BACKGROUND: Universal tumor testing for defective DNA mismatch repair (MMR) is recommended for all women diagnosed with endometrial cancer to identify those with underlying Lynch syndrome. However, the effectiveness of these screening methods in identifying individuals with Lynch syndrome across the population has not been well studied. The aim of this study was to evaluate outcomes of MMR immunohistochemistry (IHC), mutL homolog 1 (MLH1) methylation, and microsatellite instability (MSI) analysis among patients with endometrial cancer. METHODS: A complete systematic search of online databases (PubMed, EMBASE, MEDLINE, and the Cochrane Library) for 1990-2018 was performed. A DerSimonian-Laird random effects model meta-analysis was used to estimate the weighted prevalence of Lynch syndrome diagnoses. RESULTS: The comprehensive search produced 4400 publications. Twenty-nine peer-reviewed studies met the inclusion criteria. Patients with endometrial cancer (n = 6649) were identified, and 206 (3%) were confirmed to have Lynch syndrome through germline genetic testing after positive universal tumor molecular screening. Among 5917 patients who underwent tumor IHC, 28% had abnormal staining. Among 3140 patients who underwent MSI analysis, 31% had MSI. Among patients with endometrial cancer, the weighted prevalence of Lynch syndrome germline mutations was 15% (95% confidence interval [CI], 11%-18%) with deficient IHC staining and 19% (95% CI, 13%-26%) with a positive MSI analysis. Among 1159 patients who exhibited a loss of MLH1 staining, 143 (13.7%) were found to be MLH1 methylation-negative among those who underwent methylation testing, and 32 demonstrated a germline MLH1 mutation (2.8% of all absent MLH1 staining cases and 22.4% of all MLH1 methylation-negative cases). Forty-three percent of patients with endometrial cancer who were diagnosed with Lynch syndrome via tumor typing would have been missed by family history-based screening alone. CONCLUSIONS: Despite the widespread implementation of universal tumor testing in endometrial cancer, data regarding testing results remain limited. This study provides predictive values that will help practitioners to evaluate abnormal results in the context of Lynch syndrome and aid them in patient counseling. Cancer 2019;125:3172-3183.

STUDY QUESTION Does ovarian stimulation affect embryo euploidy rates or live birth rates (LBRs) after transfer of euploid embryos? SUMMARY ANSWER Euploidy rates and LBRs after transfer of euploid embryos are not significantly influenced by gonadotropin dosage, duration of ovarian stimulation, estradiol level, follicle size at ovulation trigger or number of oocytes retrieved, regardless of a woman’s age. WHAT IS KNOWN ALREADY Aneuploidy rates increase steadily with age, reaching >80% in women >42 years old. The goal of ovarian stimulation is to overcome this high aneuploidy rate through the recruitment of several follicles, which increases the likelihood of obtaining a euploid embryo that results in a healthy conceptus. However, several studies have suggested that a high response to stimulation might be embryotoxic and/or increase aneuploidy rates by enhancing abnormal segregation of chromosomes during meiosis. Furthermore, a recent study demonstrated a remarkable difference in euploidy rates, ranging from 39.5 to 82.5%, among young oocyte donors in 42 fertility centres, potentially suggesting an iatrogenic etiology resulting from different stimulation methods. STUDY DESIGN, SIZE, DURATION This is a retrospective cohort study that included 2230 in vitro fertilisation (IVF) with preimplantation genetic testing for aneuploidy (PGT-A) cycles and 930 frozen-thawed single euploid embryo transfer (FET) cycles, performed in our centre between 2013 and 2017. PARTICIPANTS/MATERIALS, SETTING, METHODS A total of 12 298 embryos were analysed for ploidy status. Women were divided into five age groups (<35, 35–37, 38–40, 41–42 and >42 years old). Outcomes were compared between different durations of stimulation (<10, 10–12 and ≥13 days), total gonadotropin dosages (<4000, 4000–6000 and >6000 IU), numbers of oocytes retrieved (<10, 10–19 and ≥20 oocytes), peak estradiol levels (<2000, 2000–3000 and >3000 pg/mL), and sizes of the largest follicle on the day of trigger (<20 and ≥20 mm). MAIN RESULTS AND THE ROLE OF CHANCE Within the same age group, both euploidy rates and LBRs were comparable between cycles regardless of their differences in total gonadotropin dosage, duration of stimulation, number of oocytes harvested, size of the largest follicles or peak estradiol levels. In the youngest group, (<35 years, n = 3469 embryos), euploidy rates were comparable between cycles with various total gonadotropin dosages (55.6% for <4000 IU, 52.9% for 4000–6000 IU and 62.3% for >6000 IU; P = 0.3), durations of stimulation (54.4% for <10 days, 55.2% for 10–12 days and 60.9% for >12 days; P = 0.2), number of oocytes harvested (59.4% for <10 oocytes, 55.2% for 10–19 oocytes and 53.4% for ≥20 oocytes; P = 0.2), peak estradiol levels (55.7% for E2 < 2000 pg/mL, 55.4% for E2 2000–3000 pg/mL and 54.8% for E2 > 3000 pg/mL; P = 0.9) and sizes of the largest follicle (55.6% for follicles <20 mm and 55.1% for follicles ≥20 mm; P = 0.8). Similarly, in the oldest group (>42 years, n = 1157 embryos), euploidy rates ranged from 8.7% for gonadotropins <4000 IU to 5.1% for gonadotropins >6000 IU (P = 0.3), from 10.8% for <10 days of stimulation to 8.5% for >12 days of stimulation (P = 0.3), from 7.3% for <10 oocytes to 7.4% for ≥20 oocytes (P = 0.4), from 8.8% for E2 < 2000 pg/mL to 7.5% for E2 > 3000 pg/mL (P = 0.8) and from 8.2% for the largest follicle <20 mm to 8.9% for ≥20 mm (P = 0.7). LBRs after single FET were also comparable between these groups. LIMITATIONS, REASONS FOR CAUTION Although this large study (2230 IVF/PGT-A cycles, 12 298 embryos and 930 single FET cycles) demonstrates the safety of ovarian stimulation in terms of aneuploidy and implantation potential of euploid embryos, a multi-centre study may help to prove the generalisability of our single-centre data. WIDER IMPLICATIONS OF THE FINDINGS These findings reassure providers and patients that gonadotropin dosage, duration of ovarian stimulation, estradiol level, follicle size at ovulation trigger and number of oocytes retrieved, within certain ranges, do not appear to significantly influence euploidy rates or LBRs, regardless of the woman’s age. STUDY FUNDING/COMPETING INTEREST(S) No external funding was received and there are no competing interests to declare. TRIAL REGISTRATION NUMBER N/A

The inclusion of DNA mismatch repair (MMR) evaluation as a standard of care for endometrial cancer management will result in a growing population of patients with MMR deficiency and negative germline Lynch syndrome testing (MMR-deficient). In this systematic review and study, the clinicopathologic features of endometrial cancer in patients with MMR-intact, MLH1 methylation positive, MMR-deficient or Lynch syndrome are evaluated. A systematic search of online databases between 1990 and 2018 identified studies of endometrial cancer patients with tumour testing (MMR protein immunohistochemistry or microsatellite instability) and germline assessment for Lynch syndrome. Extracted data included tumour testing, germline genetic testing, age, body mass index (BMI), family history, tumour stage, grade and histologic type. Associations between MMR-intact, MLH1 methylation positive, MMR-deficient and Lynch syndrome groups were analysed using descriptive statistics. The comprehensive search produced 4,400 publications, 29 met inclusion criteria. A total of 7,057 endometrial cancer cases were identified, 1,612 with abnormal immunohistochemistry, 977 with microsatellite instability. Nine-hundred patients underwent germline genetic testing, identifying 212 patients with Lynch syndrome. Patients in the Lynch syndrome and MMR-deficient groups were significantly younger than patients in the MMR-intact and MLH1 methylation positive groups. Patients with MMR-intact tumours had the highest BMI, followed by MMRdeficient, then Lynch syndrome. MMR-intact tumours were more likely to be grade I at diagnosis than other groups. Patients with Lynch syndrome and MMR-deficient tumours were less likely to have stage I disease as compared to patients with MMR-intact tumours. Endometrial cancer patients with MMR-deficient tumours have similar features to those with germline Lynch syndrome mutations, including age, grade, histology and stage. Even in the absence of a germline mutation, tumour evaluation for MMR status may have important clinical implications.

(Abstracted from Cancer 2019;125(18):3172–3183) Lynch syndrome caused by germline mutations in several DNA mismatch repair (MMR) proteins is the most common cause of hereditary and endometrial cancers. Lynch syndrome accounts for approximately 2% to 3% of all endometrial cancers, although the lifetime risk of endometrial cancer for patients with Lynch syndrome varies depending on the specific mutation.

e17119 Background: Universal tumor testing for defective DNA mismatch repair (MMR) is recommended for all women diagnosed with endometrial cancer (EC) to identify those with underlying Lynch syndrome (LS). However, since its implementation in 2013, the effectiveness of this screening method on identifying individuals with LS across the population has not been well studied. The aim of this study was to evaluate outcomes of MMR immunohistochemistry (IHC), MLH1 methylation, and microsatellite instability (MSI) analysis among EC patients. Methods: We conducted a complete systematic search of online databases PubMed, Embase, Medline, and Cochrane Library between 1990-2018. A DerSimonian–Laird random-effects model meta-analysis was utilized to estimate the weighted prevalence of LS diagnoses. Results: The comprehensive search produced 3,427 publications. 29 peer-review studies met the inclusion criteria. 6,649 EC patients were identified, 206 (3%) were confirmed to have LS following positive universal tumor molecular screening.5,917 patients underwent tumor IHC, 28% had abnormal staining. 3,140 patients underwent MSI analysis, 31% had MSI instability. Among EC patients with deficient IHC staining or positive MSI analysis, the weighted prevalence of LS was 15% and 19% respectively. 1159 patients exhibited loss of MLH1 staining, 143 (13.7%) were found to be MLH1 methylation negative, 32 demonstrated a germline MLH1 mutation (2.8% of all MLH1 absent staining; 22.4% of all MLH1 methylation negative). 43% of EC patients diagnosed with LS via tumor typing would have been missed by family history-based screening alone. Conclusions: Despite widespread implementation of universal tumor testing in EC, data regarding results have previously been limited. For the first time, this study provides large-scale predictive values that will help practitioners evaluate abnormal results in the context of LS and aid in patient counseling. [Table: see text]