Health-related quality of life after risk-reducing hysterectomy (PRESCORES): a cross-sectional study Authors: Samuel George Oxley, Milly Lake, Amanda Dibden, Tia Whenham, Xia Wei, Lea Mansour, Ashwin Kalra, Subhasheenee Ganesan, Jacqueline Sia, Michail Sideris, Caitlin Fierheller, Margarida Kurzer, Rachel Perfect-Lake, Jyotsna Pundir, Ruth Armstrong, Janos Balega, Ishita Bhatnagar, Michela Cicchetti, Cecilia Compton, Jacqueline Cook, Rosemarie Davidson, Lesa Dawson, Sadaf Ghaem-Maghami, Helen Hanson, Rachel Harrison, Rachel Hart, Alison Kraus, Fiona Lalloo, D Gareth Evans, Terri P McVeigh, Kevin Monahan, Laura Monje-Garcia, Esther Moss, Alex Murray, Kai Ren Ong, Myfanwy Rawson, Dimitra Repana, Adam N Rosenthal, Neil Ryan, Lyndall Sarkies Published: 2026 ABSTRACT SummaryBackgroundUnderstanding the impact of risk-reducing hysterectomy (RRH) on health-related quality-of-life is essential for informed decision-making around endometrial cancer prevention and has not been previously investigated. This study aimed to estimate health-related utilities in women who received RRH.MethodsPRESCORES was a multicentre cross-sectional study (ISRCTN17432105). We invited 4031 UK women ≥18 years with confirmed Lynch Syndrome, without prior gynaecological cancer, to participate between May 2023 and November 2024. The main outcome was a complete-case adjusted mean-difference in EQ-5D utility scores between participants who had undergone RRH or not, and by pre/post-menopausal age groups. Quality-of-life was measured at the time of survey completion, and retrospectively 4 months after surgery in the RRH-group. The Cancer–Worry Scale instrument was also evaluated by RRH group. Sensitivity analysis considered alternative statistical methods, multiple imputation, and restricting analysis to people 35–80 years.FindingsAnalysis included 1120 participants: 529 had undergone RRH with a median age of 55 years (IQR 48–62), compared with 39 years (33–52) without RRH. The RRH-group was significantly more likely to have had cancer, physical co-morbidities, be parous and used aspirin. Adjusted mean utilities were slightly lower in the RRH than no-RRH group (difference −0.020, 95% CI −0.045 to 0.004; p = 0.11), much lower for premenopausal women (−0.041, −0.074 to −0.008), but similar for postmenopausal women (0.003, −0.036 to 0.042). Cancer-Worry-Scale scores were lower in the RRH group (mean-difference 0.48, 95% CI −0.807 to −0.157; p = 0.0037). Interpretation was robust to sensitivity analyses.InterpretationHealth-related quality-of-life in women with Lynch Syndrome is likely to be lower following RRH in younger, pre-menopausal women. This is relevant for counselling, RRH decision making, and economic evaluations of women with Lynch Syndrome, with increased endometrial cancer risk.FundingRosetrees Trust (CF1\100001) and Barts Charity (G-002286 and MRC0167).Keywords: Lynch, Utilities, Risk-reducing, Hysterectomy, Quality-of-life, MMR Research in context Evidence before this study We searched PubMed, Medline & Embase from inception to January 2026, using the terms “quality of life”, “utilities”, “hysterectomy”, “risk-reduction”, “Lynch”, including synonyms and variant spellings. No previous studies have estimated the health-related quality-of-life following risk-reducing hysterectomy for endometrial cancer prevention, using EQ-5D or any other validated questionnaires. Studies which used qualitative interviews or non-validated questionnaires described a reduction in cancer worry and high satisfaction with low decision regret. Added value of this study This study is the first to report quality-of-life in women with Lynch Syndrome who have undergone risk-reducing hysterectomy and those who have not, and the first to report health utilities. It compared those who have and have not undergone risk-reducing hysterectomy to estimate the impact of surgery on quality-of-life. It recruited a large number of women with confirmed Lynch Syndrome from across the UK, and used an internationally validated and recommended tool (EQ-5D). This found that women younger than 52 years with Lynch syndrome who had risk-reducing hysterectomy report a lower quality-of-life than premenopausal women without hysterectomy, but there was no difference in older women. Implications of all the available evidence This is relevant for women with Lynch Syndrome, and may be relevant for other women at increased endometrial cancer risk, who are considering risk-reducing hysterectomy or other risk-management approaches. This information can inform counselling and shared decision-making on endometrial cancer prevention strategies, which should address not only the preventive benefit of surgery but also the potential quality-of-life trade-offs. It is also important for health economic evaluations internationally. Introduction Endometrial cancer is the most common gynaecological cancer in developed countries affecting ∼3% of women,1 and incidence is projected to increase by 53% by 2040.2,3 Lynch Syndrome affects 1 in 280–370 people,4,5 and accounts for 3% of endometrial cancer cases,6 predisposing carriers to 13–49% lifetime-risk of endometrial cancer, along-with an increased risk of ovarian and colorectal cancer and other malignancies.7 Although only an estimated <5% of carriers have been identified,5 this will likely increase with the implementation of routine tumour and targeted-germline screening at endometrial and colorectal cancer diagnosis8 and prospective trials of unselected population genetic-testing.9 Beyond Lynch Syndrome, many more women may be identified at elevated endometrial cancer-risk with the advent of validated personalised endometrial cancer risk-prediction tools.10 For women at increased endometrial cancer-risk, risk-reducing hysterectomy (RRH) is the gold-standard preventive strategy which reduces endometrial cancer-risk, likely with ∼100% efficacy.11 Gynaecological surveillance with annual transvaginal ultrasound and/or endometrial biopsy may be used as an interim strategy, however there is no high-quality evidence that surveillance improves endometrial cancer outcomes.12,13 This is recommended to Lynch-carriers from 35 to 40 years after completing childbearing, usually with bilateral salpingo-oophorectomy (BSO) to reduce the concomitant ovarian cancer-risk.14,15 RRH has potential surgical complications and cessation of natural fertility.16 Premenopausal RRH-BSO leads to surgical menopause, for which hormone-replacement-therapy (HRT) helps treat vasomotor and other symptoms, and avoid long-term detrimental sequaelae.17 The decision for risk-reducing surgery is complex, guided by perceived cancer risk, perceived treatment benefit, fertility wishes, long-term health consequences, and other psychosocial factors.18,19 Patients must weigh the benefits of cancer prevention against the potential impact on quality-of-life (QoL).20 An understanding of QoL after RRH is therefore highly relevant for counselling and decision-making around RRH, and essential for health-economic evaluations of endometrial cancer prevention, which inform clinical guidelines. These require QoL ratings in the form of utilities, which represent QoL on a spectrum from ‘1’ being perfect QoL to ‘0’ being death. No studies have reported utilities following RRH, despite its need, including for economic evaluations,21 and being identified as a research priority by policy-makers like National-Institute of Health-and Care-Excellence (NICE).22 As women with Lynch Syndrome are the major patient-group who currently undergo RRH, any direct assessment of QoL following RRH must involve such patients. Given that a randomised controlled trial of risk-reducing surgery versus control/surveillance would be unethical, and unacceptable to high-risk patients, any study must be non-randomised. A prospective study would take numerous years to recruit sufficient numbers to enable a comparison of RRH versus no RRH groups and require substantial funding, and therefore a cross-sectional retrospective study is required to provide an estimation of utilities within a reasonable timeframe. This study aims to determine the health-related utility scores for risk-reducing hysterectomy for endometrial cancer prevention, overall and separately for pre- and post-menopausal patients. Methods Study design and participants PRESCORES (ISRCTN17432105) was a multi-centre cross-sectional study of women with Lynch syndrome. Eligible adult UK-women (≥18 years) were proven Lynch Syndrome pathogenic-variant carriers, who could understand written English, and had no history of endometrial, ovarian or cervical cancer (given their treatment usually precludes RRH). Participants were identified by 25 National-Health-Service (NHS) trusts across England, Scotland and Wales, with additional publicity through the patient charity Lynch syndrome UK. Participants were sent the study information pack and could consent and complete the study questionnaire electronically or on paper.23,24 Ethics This study was approved by the Surrey Research Ethics Committee on 05/12/2022 (22/PR/1167). The questionnaire and other patient-facing materials were peer-reviewed by patient representatives including from the Lynch Syndrome UK charity (https://www.lynch-syndrome-uk.org/). Lynch Syndrome UK supported increasing awareness of the research, including through social-media channels, and were represented on the study management group and manuscript co-authorship. Procedures The study questionnaire included the EQ-5D-5L questionnaire (NICE recommended25) to measure QoL at time of survey completion.26 Patient demographics, medical and surgical history, visual-analogue scale (VAS), and the Cancer–Worry Scale were also captured.27 A retrospective EQ-5D for QoL of patients 4 months after RRH was also recorded for those who had RRH. Outcomes The primary endpoint was the adjusted mean-difference in current utilities between those who had RRH versus not overall, and secondarily in subgroups by (a) women <51 years in the RRH group versus premenopausal women in no-RRH group; (b) women ≥51 years in the RRH group versus postmenopausal women in no-RRH group. An exploratory outcome was the age-standardised difference in utilities between women with Lynch Syndrome and the general female-population in England. Statistics Sample size was planned by considering the population of 5112 identified female Lynch Syndrome carriers in England,28 of which around 30% have undergone RRH29; around 1500 women. We aimed to recruit 750 women with Lynch Syndrome overall (range 300–1200) including 250 women who had RRH (range 100–400). This was consistent with previous questionnaire studies in women at inheritable risk of gynaecological cancers, and was expected to include a representation from all types of carriers,30 and to allow for informative post-hoc exploratory analysis. Utilities were calculated by mapping between EQ-5D-5L and 3L using the NICE-recommended UK value-set.25,26 Patients were excluded if responses for the following variables were incomplete or uncertain: age, RRH-status, and EQ-5D domains. Participants were separated into “RRH” and “no-RRH” groups based on self-reported RRH ± BSO status. Patients who reported hysterectomy for cancer treatment or benign gynaecological disease, and BSO (without hysterectomy) were excluded. Patients who underwent unilateral salpingo-oophorectomy (without hysterectomy) were classified as no-RRH. Menopausal status of ‘no-RRH group’ women was defined in relation to menstrual periods; patients reporting no periods in the previous 12-months were classed as postmenopausal, unless this was due to contraception or other hormonal medication. Patients reporting ≥1 period in the last 12-months were classed as premenopausal, unless they were ≥56 years.31 Patients in the RRH-group were compared against pre/postmenopausal patients in the no-RRH group based on current age under/over 51 years (mean-age of menopause in UK women31), respectively. This enabled a comparison of patients who underwent RRH against an equivalent cohort of participants who have not, to estimate the impact of RRH on current QoL. For the sensitivity analysis, patients in the RRH-group were compared based on age at RRH under or over 51 years. A physical health condition was defined as any physical health condition other than cancer. A mental health condition was defined as any such condition, including anxiety, depression or other. Categorical variables were dichotomised post-hoc for the regression-analysis, including marital status (married/cohabiting versus other), household income (>£40,000 versus <£40,000, given median household-income of £38,00032), and education level (university/college versus secondary/high-school). Primarily, complete-case analysis was performed to evaluate the adjusted mean difference in utilities between RRH groups using linear regression overall, and by pre/post menopause. This used Ordinary-Least-Squares (OLS) regression, using the normal approximation of the bootstrap standard error for confidence intervals.33, 34, 35, 36 Other analysis tested independence between categorical variables by RRH group using Chi-squared, or Fisher’s-exact test if the expected cell counts were less than five for any cell. For continuous variables, T-tests were used if the data appeared to be normally distributed, and Mann-Whitney-U tests if the data did not appear normally distributed, based on visual inspection. Correlation between total Cancer-Worry-Scale scores and utilities used Spearman’s-coefficient. Due to the observational nature of this study, adjustment for confounders is vital for interpretation. We focused on factors that were associated with utilities in the no-RRH group and judged unlikely to be substantially caused by RRH. Since many potential variables were captured in the study, we opted to use a partly data-driven approach to choose those most strongly associated with utilities in the no-RRH group, using a Least-Absolute-Shrinkage and Selection-Operator (LASSO) linear-regression model. The impact of time since RRH to survey completion on utilities was evaluated within the RRH-group only, and in separate analysis stratifying the RRH-group by years since RRH: 0–2 years, 3–5 years, ≥6 years, comparing each against the no-RRH group. Exploratory analysis compared utilities from the PRESCORES sample with general female-population values for England, using a recent Health-Survey for England analysis based on 8094 women.37 EQ-5D utilities from PRESCORES participants were age-standardised against the reference population37 using direct standardisation, and the adjusted difference in mean-utilities between the PRESCORES and reference populations was assessed by OLS. We performed further subgroup-analysis for participants with/without a personal cancer-history. Several sensitivity analyses to evaluate robustness were conducted: 1) The primary analysis was repeated for people 35–80 years (RRH is only recommended for those ≥35 years); 2) The impact of missing data was evaluating using multiple imputation based using multiple imputation by chained equations, drawing an imputed sample in each bootstrap re-sample38; 3) An alternative model was used for primary analysis in people aged 35–80 years old, embedding propensity score weighting within bootstrap resampling, in a complete-case or multiple imputation framework. This used the same adjustment factors as the primary analysis. Here the marginal average treatment effect was estimated with inverse probability weighting applied to trimmed (99% percentile) weights, and through a doubly-robust approach39; 4) An unadjusted sensitivity-analysis was conducted for the subset of patients who underwent RRH without BSO; 5) A further sensitivity analysis aimed to determine the adjusted mean-difference in utilities 4-months after RRH between RRH and no-RRH groups, comparing utilities 4-months after RRH in the RRH-group to current utilities in ‘no-RRH group’ women. The three additional retrospective analyses were: 1) Utilities 4-months after RRH in the RRH group versus current utilities in no-RRH group; 2) Utilities 4-months after RRH in women who underwent RRH < 51 years versus current utilities of premenopausal women in no-RRH group; 3) Utilities 4-months after RRH in women who underwent RRH ≥ 51 years versus current utilities of postmenopausal women in no-RRH group. All statistical tests were two-sided, at 5% significance level, without adjustments for multiple testing. Data cleaning were undertaken using SPSS version 29, with primary, secondary and exploratory analyses undertaken using Stata version 18, and some sensitivity analyses using R version 4.4.2, with R packages MatchIt version 4.7.2 for propensity score matching and WeightIt for inverse probability weighting.40 Role of funding source This study was funded by the Rosetrees Trust (CF1\100001 5) and Barts Charity (G-002286 and MRC0167). Funders had no role in the design, data collection, data analysis, interpretation or writing of the report. Results In total, 4031 patients with Lynch Syndrome were directly invited to participate from 25 NHS trusts across England, Scotland and Wales (Appendix Fig. S1) between May 2023 and November 2024. It is not possible to know how many additional patients saw Lynch Syndrome UK publicity. Overall, 1321 participants provided informed consent to participate in the study and returned questionnaires. After excluding 201 patients, 1120 participants were included for analysis (Fig. 1). Study flow diagram. The clinical and demographic characteristics of included participants are summarised in Table 1. Of the 529 participants who underwent RRH, 500 had RRH-BSO and 29 had RRH without BSO. RRH and no-RRH groups differed significantly with respect to age and some clinical and demographic factors. Table 1Characteristics of included participants, provided as numbers (% of column total).RRHno-RRHOverallp-valueN5295911120Mean age, yrs [SD]54.9 [9.8]42.9 [14.2]48.6 [13.7]<0.0001aMedian age, yrs [IQR, range]55 [48–62, 32–82]39 [33–52, 18–87]49 [38–59, 18–87]Lynch Syndrome genes MLH1120 (22.7)127 (21.5)247 (22.1)<0.0001a MSH2150 (28.4)137 (23.2)287 (25.6) MHS6118 (22.3)97 (16.4)215 (19.2) PMS248 (9.1)90 (15.2)138 (12.3) EPCAM2 (0.4)8 (1.4)10 (0.9) Not sure/missing91 (17.2)132 (22.3)223 (19.9)Mean age at RRH, years (SD)47.4 (8.9)–––Median years since RRH (IQR)6 (2–11)–––Menopausal status/age Premenopausal189 (35.7)433 (73.3)–<0.0001a Postmenopausal340 (64.3)158 (26.7)–HRT ever user No181 (34.2)511 (86.5)692 (61.8)<0.0001a Yes336 (63.5)67 (11.3)403 (36.0) Not sure1 (0.19)5 (0.85)6 (0.54) Missing11 (2.1)8 (1.4)19 (1.7)HRT current user No266 (51.5)532 (92.0)798 (72.9)<0.0001a Yes251 (48.6)46 (8.0)297 (27.1) Missing12 (2.3)13 (2.2)25 (2.2)Personal cancer history No326 (61.6)445 (75.3)771 (68.8)<0.0001a Yes202 (38.2)145 (24.5)347 (31.0) Missing1 (0.19)1 (0.17)2 (0.18) Cancer typeb: Bowel144 (71.3)108 (74.5)252 (72.6) Breast36 (17.8)28 (19.3)64 (18.4) Other43 (21.3)22 (15.2)65 (18.7)Number of physical comorbidities 0257 (48.6)334 (56.5)591 (52.8)0.0061 1152 (28.7)163 (27.6)315 (28.1) 2+120 (22.7)94 (15.9)214 (19.1)Mental health conditionb No386 (73.0)421 (71.3)807 (72.1)0.32 Yes136 (25.7)163 (27.6)299 (26.7) Missing7 (1.3)7 (1.2)14 (1.3) Anxiety110 (80.9)136 (83.4)246 (82.3) Depression67 (49.3)74 (45.4)141 (47.2) Other9 (6.6)28 (17.2)37 (12.4)Have children No65 (12.3)215 (36.4)280 (25.0)<0.0001a Yes454 (85.8)372 (63.0)826 (73.8) Missing10 (1.9)4 (0.68)14 (1.3)Aspirin use No321 (60.7)416 (70.4)737 (65.8)<0.0001a Yes195 (36.9)166 (28.1)361 (32.2) Not sure1 (0.2)3 (0.51)4 (0.36) Missing12 (2.3)6 (1.0)18 (1.6)Attend colorectal screening No26 (4.9)89 (15.1)115 (10.3)<0.0001a Yes491 (92.8)491 (83.1)982 (87.7) Not sure1 (0.19)6 (1.0)7 (0.62) Missing11 (2.1)5 (0.85)16 (1.4)Attend endometrial cancer screening No–393 (66.5)–– Yes–162 (27.4)– Not sure–31 (5.3)– Missing5 (0.85)Family history of cancer No13 (2.5)13 (2.2)26 (2.3)0.77 Yes503 (95.1)565 (95.6)1068 (95.4) Not sure3 (0.57)8 (1.4)11 (0.98) Missing10 (1.9)5 (0.85)15 (1.3)Ethnicity White483 (91.3)537 (90.9)1020 (91.1)0.088 Asian12 (2.3)19 (3.2)31 (2.8) Black7 (1.3)6 (1.0)13 (1.2) Other7 (1.3)21 (3.6)28 (2.5) Missing20 (3.8)8 (1.4)28 (2.5)Marital status (%) Single40 (7.6)116 (19.6)156 (13.9)<0.0001a Married/Cohabiting395 (74.7)407 (68.9)802 (71.6) Divorced/Separated51 (9.6)37 (6.3)88 (7.9) Widowed19 (3.6)17 (2.9)36 (3.2) Missing24 (4.5)14 (2.4)38 (3.4)Education level No formal20 (3.8)26 (4.4)46 (4.1)<0.0001a GCSE/Equivalent152 (28.7)98 (16.6)250 (22.3) A-level/Equivalent93 (17.6)93 (15.7)186 (16.6) Higher Education242 (45.8)359 (60.7)601 (53.7) Missing22 (4.2)15 (2.5)37 (3.3)Income Less than £10,00020 (3.8)29 (4.9)49 (4.4)0.17 £10,000–£19,99973 (13.8)58 (9.8)131 (11.7) £20,000–£29,99972 (13.6)82 (13.9)154 (13.8) £30,000–£39,99957 (10.8)83 (14.0)140 (12.5) £40,000–£49,99968 (12.9)72 (12.2)140 (12.5) £50,000 or more190 (35.9)236 (39.9)426 (38.0) Missing49 (9.3)31 (5.3)80 (7.1)RRH—risk-reducing hysterectomy, IQR—interquartile range; HRT—Hormone replacement therapy. p-value indicates the result of a Chi squared/Fisher’s exact test or Mann-Whitey U test to compare RRH and no-RRH groups and are calculated using known values only.aIndicates statistical significance at 0.05.bValues may sum to over 100% due to multiple non-exclusive comorbidities/variable options. Characteristics of included participants, provided as numbers (% of column total). RRH—risk-reducing hysterectomy, IQR—interquartile range; HRT—Hormone replacement therapy. p-value indicates the result of a Chi squared/Fisher’s exact test or Mann-Whitey U test to compare RRH and no-RRH groups and are calculated using known values only. Indicates statistical significance at 0.05. Values may sum to over 100% due to multiple non-exclusive comorbidities/variable options. Responses to the five EQ-5D domains are reported in Appendix Table S1. No-RRH group patients were less-likely to have problems with mobility (p < 0.0001), doing usual activities (p = 0.0010), and pain/discomfort (p = 0.013), compared to RRH-group patients. The overall unadjusted mean-utilities within the whole study population was 0.812 (SD 0.207), 0.793 (0.225) within the RRH-group and 0.828 (0.189) in the no-RRH-group. These results are shown in Fig. 2, including premenopausal/postmenopausal-age comparison groups. See Appendix Table S2 for the full dataset. Unadjusted utilities of “Risk-reducing hysterectomy (RRH)” and “No-RRH” groups. The factors associated with utilities in the no-RRH-group and used for adjustment were: presence of any physical health condition (−0.091, 95% CI −0.117 to −0.065; p < 0.0001), mental health condition (−0.165, −0.203 to −0.128; p < 0.0001), household income ≥£40,000/year (0.042, 0.017 to 0.068; p = 0.0013), and a personal cancer history (−0.053, −0.086 to −0.021; p = 0.0013). Age was not significantly associated with utilities after adjustment for these factors but was also included as an adjustment because it is still an important confounder that is associated with EQ-5D, and differed substantially between groups. Primary analysis revealed slightly lower utilities in the RRH group (adjusted mean difference −0.020, 95% CI −0.045 to 0.004; p = 0.11; complete-case n = 1032, n = 478 RRH). Utilities were significantly lower in the premenopausal RRH group (adjusted mean −0.041, −0.074 to −0.008; p = 0.014, n = 584) but similar (0.003, −0.036 to 0.042; p = 0.88, n = 448) in the postmenopausal group (Table 2). Table 2Adjusted mean-difference in current utilities between “Risk-reducing hysterectomy (RRH)” and “no-RRH” groups, overall, and by current age in the RRH group (