Precise Personalized Medicine in Gynecology Cancer and Infertility.

P-YZ drafted the manuscript. YY conceived the idea and revised the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Reproductive medicine has already accomplished a lot, but there is still a long road ahead to truly achieve precision medicine. There are difficulties in transforming the research results of genetics and biology into clinical application. Precision medicine might be able to fill this gap between basic research and clinical application. Apart from the urgency of increasing the success rates and reducing the cost, reproductive medicine needs to be more elucidative. Complex diseases involve many variants, and identifying these determinants is a challenge for reproductive medicine. Mendelian diseases are controlled by a single locus, and the phenotype is easy to predict. Nevertheless, Mendelian diseases can only explain a small fraction of infertility causes, and a fairly large number of infertility causes cannot be explained by one single gene’s mutation. Many couples have received extensive diagnostic testing but never received a definitive diagnosis for their reproductive difficulties ( Luke et al., 2012 ). The diagnosis of many causes of infertility is descriptive, such as recurrent pregnancy loss (RPL), decline in ovarian reserve (DOR), and POI. The mechanisms underlying these diseases are not clear. These uncertainties cause patients to have difficulties going through these time-intensive and emotionally taxing treatment cycles. We need to understand the mechanisms underlying symptoms and offer patients genetic-based tests and specific diagnoses and personalized treatments. We started noticing disease heterogeneity in oncology at the very beginning. Cancers are divided by organs and then subdivided by tumor characteristic molecules and genotypes, and targeted treatment is carried out based on this information, e.g., herceptin treatment is carried out in patients with HER2 overexpression. However, not enough attention is paid to disease heterogeneity in reproductive medicine’s diagnosis and treatment, as in oncology. Only 50% of the cases have a definitive diagnosis, and the definitive diagnosis for these 50% of cases is still heterogeneous. They might have disparate etiologies with distinct biological and molecular bases ( Beim et al., 2013 ). Multigene panel tests might reveal the mechanisms and lead us to a better understanding. For example, a multigene panel test can be used to assess ovarian function. Interventions can be offered on the basis of the multigene panel test results and clinical evidence ( Noyes et al., 2011 ). Since genes at high risk and genes of lesser magnitude are all tested, multigene panel tests can reduce the likelihood of false-negative assessment and are cost-effective and time-efficient in oncology. As an ideal goal of reproductive medicine, identifying pathogenic genetic variants for patients with specific infertility phenotypes, such as ovarian dysgenesis or azoospermia, can be realized with focused gene panels (such as those offered by Centogene and Evolve Gene). Moreover, screening rare mutations that directly cause overt infertility, such as impaired oocyte maturation and fertilization defects, is also required. In 2017, Celmatix Inc. announced the first commercially available product, Fertilome, in the United States. Celmatix Lab obtained 49 specific single-nucleotide variants in 32 genes by masked analysis of targeted NGS, and Fertilome can offer services to patients with these data, which have been implicated in a variety of reproductive conditions, such as POI, RPL, PCOS, and endometriosis ( Celmatix Inc., 2017 ). While the power of multigene panel testing depends on the amount of genetic information it provides, this level of gene information carries some inherent downsides. Identification of variants of unknown significance (VUS) can confuse clinicians and patients, for which there is inadequate evidence (e.g., from family or population studies, prior instances of the variant, in vitro or in vivo functional studies, or in silico predictions of function) to remark on its pathogenicity. We lack a thorough understanding of all the genes we test, and we might have difficulties explaining what the test results mean to patients. Some positive results of less relevant genes may cause anxiety for patients if clinicians cannot explain it properly. A close scrutiny of the literature in which risk alleles have been identified is needed to ensure clinical validity. Advancements in high-throughput omics technologies allow the generation of large volumes of biomedical data that can be usefully exploited for precision medicine. The big data and developing artificial algorithms provide the basis for novel insights into diseases that can be translated into personalized treatments ( Krawetz, 2017 ). Big data can offer help in randomized controlled trials (RCTs) and provide clinical evidence of high quality. In the field of reproductive medicine, big data can enable us to obtain more patient information, which is conducive to drug development, remote monitoring, and providing a lot of information for the exploration of the causes of infertility diseases. Genetic disease susceptibility genes have ethnic specificity, and big data can be used to find the specific susceptibility genes of the right population. Big data could help clinicians develop more specific guidelines for different situations and make more personalized determinations for patients. As discussed above, reproductive medicine currently still has a low success rate owing to many patients’ explicit diagnoses. For patients without a defined diagnosis, intrauterine insemination (IUI) and IVF are still considered as treatment. Theoretically, these interventions should be given only if the expected success rates exceed the probabilities of spontaneous pregnancy. However, it is not easy for gynecologists to evaluate when objective criteria are lacking. Thus, prediction models are needed to help clinicians precisely estimate the success rate and make decisions that are in the best interest of the patient ( Leushuis et al., 2009 ). Clinical studies have identified some predictors for the success rate of assisted reproduction technologies. Serum hormone levels including estradiol, FSH, and anti-Müllerian hormone, history of gestation, age and many other indicators are all strong predictors and are widely used in clinical decision and success rate predictions. These data can be integrated, and statistical and machine learning techniques are applied to derive an algorithm combining the clinical factors and yield a predicted likelihood of success for each patient, therefore making reproductive medicine more precise and individualized ( Mayer-Schönberger and Ingelsson, 2018 ). Some clinical factor models based on logistic regression or Cox models are published but cannot be applied because of low currency. The algorithms need to be rectified, and iterative, adaptive, and learning models are needed ( Leushuis et al., 2009 ). This algorithm can be repeatedly recalibrated by performing calculations on patients until its prediction coincides with real models ( Beim et al., 2013 ). Big data promise huge benefits for medical research. Apart from the prediction models, big data can offer us new insights in medical exploration. Unlike conventional analysis of data samples, big data can help us find some factors that we did not consider as causes for infertility. Data are captured more comprehensively relative to the phenomenon under study. This type of data collection reduces some bias and brings to the surface important trade-offs between data quantity and data quality. Data are analyzed by machine learning tools, such as neural networks, rather than conventional statistical methods. This learning system can capture insights implicit in data, but it rarely reveals causal connections, which remain as black boxes. The purpose of the analysis of data is no longer simply answering existing questions but hinting at novel ones and generating promising new hypotheses ( Leushuis et al., 2009 ; Mayer-Schönberger and Ingelsson, 2018 ). Based on the characteristics of the big data mentioned above, we can note new factors related to the field of reproductive medicine and explore new directions for scientific research. Big data can narrow down the range of factors that need to be studied and help make medicine research more precise ( Figure 2 ). Comparison of conventional statistical methods and big data analysis. Targeted molecular medicine has been widely studied in oncology, and many targeted drugs have already been applied clinically. Many drugs have been developed; however, not many have clinical benefits. In addition, drug resistance is an issue that cannot be ignored. Since the first biotechnology-derived pharmaceutical recombinant human insulin was introduced and licensed for human use in 1982, huge progress has been made in molecular biology, functional genomics, and proteomics, and the number of large molecule biopharmaceuticals for human use has rapidly increased. Unlike conventional small-molecule drugs that are chemically synthesized, biopharmaceuticals cannot be tested in standard processes. Biopharmaceuticals are derived from living sources, such as humans, animals, plants, and microorganisms, using manufacturing processes that are often complex. Most biopharmaceuticals are large molecules or heterogeneic mixtures of large molecules that are not easily characterized, often display species specificity and pleiotropism in their pharmacologic effects, are immunogenic, and tend to be heat-sensitive and susceptible to microbial contamination ( Brennan et al., 2004 ). Identification of effective combination therapies is critical to address the emergence of drug-resistant cancers, and many new screening techniques have been developed. The CRISPR-based double knockout (CDKO) system provides new possibilities for cancer medicine screening. This screening method combines genomic information so that it can proceed in a precise way ( Han et al., 2017 ). For reproductive medicine, apart from the targeted drugs for gynecological cancer, many causes of infertility, such as PCOS and endometriosis, are associated with molecular mechanisms, and some drugs have been developed addressing these mechanisms. Similar to anastrozole and letrozole, third-generation aromatase inhibitors, which selectively inhibit the action of aromatase, have been investigated for the treatment of endometriosis as monotherapy or in combination with other hormonally active drugs ( Ferrero et al., 2018 ). Atosiban, a mixed vasopressin V1a and oxytocin receptor antagonist, has been registered for the treatment of imminent premature birth with minimal side effects. Now atosiban and its advanced version barusiban can reduce uterine contractions, and increase the clinical PR per cycle and IR per transfer ( Blockeel et al., 2009 ; Visnova et al., 2009 ; Moraloglu et al., 2010 ). Besides, efforts on targeted drug delivery system also bring new angle of precision medicine. Mesoporous silica, a kind of targeted nanovectors for biological delivery facilitate the precise transport of large amounts of any type of compound to specific destinations, followed by rapid internalization via physiological uptake mechanisms and the ultimate targeting of specific intracellular pathways ( Barkalina et al., 2014 ). Other drugs are being developed for different molecular mechanisms, and gene editing technologies can also be used to screen drugs for effectiveness and personal fitness. However, targeted medicine are not well developed in the area of reproductive medicine like oncology does, even though molecular mechanisms and genomic variations of reproductive diseases are widely studied and explained. We hope cancer drug screening system can be applied in the field of reproductive medicine and search for more targeted drugs aims at genetic variations and signaling pathways. The research on stem cells brings us a promising future wherein stem cell therapy can be possible. For some patients seeking the help of ART, there is the problem of them having no functional gametes. Donor gametes can solve the problem but cannot be accepted by every couple, especially those people who want their genetically own offspring. The emergence of induced pluripotent stem cells (iPSCs) ( Takahashi and Yamanaka, 2006 ; Takahashi et al., 2007 ) brings new solutions to this problem. Gametes induced by autologous cells offer possibilities for people without functional gametes. However, the safety is uncertain, and it cannot be tested in conventional procedures because iPSCs induced from one person’s somatic cells can only be tested on himself/herself ( Vassena et al., 2015 ). A case report about a patient treated with retinal pigment epithelium derived from human embryonic stem cells indicates that stem cell therapy is on the brink of being applied in other medical specialties ( Schwartz et al., 2012 ). There are no stem cell-based therapies available to the larger public or outside of clinical trials directed at ameliorating or solving reproductive medicine issues yet. However, we believe that this day will come.

Starting from oncology, precision medicine has already permeated into various fields of medicine. For reproductive medicine, precision has always been a criterion in every procedure, including etiology-oriented examination, specific diagnosis, identifying healthy embryos, WOI, and accurate implantation. Combined with genetic information and a large volume of biomedical data, an unknown territory of reproductive medicine will be explored, and the mechanisms underlying the causes of infertility that we do not yet know will be elucidated. The application of precision medicine has become a guideline for the development of medicine, especially for reproductive medicine. In many other specialties, more and more available clinical and genetic biomarkers have helped physicians make overall disease prediction and have informed clinical decision making. Chronic diseases such as cancer, cardiovascular disease, and type-2 diabetes place a heavy burden on society, and the health care model is gradually evolving into a proactive system. A 4P health care model (Predictive, Preventive, Personalized and Participatory) has been put forward and has been increasingly popularized in the care of chronic disease. The outcome for a breast cancer patient can be predicted by the expression of HER2 and ER; likewise, can the risk of infertility and success rate of ART be predicted? Can this 4P model be applied to reproductive medicine? Can the application of the 4P health care model make reproductive medicine more precise? We believe the answers are affirmative. Regarding prediction, avoiding the accumulation of stressors and identifying early symptoms are primary objectives. For reproductive medicine, current biomarkers such as hormone levels and ultrasound examination are still at the core of predicting disease. The next generation of biomarkers combined with big data technology will increase the precision of identifying dysfunctions. Information on the genome, epigenetic groups, transcriptome, proteome, metabolome, and molecular diagnosis can be integrated, and based on this information, screening and identification platforms can be established. For prevention, the essence is avoiding the causes of infertility. Among all the causes of infertility, some are related to an unhealthy lifestyle. Strengthening education on female reproductive health will be a strong prevention for infertility. With our deepening understanding of different causes of infertility, more personalized diagnoses can be made, and more precise treatment can be given, and in this way, the success rate of ART can be improved. Meanwhile, it is important to strengthen the psychological treatment and education of patients to overcome the emotional barriers and prevent dropout caused by a lack of awareness. Involving the individual in personalized treatment and preventive interventions and improved data collection through self-tracking will be important factors. To achieve this goal, we have to gain a much better understanding of an individual’s “health literacy” ( Sagner et al., 2017 ). With the continued promotion of precision medicine, reproductive medicine will reach new heights and bring hope for more patients who are suffering from infertility.

Infertility is defined as the inability to achieve clinical pregnancy after 1 year of regular unprotected intercourse and can affect up to 14% of couples of reproductive age ( Zegers-Hochschild et al., 2009 ). Reproductive medicine brings good news to many infertile couples. The first test-tube baby, Louise Brown, was born in 1978, and great progress has been made in improving the success rates and efficiency of fertility treatments over the past four decades. A number of new in vitro fertilization (IVF) practical regimens have been applied in different regions, and their viabilities still need to be evaluated. For example, embryo cryopreservation and subsequent thawed embryo transfer have replaced fresh embryo transfer ( Roque et al., 2013 ), blastocyst-stage transfer has replaced cleavage-stage embryo transfer ( Blake et al., 2007 ), single-embryo transfer (SET) has replaced double embryo transfer (DET) ( Gelbaya et al., 2010 ), and preimplantation genetic screening (PGS) ( Schoolcraft and Katz-Jaffe, 2013 ) and minimal stimulation protocols ( Teramoto and Kato, 2007 ; Kato et al., 2012 ) have been applied. Still, there is a lot of room for improvement of reproductive medicine, and it is required to be more cost-effective and efficient. In the United States, IVF treatments fail 70% of the time on a per cycle basis and cost $12,400 per treatment cycle on average ( Figure 1 ; Centers for Disease Control and Prevention, 2011 ). It is estimated that approximately one in seven couples is struggling with infertility, but only a small fraction of these individuals could receive assisted reproductive technologies (ART) treatment ( Beim et al., 2013 ). The number of patients who have received ART is increasing annually. An analysis focused on the development trend of assisted reproductive technology in Liaoning province from 2012 to 2016 shows that the number of fresh cycles and thawed cycles increased onefold. The success rate has been elevated, but it is still approximately 40%, which is consistent with the data provided from other regions. In summary, efficiency is a problem that urgently needs to be solved, and reproductive medicine should be more precise ( Fang et al., 2018 ). Outcomes of ART cycles using fresh non-donor eggs or embryos, 2015 (United States). The total does not equal 100% due to rounding. After experiencing evidence-based medicine (EBM), we have entered the era of precision medicine. EBM emphasizes the relationship between clinical research and clinicians, but the gap between clinicians and patients has not been filled. We have gradually realized that we need a new health care model that can adapt to various patients. The conception of this new health care model of precision medicine is beyond EBM. This conception started from oncology, which refers to selecting chemotherapeutic agents based on patients’ genetic profiles to maximize the efficiency and safety. Currently, the conception of precision medicine has been employed in many areas of medicine, such as psychiatry and cardiovascular disease. Genetic profiles have been identified that predict the necessary therapeutic dose for opioid analgesia and the amount of heart disease risk reduction a patient will gain from statins ( Krawetz, 2017 ; Smith et al., 2017 ). According to the definition promulgated by the National Institutes of Health (NIH), precision medicine refers to a thorough understanding of the consequences of unique features of individual patients, such as environmental exposures, lifestyle, and genetic profiles, and requires new individualized treatments and prevention methods. The goal of precision medicine is a better understanding and identification of patients who are at risk for particular diseases. It requires refinement of screening guidelines, to start early intervention and prevent unwanted outcomes and to urge potential patients to change their lifestyle and carefully consider overall life plans ( Chang and Lee, 2018 ). In the realm of reproductive medicine, this conception refers to better information collection of health history, family history, lifestyle, and genotype, and customized treatment protocols are made based on this information. Instead of treating all patients with the same procedures, this type of personalized health model can increase the efficacy and efficiency of ART. Precision medicine has already been widely accepted and implicated in reproductive medicine. However, reproductive medicine still faces severe challenges. The number of people facing fertility disorder is increasing gradually, and the reasons for fertility disorder are various. Reproductive medicine should be more precise in diagnosis, healthy embryo identification, treatment processes, and many other aspects. This review discusses the current application of precision medicine and the potential that exists for reproductive medicine.