nowadays in the MIC countries such as Argentina. None of the health sub-sectors recognizes this study (not included in the oncological diagnosis and treatment protocols accepted by law in our country) for which they do not reimburse their cost. The few cases performed were done privately, paid by the patients and performed abroad the country. For all the mentioned the current tendency, encouraged by research groups, is to use clinic pathologic variables for prediction of low-risk or high-risk OncotypeDx® Recurrence Score (ODXRS) using nomograms models. Quality prediction models depends on the amount and quality of data derived. In some aspects of the diagnosis of BC our country performs better than expected due to its economic development level. For example, immunohistochemistry (IHC) for HR and Her2 / neu is performed routinely in a high percentage of cases, in almost most laboratories for at least more than a decade. As such, IHC is an accessible and relatively inexpensive test and one that can be performed quite quickly. This is in sharp contrast to genomic test that are routinely performed abroad the country resulting in a prolonged time of realization and increased costs. Having, in our series, the results of HR and Her2 / neu in almost all cases, this allows us to apply 5/10 of the selected nomograms in which these required SPs are available between 88.5% and 99.6% of the cases. On the contrary, when it is necessary to know the proliferation index studied by the Ki67 expression as SP to fulfill a nomogram, this data is only present in 61.5% of the cases in the remaining 5 nomograms. # Conclusion Despite the fact that MIC oncologists are theoretically well informed in the use of genomic platforms [8,[33][34][35][36] they are far away from developed in the real-world a practice based in precision oncology; this kind of practice is a great challenge for themand frequently limited, when it is possible, to private practice. Our study, despite its limitations, provides some evidence for the design and orientation in our country of health policies and diagnostic interventions such as nomograms in the treatment of BC cancer, especially in early stages. These nomograms are useful tools to help to decide whether further OncotypeDx® testing is necessary and are excellent surrogates for patients for which OncotypeDx® testing is not affordable or even available. We demonstrate that, although there are gaps in the updated care process (for example: genetic profile) of BC, it is possible to use nomograms in our country through a comprehensive and complete collection and a careful subsequent analysis of the conventional parameters present in the diagnosis of BC. Special emphasis should be noted on the histopathological and immunohistochemical results, available almost routinely in our series, a fact that allows us to evaluate the RS and thus apply the corresponding therapy trying to achieve results similar to those that would be obtained with the use of GP. 1ReferenceBreast cancer cases (n, %) ER(+; HER2 (-);lymph node-negative/TotalNotification period (years). CountryClinical and pathologic markers required as SPAuerbach J et138 cases with availableUSATumorsize,patientage,al., 2010 (11)Oncotype DX recurrencelaterality, ER receptor, PRscoresreceptor results and HER2/neuresultDellapasqua Set al. 2012 © 2022 Global Journals Surrogate Parameters, Instead of the Genetic Profile, for Recurrence Risk Evaluation in "Early-Stage Breast Cancer Cases" in a Middle -Income Country Like Argentina Acknowledgement: Marcelo Guruceaga and Natalia Gonçalves for their continued support in the development of this project. * Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries FBray JFerlay ISoerjomataram RLSiegel LATorre AJemal 10.3322/caac.21492 30207593 CA Cancer J Clin 68 6 2018 Nov. 2018 Sep 12 * Breast cancer in young women in Latin America: an unmet, growing burden CVillarreal-Garza CAguila MCMagallanes-Hoyos AMohar EBargalló AMeneses 10.1634/theoncologist.2013-0321 Oncologist 18 12 2013 * Breast Cancer-Major changes in the American Joint Committee on Cancer Eighth edition cancer staging manual AEGiuliano JLConnolly SBEdge EAMittendorf HSRugo LJSolin 10.3322/caac.21393 28294295 CA Cancer J Clin 67 4 345 2017 Jul 8. 2017 Mar 14. 2017 Jul 8 CA Cancer J Clin * Spotlight on the utility of the Oncotype DX(®) breast cancer assay ZRSiow DeBoer RHLindeman GJMann GB Int J Womens Health 2018 Feb 21 * 10.2147/IJWH.S124520 29503586 PMC5827461 Review 10 2018 * Online accounts of gene expression profiling in early-stage breast cancer: Interpreting genomic testing for chemotherapy decision making. Health Expect ERoss JSwallow AKerr SCunningham-Burley 30387238 PMC6351409 2019 Feb. 2018 Nov 1 22 * Utilization and impact of 21-gene recurrence score assay for breast cancer in clinical practice across the United States: lessons learned from the 2010 to 2012 National Cancer Data Base analysis AOrucevic REHeidel JLBell 10.1007/s10549-016-3833-9 27206678 PMC4903105 Breast Cancer Res Treat 157 3 2016 Jun. 2016 May 20 * Impact of Oncotype DX breast Recurrence Score testing on adjuvant chemotherapy use in early breast cancer: Real world experience in Greater Manchester, UK JLoncaster AArmstrong SHowell GWilson RWelch AChittalia 10.1016/j.ejso.2016.12.010 28111076 Eur J Surg Oncol 43 5 2017 May. 2017 Jan 9. 2017 Nov 23 Eur J Surg Oncol * Changes in Recurrence Score by neoadjuvant endocrine therapy of breast cancer and their prognostic implication. ESMO Open TUeno SSaji NMasuda HIwata KKuroi NSato 10.1136/esmoopen-2018-000476.eCollection2019 30962956 PMC6435245 2019 Feb 27 4 e000476 * Cost Effectiveness of Gene Expression Profile Testing in Community Practice YChandler CBSchechter JJayasekera ANear SCO'neill CIsaacs 10.1200/JCO.2017.74.5034 29309250 PMC581540 J Clin Oncol 36 6 2018 Feb 20. 2018 Jan 8 * Breast Cancer in Argentina: Analysis from a Collaborative Group for the Study of Female Breast Cancer MeissKress RPChuit RNovellije ,Abalo ELorusso A 10.24218/jcet.2016.10 J Can Epi Treat 1 2 2016 * Can Features Evaluated in the Routine Pathologic Assessment of Lymph Node-Negative Estrogen Receptor-Positive Stage I or II Invasive Breast Cancer Be Used to Predict the Oncotype DX Recurrence Score? JAuerbach MimiKim SusanFineberg Archives of Pathology & Laboratory Medicine 134 11 2010. November 2010 * Risk score based on histopathological features predicts higher risk of distant recurrence in premenopausal patients with lymph node-negative endocrine-responsive breast cancer SDellapasqua VBagnardi MMRegan NRotmensz MGMastropasqua GViale 10.1016/j.breast.2012.06.003 22749924 PMC3566763 Breast 21 5 2012 Oct. 2012 Jun 29 * HER2 overexpression a major risk factor for recurrence in pT1a-bN0M0 breast cancer: results from a French regional cohort PRouanet PRoger ERousseau SThibault GRomieu AMathieu 10.1002/cam4.167 24407937 PMC3930398 Cancer Med 3 1 2014 Feb. 2014 Jan 10 * Use of modified Magee equations and histologic criteria to predict the Oncotype DX recurrence score BMTurner KASkinner PTang MCJackson NSoukiazian MShayne 10.1038/modpathol.2015.50 25932962 Mod Pathol 28 7 2015 Jul. 2015 May 1 * A Validated Model for Identifying Patients Unlikely to Benefit From the 21-Gene Recurrence Score MGage MRosman WMylander EGiblin HKim LCope Assay Clin Breast Cancer 2015 * 10.1016/j.clbc.2015.04.006 December 15 * Correlations Between Oncotype DX Recurrence Score and Classic Risk Factors in Early Breast Cancer: Results of A Prospective Multicenter Study in Turkey VÖzmen AAtasoy EGökmen MÖzdo?an NGüler CUras 10.5152/tjbh.2016.2874 28331745 PMC5351479 J Breast Health 12 3 2016 Jul 1. 2016 Jul * Algorithms for prediction of the Oncotype DX recurrence score using clinicopathologic data: a review and comparison using an independent dataset MRHarowicz TJRobinson MADinan ASaha JRMarks PKMarcom 10.1007/s10549-016-4093-4 28064383 PMC5909985 Breast Cancer Res Treat 162 1 2017 Feb. 2017 Jan 7 * Magee Equation 3 predicts pathologic response to neoadjuvant systemic chemotherapy in estrogen receptor positive, HER2 negative/equivocal breast tumors DJFarrugia ALandmann LZhu EJDiego RRJohnson MBonaventura 10.1038/modpathol.2017.41 28548119 Mod Pathol 30 8 2017 Aug. 2017 May 26 * Correlation of Oncotype DX Recurrence Score with Histomorphology and Immunohistochemistry in over 500 Patients MGHanna IJBleiweiss ANayak SJaffer Int J Breast Cancer 2017 1257078 2017. 2017/1257078 Epub * Nomogram update based on TAILORx clinical trial results -Oncotype DX breast cancer recurrence score can be predicted using clinicopathologic data AOrucevic JLBell MKing APMcnabb REHeidel 10.1016/j.breast.2019.05.006 31146185 Breast 46 2019 Aug. 2019 May 10 * Breast Cancer Thirty Years Later: A Comparative Study between A 1983-1984 AND A 2012-2013 Cohorts of Argentine Women MeissKress RPNovelli JE AbaloELorusso AGualtieri A 10.24218/jcet.2017.13 J Can Epi Treat 1 3 2017 * Breast Cancer in Argentina: Feasibility for the Implementation of the New TNM Staging System in A Middle-Income Country MeissKress RPNovelli JEGago FERobles MMorales SEt La 10.24218/jcet.2018.20 J Can Epi Treat 2 1 2018 * Demographic changes in breast cancer incidence, stage at diagnosis and age associated with population-based mammographic screening FCVerdial REtzioni CDuggan BOAnderson 28194807 PMC5701282 J Surg Oncol 115 5 2017 Apr. 2017 Feb 14 * Burden of early, advanced and metastatic breast cancer in The Netherlands GTVondeling GLMenezes EPDvortsin FgaJansman IRKonings MJPostma MHRozenbaum 10.1186/s12885-018-4158-3 29514651 PMC5842550 BMC Cancer 18 1 262 2018 Mar 7 * Trends in stage-specific breast cancer incidence in New South Wales, Australia: insights into the effects of 25 years of screening mammography GJacklyn KMcgeechan LIrwig NHoussami SMorrell KBell ABarratt 10.1007/s10549-017-4443 28822001 Breast Cancer Res Treat 166 3 2017 Dec. 2017 Aug 19 * Use of Biomarkers to Guide Decisions on Adjuvant Systemic Therapy for Women With Early-Stage Invasive Breast Cancer: American Society of Clinical Oncology Clinical Practice Guideline LNHarris NIsmaila LMMcshane FAndre DECollyar AMGonzalez-Angulo 10.1200/JCO.2015.65.2289 26858339 J Clin Oncol 34 10 2016 Apr 1. 2016 Feb 8 * PubMed Central PMC4933134 * Value of Information Analysis of Multiparameter Tests for Chemotherapy in Early Breast Cancer: The OPTIMA PSHall ASmith CHulme AVargas-Palacios AMakris LHughes-Davies * 10.1016/j.jval.2017.04.021 29241890 2017 Dec. 2017 Jul 11 20 * The influence of 21-gene recurrence score assay on chemotherapy use in a population-based sample of breast cancer patients YLi AWKurian IBondarenko JmgTaylor RJagsi KCWard ASHamilton SJKatz TPHofer 10.1007/s10549-016-4086-3 28012085 PMC5243200 Breast Cancer Res Treat 161 3 2017 Feb. 2016 Dec 23 * Cost of breast cancer based on real-world data: a cancer registry study in Italy SCapri ARusso 28122558 PMC5267401 BMC Health Serv Res 17 1 84 2017 Jan 26. 1186/s12913-017-2006-9 * Global treatment costs of breast cancer by stage: A systematic review LSun RLegood Dos-Santos-Silva IGaiha SMSadique Z 10.1371/journal.pone.0207993 30475890 PMC6258130 PLoS One 13 11 e0207993 2018 Nov 26. 2018 * Healthcare Utilization and Costs During the Initial Phase of Care Among Elderly Womenwith Breast Cancer AVyas SSMadhavan USambamoorthi XLPan MRegier HHazard SKalidindi 10.6004/jnccn.2017.0167 29118 PMC5817990 J Natl ComprCancNetw 15 11 2017 Nov * Annual Report 2018 * Clinical relevance of the 21-gene Recurrence Score(®) assay in treatment decisions for patients with node-positive breast cancer in the genomic era EPMamounas CARussell ALau MPTurner KSAlbain 10.1038/s41523-018-0082-6.eCollection 30155517 PMC6102296 NPJ Breast Cancer 4 27 2018 Aug 20. 2018 Review * Practical Consequences Resulting from the Analysis of a 21-Multigene Array in the Interdisciplinary Conference of a Breast Cancer Center HUVoelker LFrey AStrehl MWeigel Int J Breast Cancer 2018 * 10.1155/2018/2047089 30112216 Jul 2047089 10. 2018. 2018 * PubMed Central PMC6077570 * Recent Trends in Chemotherapy Use and Oncologists' Treatment Recommendations for Early-Stage Breast Cancer AWKurian IBondarenko RJagsi CRFriese MCMcleod STHawley 10.1093/jnci/djx239 29237009 J Natl Cancer Inst 110 5 2018 May 1 * PubMed Central PMC5946952 * The Distribution and Outcomes of the 21-Gene Recurrence Score in T1-T2N0 Estrogen Receptor-Positive Breast Cancer with Different Histologic Subtypes. Front Genet JWang ZYHe YDong JYSun WWZhang SGWu 10.3389/fgene.2018.00638.eCollection 2018 Dec 17. 2018 9 638 * Pubmed 30619463 PMC6304349