Risk-reducing gynecologic surgery decisions in Lynch syndrome at a safety-net hospital and university medical center.

Medical record review for Lynch syndrome patients who underwent genetic counseling was approved by the Institutional Review Board. Patients with Lynch syndrome seen between 2012-2024 were first identified by querying our internal Progeny Genetics software database (Progeny version 10.6.2.0) which includes all patients receiving genetic counseling by a board-certified genetic counselor (CGC) or an advanced practice genetics nurse practitioner (APNG) at our university medical center or safety-net hospital facilities. Subsequently, the medical records at these facilities were searched, and patients were included in the study if they received gynecology care following genetic counseling. The population included patients referred to genetic counseling from a variety of providers, including primary care, oncology, and self-referrals. Patients at the safety-net hospital received care from genetic counselors who were bilingual in Spanish and English. Patients at the university medical center were seen by English-speaking genetic counselors with the assistance of a certified medical interpreter in their preferred language. Genetic counselors at both sites used the same risk counseling methods. After undergoing genetic counseling, patients at the university medical center were typically cared for by gynecology or gynecologic oncology faculty, while patients at the safety-net hospital were primarily cared for by residents in gynecology or fellows in gynecologic oncology who were supervised by the same attending faculty physicians who practice at the university medical center. Patients at the university medical center had the option to be followed in a high-risk genetic cancer screening clinic for their gynecology care, while patients at the safety-net hospital were followed in general gynecology clinics. Patients at both sites received the same pre-operative counseling. There were no specific menopausal support clinics or programs at either site during the study period. Eligible patients had an intact uterus and/or ovaries and no prior uterine or ovarian cancer at the time of genetic testing that confirmed their Lynch syndrome diagnosis. We defined “age censored” as the age of risk-reducing surgery or the age at last observation in the medical record if the patient did not undergo surgery. Medical record reviews were conducted by a genetic counselor or a research assistant. Data collected included demographic characteristics, cancer diagnoses, dates of surgeries performed for risk reduction, parity, family history, benign gynecologic diagnoses, and use of hormone replacement therapy following risk-reducing surgery. After reviewing eligible patients of all ages, women aged 30 or above were included in the analysis since women under age 30 are not typically advised to undergo risk-reducing surgery, and the youngest patient to undergo risk-reducing surgery in this cohort was 33 years old. Reasons for declining or deferring risk-reducing surgery were abstracted if noted in the clinical documentation. This was a retrospective, hypothesis-generating study designed to describe cancer risk-management patterns across hospital sites. No formal sample size or power calculation was performed, as the analyses were exploratory in nature and intended to generate hypotheses for future, adequately powered studies. A chi-squared analysis was conducted to examine potential factors associated with undergoing risk-reducing gynecologic surgery including hospital site, age censored, age at Lynch syndrome diagnosis, race/ethnicity, preferred language, Lynch-associated pathogenic variant, prior history of cancer, family history of gynecologic cancer, history of benign gynecologic diagnoses, and parity. After recognizing that hospital site was the strongest associated factor, chi-squared analyses were completed to assess potential differences between patients across the two hospital sites including age censored, age at Lynch syndrome diagnosis, age at first risk-reducing surgery, race/ethnicity, preferred language, prior history of cancer, family history of gynecologic cancer, history of benign gynecologic diagnoses, parity, and post-surgical hormone replacement therapy use among those undergoing surgery. Cox proportional hazards regression models were used to evaluate factors associated with the outcome while controlling for potential confounders. The proportional hazards assumption was tested and not violated. Because of multicollinearity between hospital site and race/ethnicity (condition number = 263), only age at censoring and hospital site were included in the multivariable models. Time to risk-reducing surgery was analyzed using cumulative incidence functions, and differences among hospital sites were assessed with Gray’s test. All analyses were performed in R (version 4.5.0), and the critical value was set at 0.05 [ 8 ].

A total of 65 patients met inclusion criteria, including 28 from the safety-net hospital and 37 from the university medical center. Demographic information is shown in Table 1 , including age censored, age at Lynch syndrome diagnosis and at first risk-reducing surgery, race/ethnicity, preferred language, prior history of cancer, family history of gynecologic cancer, history of benign gynecologic diagnoses, parity, and use of hormone replacement therapy post risk-reducing surgery. All patients identified as women. The median age censored of the 65 patients was 43.0 years (IQR 36.0 – 50.0), which did not differ significantly between the university medical center and safety-net hospital. The median age of Lynch syndrome diagnosis was 41.0 (IQR 31.0 to 48.0), which differed significantly between the university medical center (median age 35.0 [IQR 28.0 to 46.0]) and the safety net hospital (median age 45.0 [IQR 37.8 to 48.2]) (p = 0.026). The sample included a diverse population, with more Hispanic individuals (46.2%) than any other race/ethnicity, and over a third 24/65 (36.9%) whose preferred language was Spanish. None of the Hispanic patients reported Black/African American ancestry. Most patients at the safety-net hospital were Hispanic (24/28, 85.7%), and the majority at the university medical center were White, non-Hispanic (26/37, 70.3%) (p<0.001). Additionally, most patients at the safety-net hospital were Spanish-language preferring (23/28, 82.1%), and the majority at the university medical center preferred English (35/37, 94.6%) (p<0.001). More than half of patients (35/65, 53.8%) had at least one prior cancer diagnosis, with colorectal cancer being the most common cancer reported (24/35, 68.6%), followed by breast cancer (7/35, 20.0%). A greater proportion of patients at the safety-net hospital had a prior history of cancer (75.0% versus 37.8%, p = 0.006). The distribution of Lynch-associated pathogenic variants among the 65 patients is shown in Figure 1 , and pathogenic variants in MLH1 were most common (27/65, 41.5%). The distribution of Lynch-syndrome causative genes did not differ between the two hospital sites (p = 0.113). Less than half of the patients (26/65, 40.0%) had a family history of gynecologic cancer, and this did not differ between the two hospital sites. One-fifth of the study population (14/65, 21.5%) had a personal history of a benign gynecologic diagnosis, with ovarian cysts being the most common (6/14, 42.9%), and there was no significant difference between the two hospital sites. Almost half of the patients underwent risk-reducing gynecologic surgery (32/65, 49.2%), with a higher rate of risk-reducing surgery at the safety-net hospital (19/28, 67.9%) compared to the university medical center (13/37, 35.1%) (p = 0.018). In a univariate analysis, hospital site was the only factor associated with undergoing risk-reducing surgery that reached statistical significance. There were several factors that neared but did not reach statistical significance including age censored, race/ethnicity, preferred language, prior history of cancer, and family history of gynecologic cancer. The median age of patients who underwent risk-reducing surgery was 44.5 (IQR 40.8 - 53.2) compared to the median age for non-surgical patients which was 38.0 (IQR 36.0 - 48.0) (p = 0.118). The majority of Hispanic patients underwent surgery (19/30, 63.3%) while 37.1% (13/35) of non-Hispanic patients underwent surgery (p = 0.063). A smaller proportion of English-preferring patients underwent surgery (15/39, 38.5%) compared to non-English-preferring patients (17/26, 65.4%) (p = 0.061). Of the 35 patients with a prior cancer diagnosis, 21 (60.0%) underwent risk-reducing surgery, while 11/30 (36.7%) patients without a cancer history underwent surgery (p = 0.104). There were 9/26 (34.6%) patients with a family history of gynecologic cancer compared to 23/39 (59%) patients without a family history of gynecologic cancer who underwent surgery (p = 0.095). Other factors that were analyzed that did not reach significance included Lynch pathogenic variant (p = 0.184), history of benign gynecologic diagnoses (p = 0.714), and parity (p = 0.234). The median age at first risk-reducing surgery was 46 (range 33-57) at the safety-net hospital and 41 (range 33-60) at the university medical center. In nearly all cases, the surgeries included both risk-reducing hysterectomy and RRSO (29/32, 90.6%). One patient at the safety-net hospital with a germline MLH1 pathogenic variant underwent risk-reducing hysterectomy and bilateral salpingectomy only, declining risk-reducing bilateral oophorectomy, at age 54. One patient at the university medical center underwent RRSO at age 60 given she had a prior hysterectomy for uterine fibroids. Another patient at the university medical center underwent risk-reducing hysterectomy and bilateral salpingectomy at age 42. After undergoing surgery, some patients (15/32, 46.9%) received hormone replacement therapy. Of the 17 patients who underwent surgery and did not receive hormone replacement therapy, 10/17 (58.8%) were age 50 and older and 7/17 (41.2%) were under 50. The rate of hormone replacement therapy use did not differ between the two hospital sites (p = 1.00). Of the 33 patients who did not complete risk-reducing surgery, nine (27.3%) were 40-49 years old, including four (44.4%) from the safety net hospital and five (55.6%) from the university medical center. There was documentation for reasons surgery was deferred or declined for all nine patients. Similar reasons were reported across the two hospital sites and included a desire to preserve fertility, avoid menopausal symptoms, other medical comorbidities, and/or because of a lower risk of cancer associated with PMS2. Seven (21.2%) of the 33 patients who did not undergo surgery were age 50 or older including one patient (14.3%) from the safety-net hospital and six patients (85.7%) from the university medical center. Reasons for not undergoing surgery were reported for 6 (85.7%) of these patients and included fear of surgical complications, wanting to avoid additional surgery after treatment for colorectal cancer, other medical comorbidities, and lower risk of cancer associated with PMS2 . As shown in the Cox regression in Table 2 , after adjusting for history of prior cancer and age censored, patients at the safety-net hospital were more likely to undergo risk-reducing surgery than patients at the university medical center (HR 2.43, p = 0.02). Additionally, the time to risk-reducing surgery after Lynch syndrome diagnosis was shorter for patients at the safety-net hospital compared to those at the university medical center ( Figure 2 ). Some patients received gynecologic cancer screening as part of their Lynch syndrome management. A total of 28/65 (43.1%) patients received at least one endometrial biopsy for screening purposes, 11/65 (16.9%) patients had CA-125, and 33/65 (50.8%) underwent transvaginal ultrasound for screening. Between the two hospital sites, there were no significant differences observed in the rates of screening. Transvaginal ultrasound was performed in 50.0% of patients at the safety-net hospital and 51.4% of patients at the university medical center (p = 1.0). CA-125 screening was higher in the safety-net hospital than in the university medical center (28.6% versus 8.1%), and the same trend was observed for endometrial biopsy (57.1% versus 32.4%), but the differences were also not statistically significant (p = 0.065 for CA-125 and p = 0.082 for endometrial biopsy).

This study explored the utilization of surgery and screening for gynecologic cancer risk management among patients with Lynch syndrome and found that patients followed at a safety-net hospital were over twice as likely to undergo risk-reducing gynecologic surgery than patients at a university medical center. Additionally, patients at the safety-net hospital had a quicker uptake of surgery with most surgeries occurring within three years after Lynch diagnosis. Hospital site was the only factor that was significantly associated with undergoing risk-reducing surgery, but other factors trended towards an association, such as older age, history of another cancer, Hispanic ancestry, and preferring a language other than English. Notably, other factors such as parity, having a benign gynecologic diagnosis, or a family history of cancer were not associated with risk-reducing surgery. While limited, prior studies at safety-net hospitals have also reported high rates of cancer risk-reducing surgery, mostly in the context of breast and ovarian cancer predisposition. One study investigating the uptake of contralateral prophylactic mastectomy in breast cancer patients at a safety-net hospital found that patients with pathogenic variants in breast cancer susceptibility genes had an 86.3% rate of contralateral prophylactic mastectomy, which was higher than expected rates from similar studies in university and private medical center settings [ 7 ]. Another study investigating the uptake of RRSO in a diverse population of BRCA1/2 pathogenic variant carriers at a public urban hospital found that 52% of patients underwent RRSO, which is similar to reported rates of 50-70% at private centers [ 9 ]. It is possible that patients at safety-net hospitals may have uncertain access to future preventive care and may view risk-reducing surgery as a more long-term cost-effective approach at managing cancer risk [ 7 ]. This may contribute to why patients at the safety-net hospital underwent risk-reducing surgery more quickly after receipt of their Lynch syndrome diagnosis, to engage in risk-reduction while they had health care coverage. Historically, medical redlining in Los Angeles (LA) has been associated with lower access to colorectal, breast, and cervical cancer screening [ 10 ]. The impact of medical redlining has not been researched in racial and ethnic minorities with Lynch syndrome specifically. While individuals with Lynch syndrome living in redlined neighborhoods in LA may have had lower access to Lynch-syndrome related care, this does not seem to be reflected in our findings. Given that the rates of gynecologic cancer screening for patients with Lynch syndrome did not differ significantly between the two hospital sites, it is possible that access to care through a large safety-net hospital along with healthcare initiatives such as the Affordable Care Act, Every Woman Counts programs, and expanded Medicaid programs may be contributing to more equitable access and may have helped to alleviate the historical disparities in the area. Previous studies have found an association between higher education level and being more active in medical decision-making [ 11 , 12 ]. Our study found a 35% uptake of risk-reducing surgery among patients at the university medical center, which is similar to rates reported by an international multicenter, prospective study of 2292 Lynch syndrome carriers aged 30-69 years, which reported that 29% underwent risk-reducing hysterectomy and 26% underwent RRSO [ 13 ]. Patients with more education may be more attuned to possible surgical risks, such as side effects of surgical menopause, and be more likely to question medical authority causing increased hesitation to undergo surgery. Prior research on confidence in medicine in the US demonstrated that obedience to doctors’ authority was lower among respondents who were White, completed more years of education, and had a higher income [ 14 ]. This would suggest that patients at the university medical center may be more likely to question the option of risk-reducing surgery presented by their physicians causing them to delay or decline surgery. Of note, risk-reducing hysterectomy has not been shown to reduce mortality in patients with Lynch syndrome [ 2 ]. The distinction between reduced cancer incidence and overall survival may influence surgery uptake. For instance, patients with higher levels of education may be more likely to engage in shared decision-making and critically evaluate these more nuanced trade-offs. Future qualitative studies could help elucidate what factors patients with diverse educational backgrounds consider when making decisions about risk-reducing surgery, as well as how these discussions may vary across care settings. A previous study of patients with Lynch syndrome found that having less than a college education was associated with higher uptake of risk-reducing surgery such that 85% (15/23) without a college degree had a risk-reducing hysterectomy and RRSO in comparison to 40% (37/92) of those with a college degree or higher [ 3 ]. Although highest education level was not collected in this study, it is possible that educational level is an underlying factor that makes women from a safety-net hospital more likely to undergo risk-reducing surgery. Further research is needed to explore the role of education in surgical decision making so that strategies can be tailored to facilitate informed decision making across educational levels. Previous studies have found older age to be a factor associated with undergoing risk-reducing gynecologic surgery in patients with Lynch syndrome [ 3 – 5 ]. The current study also found a non-significant trend that women who were older were more likely to undergo risk-reducing surgery. Factors correlated with increasing age that could influence this association include completion of childbearing and menopausal status. Several studies have demonstrated that having a prior history of cancer is associated with undergoing risk-reducing gynecologic surgery in patients with Lynch syndrome and BRCA1/2 pathogenic variants [ 3 , 15 – 17 ]. Patients with a history of cancer may be more likely to want to avoid a second cancer diagnosis after difficult experiences undergoing cancer treatment. In a study surveying women with Lynch syndrome in support groups, one individual remarked that she decided to undergo RRSO because she would, “much rather deal with hot flashes than go through chemo again.”[ 3 ] Given that more women at the safety-net hospital had a prior history of cancer, our multivariable analysis adjusted for cancer history as a possible confounder; despite this, those who received care at the safety-net hospital were still significantly more likely to undergo risk-reducing surgery. Given that the safety-net hospital serves a primarily Hispanic population, our data support emerging evidence that Hispanic women with pathogenic variants are increasingly choosing preventive surgery at the same or higher rates as non-Hispanic women as demonstrated through prior research involving women with BRCA1/2 pathogenic variants. For instance, the UPTAKE study explored the rate of RRSO among 100 Latina women with BRCA1/2 pathogenic variants living in the United States from a diverse range of hospital settings and found a high rate of surgery uptake, with 68% undergoing RRSO, which is close to the estimated 70-80% uptake of RRSO in non-Hispanic, White, US populations [ 18 ]. An international cohort of 1736 Hispanic women with pathogenic variants in BRCA1/2 or other breast or ovarian cancer susceptibility genes found that 56.7% of BRCA carriers underwent RRSO and that women living in the US were more likely to undergo risk-reducing surgery than those living in Latin America (OR, 3.90) [ 19 ]. Another study observing uptake of RRSO in 90 diverse BRCA1/2 pathogenic variant carriers with a history of breast cancer reported similar results, with 90.9% of Latina patients undergoing RRSO compared to 76.6% of non-Hispanic white patients, though there were only 11 Latina patients, and the study did not report a statistical comparison. [ 20 ]. There are several potential cultural beliefs and values that could influence the uptake of risk-reducing surgery among Hispanic individuals in the United States, such as “respeto”, meaning deferring to medical authority, and “familismo,” which emphasizes a strong loyalty and obligation to support the family [ 21 , 22 ]. Our study was unable to examine cultural factors. Future qualitative research could help elucidate how Hispanic cultural beliefs could influence decision-making for risk-reducing surgery. This study was subject to several limitations. Given that Lynch syndrome is a rare condition, our study was limited by a small sample size and may have been underpowered to study the effect of other factors including age, race/ethnicity, preferred language, prior cancer history, and family history of gynecologic cancer, all which showed p-values near the 0.05 threshold. The directionality of the observed differences may reflect meaningful differences or clinical decision-making patterns that warrant further investigation with a larger sample. Additionally, the data gathered was limited to medical record abstraction. We were not able to assess more nuanced factors such as educational level, cultural beliefs, and desire for future fertility, all which could have impacted the decision to undergo risk-reducing surgery. Future mixed methods research can incorporate patient interviews to further investigate how patients incorporate these factors in their decisions. We acknowledge that the term Hispanic is used to encompass a diverse group of individuals of a variety of backgrounds and cultural perspectives. The patients included in this cohort may be representative of those seen at our safety-net hospital, but future larger studies would be needed across various facilities to explore behaviors across a more representative group of Hispanic women. While the same gynecologic oncology and gynecology faculty practice at both institutions, patients at the safety-net hospital more commonly received care from residents and fellow physicians. We did not assess the impact of physician type or physician training level on surgical decision-making. Further, some patients in the study were lost to follow-up and may have discontinued their care at our facilities before undergoing risk-reducing surgery. Future studies involving patient interviews can gather additional information regarding patients who may have continued care elsewhere. Patients at the university medical center were mostly followed in a high-risk genetic cancer screening clinic, while patients at the safety-net hospital were followed in general gynecology clinics. We did not have enough patients in our study followed in general gynecology clinics at the university medical center to examine the potential impact of this specific clinical setting on surgical outcomes. Ultimately, our findings are hypothesis-generating and should be explored in larger, prospective, multi-institutional studies including patient interviews. Such research will provide more insight into the complexities of the surgical decision-making process within a diverse patient population.

Lynch syndrome is a hereditary cancer predisposition syndrome caused by germline pathogenic variants in the mismatch repair genes and is associated with varying elevated risks for several types of cancers including colorectal, endometrial, ovarian, gastric, and other cancers [ 1 ]. The National Comprehensive Cancer Network guidelines recommend individualized management of endometrial and ovarian cancer risk, including consideration of screening and risk-reducing surgery (hysterectomy and/or risk-reducing salpingo-oophorectomy [RRSO]). Risk-reducing surgery is advised to be considered beginning at age 40 for carriers of MLH1, MSH2 , or EPCAM pathogenic variants; at age 40 for MSH6 carriers with delayed oophorectomy at age 50; and at age 50 for PMS2 carriers [ 2 ]. These guidelines have changed over time, can be gene-dependent, and are supported by limited data regarding the utility of screening for endometrial and ovarian cancer. Therefore, the decision to offer screening versus surgery is often left to the clinician’s discretion [ 2 ]. Little is known about the utilization of risk-reducing surgery and what factors may influence those decisions. The few studies performed have suggested that lower education level, increasing age, parity, post-menopausal status, no desire for future fertility, family history of Lynch syndrome-associated cancers, and having a personal history of non-gynecologic cancer may be associated with undergoing risk-reducing surgery [ 3 – 6 ]. Patients at a safety-net hospital may also face unique barriers to obtaining care, such as uncertainty about future insurance and access to care, as well as language barriers [ 7 ]. This exploratory study aimed to analyze the surgical decisions of patients with Lynch syndrome across a diverse patient cohort and expand existing research to include a safety-net hospital population.