Microbiome impacts quality of life in patients with endometrial cancer and benign gynecological conditions

As detailed in Łaniewski et al. [ 41 ], 192 patients undergoing hysterectomy at three clinical sites around the greater Phoenix area consented to and were enrolled in a study investigating host-microbiome interplay in endometrial cancer (EC) between June 2018 and February 2020. Eligibility criteria included females aged 18 and older undergoing hysterectomy for benign or malignant conditions, of any race/ethnicity. Patient diagnoses were confirmed postoperatively by histopathology of uterine tissue samples. Before hysterectomy, all patients were given seven surveys validated for measuring patient reported outcomes (PRO), which were used as proxies for assessing patient-reported preoperative QOL: National Institutes of Health Toolbox Global Health v1.2 [ 42 , 43 ], Gastrointestinal: Gas and Bloating v1.1 13a [ 44 , 45 ], Gastrointestinal: Diarrhea v1.0 6a [ 44 , 45 ], and Sexual Function and Satisfaction Brief Profile (Female) v2.0 [ 46 ] from the Patient-Reported Outcomes Measurement Information System (PROMIS), the perceived stress scale 10 (PSS10) [ 47 ], the vaginal assessment scale (VAS) [ 48 ], and the vulvar assessment scale (VuAS) [ 48 ]. All patients treated at Banner University Medical Center (BUMC) were included. Patients with EC or benign gynecological conditions (uterine fibroids, adenomyosis, and endometriosis) on postoperative histopathology were included. Patients with missing vaginal microbiome or survey data were excluded. All patients with EC were newly diagnosed and had not received treatment prior to data collection. Patients with benign gynecological conditions were used as a comparison group to patients with EC given the high prevalence, severity of symptoms, and previous work by Chase et al. outlining differences in preoperative QOL [ 40 ]. Survey responses were analyzed to ensure at least 50% response rate amongst participants. The number of responses to each survey by disease group can be found in Online Resource 1 Table  1 . All survey scores except VAS and VuAS were standardized to T-scores according to item response theory [ 49 ] for survey data analysis. The PROMIS Global Health survey was parsed into mental and physical health T-scores and analyzed separately. PROMIS surveys were solely analyzed as continuous T-scores, as described in Chase et. al.’s manuscript [ 40 ]. The PSS10, VAS, and VuAS scores were further dichotomized into ‘high’ and ‘low’ score groups using the median score of each survey (PSS10: T-score < 51 was ‘low’, ≥ 51 was ‘high’; VAS: score of 0–1 was ‘low’, 2–12 was ‘high’; VuAS: score of 0 was ‘low’, 1–12 was ‘high’). Differences in survey T-scores between patients with benign gynecologic conditions and EC were assessed with a Welch’s t test [ 50 ] (significant at p  < 0.05), and visualized in Prism 10.2.3 (GraphPad, San Diego, CA). Confounding variable analysis was performed to identify confounding demographic variables between the disease groups. Multiple linear regression analysis in Prism 10.2.3 (GraphPad, San Diego, CA) was used to investigate these differences further by adjusting for BMI and menopausal status, given the known confounding impact of these factors on QOL. BMI and menopausal status were confounding variables in all surveys except VAS and adjusted for when appropriate. Survey T-scores were also correlated to each other using a Spearman Rank correlation and visualized with a heatmap in Prism 10.2.3 (GraphPad, San Diego, CA). Inter-survey correlations were adjusted for BMI and menopausal status. Table 1 Association of demographics with disease status. While mean age was significantly higher in benign group, menopausal status was significantly more pre-menopausal in benign group. Vaginal pH was significantly more often > 4.5 in EC group, and BMI was significantly higher in EC group than benign All Benign EC p -value ( n  = 140) ( n  = 93) ( n  = 47) Age (mean (S.D.)) 50.23 (12.84) 60.60 (11.90) 44.99 (9.78) < 0.0001 Race  American Indian/Alaskan 10 (7.14) 4 (2.86) 6 (4.29) 0.0055  Asian/Far East/South East 4 (2.86) 4 (2.86) 0 (0.00)  Native Hawaiian/Pacific Islander 1 (0.71) 0 (0.00) 1 (0.71)  White/Caucasian 101 (72.14) 68 (48.57) 33 (23.57)  Black or African American 11 (7.86) 10 (7.14) 1 (0.71)  Middle Eastern/North African 1 (0.71) 0 (0.00) 1 (0.71)  Mixed or Multi Racial 9 (6.43) 7 (5.00) 2 (1.43)  Not specified, other 3 (2.14) 0 (0.00) 3 (2.14) Ethnicity  Non-Hispanic 106 (75.71) 65 (46.43) 41 (29.29) 0.0357  Hispanic 34 (24.29) 28 (20.00) 6 (4.29) Education  Less than High School 5 (3.68) 3 (2.21) 2 (1.47) 0.9166  High School Diploma or GED 28 (20.59) 18 (13.24) 10 (7.35)  Some College 35 (25.74) 22 (16.18) 13 (9.56)  Associate Degree or Technical Certification 29 (21.32) 19 (13.97) 10 (7.35)  Bachelor degree 23 (16.91) 17 (12.50) 6 (4.41)  Master/Doctoral Degree 16 (11.76) 12 (8.82) 4 (2.94) Income  <10,000 6 (4.51) 2 (1.50) 4 (3.01) 0.0002  10,000–25,000 16 (12.03) 10 (7.52) 6 (4.51)  25,000–50,000 28 (21.05) 10 (7.52) 18 (13.53)  50,000–75,000 29 (21.80) 22 (16.54) 7 (5.26)  75,000-100,000 21 (15.79) 20 (15.04) 1 (0.75)  >100,000 18 (13.53) 14 (10.53) 4 (3.01)  Don’t know/refused 15 (11.28) 10 (7.52) 5 (3.76) pH  ≤4.5 61 (43.57) 55 (39.29) 6 (4.29) 4.5 79 (56.43) 38 (27.14) 41 (29.29)  BMI (mean (S.D.)) 32.91 (9.71) 29.58 (7.11) 39.51 (10.80) < 0.0001 BMI  <25 27 (19.29) 22 (15.71) 5 (3.57) < 0.0001  25–29 41 (29.29) 37 (26.43) 4 (2.86)  30–34 25 (17.86) 17 (12.14) 8 (5.71)  ≥35 47 (33.57) 17 (12.14) 30 (21.43) Menopausal status  Pre-menopausal 89 (24.29) 80 (20.00) 9 (4.29) < 0.0001  Post-menopausal 51 (75.71) 13 (46.43) 38 (29.29) Association of demographics with disease status. While mean age was significantly higher in benign group, menopausal status was significantly more pre-menopausal in benign group. Vaginal pH was significantly more often > 4.5 in EC group, and BMI was significantly higher in EC group than benign Detailed information regarding microbiome sample collection and processing, DNA sequencing, demultiplexing, human contamination filtration, quality control and preprocessing, and taxonomic classification can be found in Online Resource 2. Diversity analyses were performed separately for the vaginal and rectal microbiomes. A rooted phylogenetic tree was created using ‘qiime phylogeny align-to-tree-mafft-fasttree’ [ 51 ]. ‘qiime diversity core-metrics-phylogenetic’ was run to compute diversity metrics. The sampling depths were confirmed to be representative of the microbiome’s diversity using ‘qiime diversity alpha-rarefaction’. ‘qiime diversity alpha-group-significance’ was used to compare Faith’s phylogenetic diversity (PD) [ 52 ] and evenness, measured by Shannon Diversity Index. Statistical comparisons of alpha diversity between disease groups were made using Mann-Whitney U [ 53 ] tests and visualized in bar graphs using Prism 10.2.3 (GraphPad, San Diego, CA). Correlations between alpha diversity metrics and survey T-scores were also performed using Spearman Rank correlations and visualized in scatterplots in Prism 10.2.3 (GraphPad, San Diego, CA). Lactobacillus species relative abundance and correlation to VAS and VuAS categories were analyzed using Mann-Whitney U [ 53 ] tests to determine whether one species of Lactobacillus was significantly increased or decreased in low or highly symptomatic patients. Survey T-scores were correlated to relative abundances of taxa using Spearman Rank correlations in Prism 10.2.3 (GraphPad, San Diego, CA). Results were visualized with dot plots in Prism 10.2.3 (GraphPad, San Diego, CA). To identify differentially abundant taxa, taxa counts at the species and genus level for vaginal samples and order level for rectal samples were conducted with analysis of compositions of microbiomes with bias correction (ANCOM-BC) package [ 54 ] utilizing R version 4.2.2. Adjustments for risk factors associated with EC, such as menopausal status (pre- and post-menopausal) and BMI (< 25, 25–29, 30–34, and ≥ 35), were performed. Adjustment analyses included menopausal status alone, BMI alone, and menopausal status and BMI together. P values were corrected for multiple comparisons using the FDR method. Taxa with a p -value < 0.05 and LFC ≥ 0.4 or q -value < 0.05 were considered significant. Heatmap visualizations of these data were created in Prism 10.2.3 (GraphPad, San Diego, CA). Methodological workflow and summary figures were made using BioRender at BioRender.com. Otherwise, visuals were created in Prism 10.2.3 (GraphPad, San Diego, CA), as noted above.

Of the 163 patients from BUMC, 9 had diagnoses other than EC or benign conditions and 14 had missing microbiome or survey data and were excluded. In total, 140 participants were identified for inclusion (Fig.  1 ): 93 patients had benign gynecological conditions and 47 had EC. Of the patients with benign conditions, 44 had adenomyosis, 21 had endometriosis, and 67 had uterine fibroids (39 patients diagnosed with > 1 benign condition). Thirty-nine of the patients with EC had grade 1 or 2 endometrial endometrioid carcinoma (EEC), while eight had a different EC subtype (grade 3 EEC, serous carcinoma, or other histological subtype). Patients with EC were more often post-menopausal ( p  < 0.0001), had higher BMI ( p   4.5) ( p  < 0.0001) than patients with benign conditions (Table  1 ). A complete table of demographic information collected can be found in Łaniewski et al. [ 41 ], and has also been described by Chase et al. [ 40 ]. Fig. 1 Inclusion/exclusion criteria and methodological workflow of the study. Differential abundance, assessed with ANCOM-BC, was adjusted for BMI and menopausal status. Created in BioRender. Gautam, N. (2025) https://BioRender.com/j32m804 Inclusion/exclusion criteria and methodological workflow of the study. Differential abundance, assessed with ANCOM-BC, was adjusted for BMI and menopausal status. Created in BioRender. Gautam, N. (2025) https://BioRender.com/j32m804 Adjusting for menopausal status and BMI where appropriate, two survey characteristics were found to be significantly different between patients with benign conditions and EC: VAS scores were increased in the benign group ( p  = 0.0131) compared to the EC group, and sexual interest T-scores were increased in the benign group ( p  = 0.0194) compared to the EC group (Online Resource 1 Fig. S1 ). In the entire study cohort, several significant correlations between survey scores were identified (Online Resource 1 Fig. S2 a): Good physical health and increased sexual interest ( p  = 0.0328, r  = 0.2979) were correlated, high perceived stress and worse vaginal ( p  = 0.0003, r  = 0.3729) and vulvar ( p  = 0.0046, r  = 0.2978) symptoms were correlated, high perceived stress and low sexual interest ( p  = 0.0352, r =-0.2141) were correlated, and high perceived stress and poor mental ( p  < 0.0001, r  = 0.-0.5604) and physical ( p  < 0.0001, r =-0.5245) health were correlated. Patients with benign gynecological conditions exhibited similar correlations (Online Resource 1 Fig. S2 b). In patients with EC, no significant correlation between physical health and sexual interest, or perceived stress and sexual interest, was seen. Perceived stress still positively correlated to worse vaginal ( p  = 0.0029, r  = 0.6440) and vulvar ( p  = 0.0400, r  = 0.4747) symptoms, and negatively correlated to good mental ( p  = 0.0005, r =-0.5301) and physical health ( p  = 0.0120, r =-0.3983), though less than in the benign group (Online Resource 1 Fig. S2 c). To understand differences in the vaginal and rectal microbiomes between patients with benign gynecological conditions and EC, we first examined and characterized the microbial species richness at each body site. Both metrics of species richness and evenness revealed patients with EC had significantly greater vaginal alpha diversity than patients with benign conditions (Faith’s phylogenetic diversity, p  < 0.0001; Shannon Index, p  < 0.0001) (Online Resource 1 Fig. S3a). Further, alpha diversity of the vaginal microbiome negatively correlated to vaginal symptoms in patients with EC ( p  = 0.0237, r =-0.4697) (Online Resource 1 Fig. S3 b)). No significant differences in species richness or correlations between QOL and alpha diversity were identified in rectal microbiomes (Online Resource 1 Figs. S3 a) & S3 c). L. iners was significantly increased in patients with EC with high VAS scores as compared to those with low VAS scores ( p  = 0.0266) (Fig.  2 a). There were no significant differences in L. iners between low and high VAS score groups in patients with benign conditions. No other significant differences were identified between Lactobacillus species abundances and VAS score groups (Fig.  2 b). Fig. 2 Patients with EC and worse vaginal symptoms had abnormally higher proportions of Lactobacillus iners than patients with better vaginal symptoms. ( a ) L. iners was significantly more abundant in patients with EC with high vaginal symptoms than patients with low symptoms ( p  = 0.0266), while ( b ) L. crispatus and other identified Lactobacillus species showed no significant differences between symptom groups in either benign or EC groups. Mann-Whitney U tests were used to compare relative abundances by VAS category. Benign with low VAS: n  = 33. Benign with high VAS: n  = 40. EC with low VAS: n  = 19. EC with high VAS: n  = 7. ( c ) ANCOM-BC analysis, adjusted for BMI and menopausal status, confirmed enrichment of L. iners in patients with EC and higher VAS scores ( p  = 0.0010). ( d ) L. iners was also enriched in patients with EC with higher VuAS scores ( p  < 0.001). Differential abundances ( p  < 0.05) are color-coded by log fold change (LFC): red (positive), grey (near or at 0), and blue (negative); LFC values shown within squares. Asterisk indicates q < 0.05 after FDR correction Patients with EC and worse vaginal symptoms had abnormally higher proportions of Lactobacillus iners than patients with better vaginal symptoms. ( a ) L. iners was significantly more abundant in patients with EC with high vaginal symptoms than patients with low symptoms ( p  = 0.0266), while ( b ) L. crispatus and other identified Lactobacillus species showed no significant differences between symptom groups in either benign or EC groups. Mann-Whitney U tests were used to compare relative abundances by VAS category. Benign with low VAS: n  = 33. Benign with high VAS: n  = 40. EC with low VAS: n  = 19. EC with high VAS: n  = 7. ( c ) ANCOM-BC analysis, adjusted for BMI and menopausal status, confirmed enrichment of L. iners in patients with EC and higher VAS scores ( p  = 0.0010). ( d ) L. iners was also enriched in patients with EC with higher VuAS scores ( p  < 0.001). Differential abundances ( p  < 0.05) are color-coded by log fold change (LFC): red (positive), grey (near or at 0), and blue (negative); LFC values shown within squares. Asterisk indicates q < 0.05 after FDR correction Several microbes were significantly associated with VAS and VuAS scores in patients with EC when controlling for BMI and menopausal status (Fig.  2 c and d), including those of the Lactobacillus genus. After false discovery rate (FDR) correction, three species were enriched in patients with high VuAS scores: L. iners (q = 0.0015, LFC = 1.2028), L. gasseri (q = 0.0340, LFC = 0.8950), and Streptococcus agalactiae (q < 0.0001, LFC = 0.9089) (Fig.  2 d). Several bacteria were also identified which were not significant after FDR correction, however still significant to p-value < 0.05 and LFC ≥ 0.4 (Fig.  2 c and d). In summary, L. iners was significantly enriched in patients with EC who reported high VAS scores, and L. iners , L. gasseri , and S. agalactiae were significantly enriched in patients with EC who reported high VuAS scores. In the entire study cohort, Porphyromonas ( p  = 0.0017, r =-0.3023) and Campylobacter ( p  = 0.0006, r =-0.3276) negatively correlated to sexual interest (Fig.  3 a and b). These correlations were not significant within individual disease groups (Online Resource 1 Figs. S4 a) & S4 b). In patients with benign gynecological conditions, Dialister B positively correlated to sexual interest ( p  = 0.0083, r  = 0.3006) (Fig.  3 c), and Lagierella correlated to lower VAS scores ( p  = 0.0030, r =-0.3427) (Online Resource 1 Figs. S4 c) & S4d). In the EC group, genera 28 L , DNF00809 ( p  = 0.0362, r =-0.4125), KA00274 ( p  = 0.0114, r =-0.4882), and Parvimonas ( p  = 0.0176, r =-0.4616) negatively correlated to VAS scores. DNF00809 ( p  = 0.0442, r =-0.4057) and UBA1822 ( p  = 0.0156, r =-0.4782) negatively correlated to VuAS scores. Fig. 3 Porphyromonas and Campylobacter B were associated with low sexual interest in all patients while Dialister B was associated with higher sexual interest in the benign group. ( a ) Porphyromonas and ( b ) Campylobacter B showed significant negative correlations with sexual interest T-scores in the entire study cohort ( p  = 0.0017 and p  = 0.0006, respectively). ( c ) Dialister B positively correlated to sexual interest T-scores in patients with benign gynecological conditions ( p  = 0.0083). See Online Resource 1 Fig. S4 for additional correlations in other disease groups Porphyromonas and Campylobacter B were associated with low sexual interest in all patients while Dialister B was associated with higher sexual interest in the benign group. ( a ) Porphyromonas and ( b ) Campylobacter B showed significant negative correlations with sexual interest T-scores in the entire study cohort ( p  = 0.0017 and p  = 0.0006, respectively). ( c ) Dialister B positively correlated to sexual interest T-scores in patients with benign gynecological conditions ( p  = 0.0083). See Online Resource 1 Fig. S4 for additional correlations in other disease groups No significant differences in Lactobacillus abundance were identified between low and high PSS10 score groups. Analysis of continuous PSS10 T-scores revealed genera DNF00809 ( p  = 0.0045, r =-0.4248) and Mobiluncus ( p  = 0.0417, r =-0.3120) were negatively correlated to perceived stress levels. ANCOM-BC further identified significant associations between vaginal microbiota and dichotomized scores (Online Resource 1 Fig. S5 ). In the entire study cohort, Anaerococcus tetradius ( p  = 0.0224, LFC=-0.6467), Lactobacillus mulieris ( p  = 0.0092, LFC=-1.6135), and Lactobacillus gasseri ( p  = 0.0324, LFC=-1.4275) were enriched in patients with high perceived stress scores. In the benign group, Lactobacillus mulieris ( p  = 0.0225, LFC=-1.7708) and Lactobacillus crispatu s ( p  = 0.0457, LFC=-2.014) were enriched in patients with high PSS10 scores. In the EC group, Lactobacillus gasseri ( p  = 0.0312, LFC=-2.033) was enriched in patients with high PSS10 scores. No significant associations were identified between rectal microbiota and PSS10 score groups. In summary, several Lactobacillus species and Anaerococcus tetradius were associated with higher perceived stress levels in both disease groups. In the entire study cohort, rectal Veillonellales ( p  < 0.0001, r  = 0.3498) (Online Resource 1 Fig. S6d) and Actinomycetales ( p  = 0.0003, r  = 0.3256) positively correlated to gas/bloating, Gastranaerophilales positively correlated to mental health ( p  < 0.0001, r  = 0.3276) (Fig.  4 c), and Actinomycetales ( p  = 0.0004, r  = 0.3377) positively correlated to sexual interest. In the benign group, Veillonellales positively correlated to gas/bloating ( p  = 0.0029, r  = 0.3229) and Gastranaerophilales positively correlated to mental health ( p  = 0.0034, r  = 0.3077) (Fig.  4 b & Online Resource 1 Fig. S6e). In the EC group, Christensenellales ( p  = 0.0356, r =-0.3375), Desulfovibrionales ( p  = 0.0344, r =-0.3397), and Flavobacteriales ( p  = 0.0207, r =-0.3693) all negatively correlated to gas/bloating (Fig.  4 a). Abundance of Veillonelalles was not significantly associated with gas/bloating in the EC group (Online Resource 1 Fig. S6d). Acidaminococcales positively correlated to diarrhea ( p  = 0.0156, r  = 0.4374). Gastranaerophilales again strongly positively correlated to good mental health in the EC group ( p  = 0.0041, r  = 0.4114) (Online Resource 1 Fig. S6e). Additionally, order RF32 ( p  = 0.0182, r  = 0.3431) and Verrucomicrobiales ( p  = 0.0056, r  = 0.3978) positively correlated to good mental health. Veillonellales negatively correlated to good mental health ( p  = 0.0350, r =-0.3083). Bacillales B ( p  = 0.0139, r  = 0.3566), Flavobacteriales ( p  = 0.0247, r  = 0.3274), Gastranaerophilales ( p  = 0.0034, r  = 0.4185), RF32 ( p  = 0.0036, r  = 0.4162), and TANB77 ( p  = 0.0125, r  = 0.3617) positively correlated with good physical health. Erysipelotrichales negatively correlated to good physical health ( p  = 0.0007, r =-0.4759). Mycobacteriales positively correlated to sexual interest ( p  = 0.0409, r  = 0.3819). Gastranaerophilales ( p  = 0.0094, r =-0.3919) and RF32 ( p  = 0.0119, r =-0.3802) negatively correlated to perceived stress, and Burkholderiales ( p  = 0.0237, r =-0.4421) negatively correlated to vaginal atrophy symptoms. Fig. 4 Increased orders Christensenellales, Desulfovibrionales, and Flavobacteriales in the rectal microbiome were found in patients with EC who experienced less gas and bloating, while Veillonellales was increased in patients with benign conditions who experienced more gas and bloating. a ) Christensenellales, Desulfovibrionales, and Flavobacteriales negatively correlated to increased gas and bloating in patients with EC ( p  = 0.0356, 0.0344, 0.0207 respectively). b ). Veillonellales positively correlated with increased gas and bloating in patients with benign conditions ( p  = 0.0029). See Online Resource 1 Fig. S6 for additional correlations in other groups. c ) Gastranaerophilales positively correlated to good mental health in all patients ( p  < 0.0001). d ) ANCOM-BC analysis revealed Burkholderiales (q = 0.0228) was negatively associated with VAS scores, and Verrucomicrobiales (q = 0.0237) was positively associated with VuAS scores, adjusting for BMI and menopausal status. All significant differential abundances ( p  < 0.05, LFC ≥ 0.4) are shown. Asterisk indicates q-value < 0.05 after FDR correction. Color scale represents log fold change (LFC): red (positive), grey (near or at 0), blue (negative); LFC values are shown within each square Increased orders Christensenellales, Desulfovibrionales, and Flavobacteriales in the rectal microbiome were found in patients with EC who experienced less gas and bloating, while Veillonellales was increased in patients with benign conditions who experienced more gas and bloating. a ) Christensenellales, Desulfovibrionales, and Flavobacteriales negatively correlated to increased gas and bloating in patients with EC ( p  = 0.0356, 0.0344, 0.0207 respectively). b ). Veillonellales positively correlated with increased gas and bloating in patients with benign conditions ( p  = 0.0029). See Online Resource 1 Fig. S6 for additional correlations in other groups. c ) Gastranaerophilales positively correlated to good mental health in all patients ( p  < 0.0001). d ) ANCOM-BC analysis revealed Burkholderiales (q = 0.0228) was negatively associated with VAS scores, and Verrucomicrobiales (q = 0.0237) was positively associated with VuAS scores, adjusting for BMI and menopausal status. All significant differential abundances ( p  < 0.05, LFC ≥ 0.4) are shown. Asterisk indicates q-value < 0.05 after FDR correction. Color scale represents log fold change (LFC): red (positive), grey (near or at 0), blue (negative); LFC values are shown within each square ANCOM-BC revealed no significant associations between rectal microbiota and survey T-scores in the entire cohort and benign group. Analysis in the EC group yielded few significant results after FDR correction: Verrucomicrobiales was enriched in patients with EC with higher VuAS scores (q = 0.0237, LFC = 0.8019), while Burkholderiales was depleted in patients with EC with higher VAS scores (q = 0.0228, LFC=-0.9160) (Fig.  4 d). Bacteria that were not significant after FDR correction, though still significant to a p-value < 0.05 and LFC ≥ 0.4 in the EC group, included: Bulkhoderiales ( p  = 0.0040, LFC=-0.7481) and Campylobacterales ( p  = 0.0204, LFC=-0.5249), which were depleted in patients with higher VuAS scores (Fig.  4 d).

In this study, we examined how preoperative patient-reported QOL correlations with vaginal and rectal microbiota differed between patients undergoing hysterectomy for EC and benign gynecological conditions. This study aids in the global understanding of the microbial impact on patient QOL and illuminates that, despite differences in pathophysiology, patients with benign gynecological conditions and EC experience several similar symptoms. Survey correlations were largely similar between disease groups, except for a lack of correlation between perceived stress and sexual interest in the EC group as compared to benign. Perceived stress may exhibit a larger negative impact on sexual interest levels in patients with benign gynecological conditions than patients with EC, despite correlating with similar physical vaginal and vulvar symptoms in both groups. While menopausal status was controlled for, this could still be an effect of age: older age has been associated with decreased sexual interest [ 17 ], better emotional regulation, and lower levels of perceived stress [ 55 , 56 ]. Vaginal microbiome findings largely reflected expected differences between pre- and post-menopausal patients [ 14 – 16 ]. A unique finding, however, was that lower alpha diversity was associated with worse symptoms of vaginal atrophy in patients with EC. Specifically, Lactobacillus iners was significantly increased. L. iners has been found to be predominant after vaginal infections with BV [ 57 ] and hypothesized to contribute to onset of vaginal dysbiosis [ 58 ] and inflammation [ 59 ], possibly explaining its association to vaginal atrophy symptoms. It also maintains high abundance in the vaginal microbiome during menstruation, while other lactobacilli might decline. Abnormal uterine bleeding from EC thus may favor abundance of L. iners over other Lactobacillus species in the vagina [ 60 ]. Nonetheless, increased L. iners was associated with poor vaginal symptoms in patients with EC. In patients with EC, Streptococcus agalactiae , L. iners , and L. gasseri were all also significantly enriched in patients with worse vulvar symptoms. S. agalactiae (group B Streptococcus ), while primarily studied for its pathogenic effects on neonates, has also been identified in cases of aerobic vaginitis and patients experiencing vulvovaginal symptoms [ 61 , 62 ]. Additionally, the link between L. gasseri and vulvar symptoms is surprising. A previous study determined that L. gasseri had the lowest odds of vaginal dryness out of all lactobacilli [ 63 ] in post-menopausal patients. Thus, this finding between L. gasseri and vulvar symptoms is conflicting. Dialister B , positively correlated to sexual interest in patients with benign gynecological conditions, has been found in both dysbiotic and symbiotic vaginal microbiomes in other studies [ 64 , 65 ]. Our finding of increased abundance of Dialister B in patients with benign gynecological conditions and higher sexual interest therefore adds to the literature on how Dialister B interacts within the vaginal microbiome and contribute to patients’ QOL. It remains unclear why this interaction was not similarly observed in EC patients. In the entire study cohort, increases in Porphyromonas and Campylobacter B in the vaginal microbiome were negatively correlated to sexual interest. Porphyromonas spp. in the vaginal microbiome has been correlated with EC when vaginal pH is above 4.5 [ 18 , 21 , 22 ]. Yet, our findings reveal an association between Porphyromonas in patients with benign gynecological conditions and decreased sexual interest, suggesting the microbe plays a role in QOL as well. In contrast, there is scant literature on Campylobacter in the vaginal microbiome and ties to QOL or vaginal symptoms [ 66 ]. A study on cervical cancer microbiota found both Porphyromonas and Campylobacter to be microbial markers of cervical cancer [ 67 ], though the authors also noted the novelty of this finding. It is worth considering a high degree of translocation between rectal and vaginal microbiota, which has been described in other bacteria [ 68 ], as Campylobacter is primarily described in the literature as a gut microbe [ 69 ], and Campylobacterales was found to be depleted in the rectal microbiome of patients with EC and high vulvar symptoms. In patients with high perceived stress levels, L. mulieris , L. gasseri , L. crispatus, and Anaerococcus tetradius were all enriched. One study found that 5-unit increases in perceived stress, as measured by PSS10, resulted in a higher risk of vaginal dysbiosis and lack of an L. iners -dominated microbiome [ 70 ], however did not include other Lactobacillus species. This is uniquely seen in our study, with perceived stress correlating to enrichment of Lactobacillus species that are not L. iners . This unusual correlation may also be affected by patients with EC or benign conditions experiencing higher levels of baseline perceived stress and rendering the vaginal microbiome altered from previous studies on otherwise ‘healthy’ patients. Additionally, while genus Mobiluncus negatively correlated to perceived stress in patients with EC, Mobiliuncus curtisii has previously been found enriched in patients with EC, with a hypothesis that it may promote EC carcinogenesis [ 71 ]. While our finding of this microbe was only in EC patients, it was protective of perceived stress. Several associations were identified between rectal microbes and gas/bloating scores, mental and physical health scores, and vaginal and vulvar symptoms. Order Veillonelales, which positively correlated to increased gas and bloating symptoms in patients with benign gynecological conditions, is also enriched in microbiomes with irritable bowel syndrome [ 72 , 73 ]. Christensenellales, negatively correlated to gas and bloating, has generally been associated with a healthy phenotype and longevity [ 74 ]. One study of Flavobacteriales, which negatively correlated to gas and bloating, found it significantly increased in gut microbiomes of female patients with bladder cancer [ 75 ]. Order Gastranaerophilales positively correlated to good mental health in the entire cohort and within each disease group, and to good physical health and low perceived stress in the EC group. The family of Gastranaerophilaceae was previously found to negatively correlate to emotional dysregulation, perfectionism, interoceptive deficits, and body dissatisfaction in patients with anorexia nervosa [ 76 ], corroborating its potential protective effects on mental health and perceived stress. Given the significant negative impacts of both benign gynecological conditions and EC on mental and physical health and perceived stress, further investigation into this link would be beneficial for patient QOL. Order Veillonellales was also negatively correlated to good mental health, validating its role in gut microbiomes of patients with IBS. Flavobacteriales was enriched in patients with good physical health, despite its aforementioned link to bladder cancer [ 75 ]. RF32, also classified as Rhodospirillales, was enriched in patients with good mental health, despite having generally negative effects on human health in the previous literature [ 77 – 79 ]. These effects are not reflected in our study. Order Burkholderiales was depleted when analyzed against increased vaginal symptoms. It has been associated with intestinal inflammation and gut dysbiosis [ 77 ] and lumbar fractures [ 78 ] thus far, though no existing literature connects Burkholderiales in the gut to vaginal symptoms. Further research on enteric Burkholderiales’s impact on vaginal symptoms is needed. Verrucomicrobiales in the gut, found enriched in patients with increased vulvar symptoms and positively correlated to good mental health, has been associated thus far with human longevity and health [ 79 – 81 ]. More research is needed on this order to understand the association to worse vulvar symptoms in patients with EC, despite correlating to good mental health. Given the profound impact that benign gynecological conditions and EC can have on patient QOL, additional research is needed to understand what affects preoperative QOL and how to improve post-hysterectomy QOL. Understanding the microbial composition of both vaginal and rectal environments is the first step toward assessing the microbiome impact on patient QOL. From our analysis, we conclude that microbial composition does indeed correlate to several aspects of patient QOL. Some microbes, such as Porphyromonas and Campylobacter , exhibit the same correlations in both disease groups, challenging previous understandings of these microbes, while others are unique to each disease group. Vaginal microbiota mainly affect sexual interest and vaginal/vulvar health. Rectal microbiota, such as Gastranerophilales, Burkholderiales, and Verrucomicrobiales, affect several aspects of patient-reported QOL including mental, vaginal, and vulvar health, respectively. Lastly, patients with EC exhibit a unique vaginal environment, with high levels of L. iners correlating with worse vaginal symptoms, despite having lower overall species diversity than patients with benign gynecological conditions, even when controlling for menopausal status (Fig.  5 ). Validating these correlations between microbes and symptoms in future cohorts and eventually determining causality and pathophysiology behind their impact on QOL will assist clinicians in better addressing patient wellbeing, through either targeted microbiome therapy or new screenings to prevent symptoms of poor QOL from occurring. Fig. 5 Synopsis of vaginal and rectal microbiota correlations to patient QOL. In both disease groups, vaginal Porphyromonas and Campylobacter were associated with decreased sexual interest, while Dialister B was positively associated with increased sexual interest in benign conditions. L. iners and S. agalactiae were enriched in patients reporting worse vaginal and vulvar symptoms, and patients with EC with higher vaginal symptom scores had lower alpha diversity. Rectal Veillonellales correlated with increased gas and bloating in benign conditions, while Christensenellalaes, Desulfovibrionales, and Flavobacteriales correlated with less gas and bloating in EC. Gastranaerophilales abundance positively correlated with better mental health scores in the entire study cohort. Rectal Verrucomicrobiales and Burkholderiales were associated with vaginal and vulvar symptoms. Created in BioRender. Gautam, N. (2025) https://BioRender.com/v45r769 Synopsis of vaginal and rectal microbiota correlations to patient QOL. In both disease groups, vaginal Porphyromonas and Campylobacter were associated with decreased sexual interest, while Dialister B was positively associated with increased sexual interest in benign conditions. L. iners and S. agalactiae were enriched in patients reporting worse vaginal and vulvar symptoms, and patients with EC with higher vaginal symptom scores had lower alpha diversity. Rectal Veillonellales correlated with increased gas and bloating in benign conditions, while Christensenellalaes, Desulfovibrionales, and Flavobacteriales correlated with less gas and bloating in EC. Gastranaerophilales abundance positively correlated with better mental health scores in the entire study cohort. Rectal Verrucomicrobiales and Burkholderiales were associated with vaginal and vulvar symptoms. Created in BioRender. Gautam, N. (2025) https://BioRender.com/v45r769 Limitations include having a smaller number of participants with a diagnosis of EC to compare to benign conditions, necessitating a combination of EC subtypes despite potential differences in symptomatology. Additionally, given the novelty of this field of research, several of the microbes identified have not been previously studied and their relationship to health and QOL is unknown. To better understand the role of the microbiome in the future, more studies are needed characterizing microbiota found in both vaginal and gut microbiomes. Additionally, more characterization of microbial correlations to QOL are needed to better understand the complex relationship between microbes and effects on QOL.

Endometrial cancer (EC) is the fourth most common cancer among females in the United States (U.S.) [ 1 ], and estimated to be the 4th leading cause of cancer deaths in this population in 2024 [ 2 ]. Elevated body mass index (BMI) and older age are known risk factors for EC [ 3 , 4 ], and rising young adulthood obesity rates are further increasing EC rates [ 4 , 5 ]. It is a highly prevalent disease that significantly impacts patients’ physical, mental, emotional, and sexual quality of life (QOL). Few studies have examined QOL in patients with EC prior to treatment or hysterectomy [ 6 – 8 ]. Benign gynecological conditions such as fibroids, endometriosis, and adenomyosis are also highly prevalent in the U.S. Around 70% of females will develop fibroids in their lifetime [ 9 ], and symptomatic fibroids are the most common indication for hysterectomies [ 10 ]. Additionally, endometriosis and adenomyosis are associated with chronic pelvic pain [ 11 , 12 ] and poor sexual QOL [ 13 ]. While EC and benign gynecological conditions are different in pathophysiology, they both decrease patients’ QOL and are indications for hysterectomy, a physically and psychologically stressful experience. Vaginal microbiomes have been well described in the literature for both pre- and post-menopausal patients. During reproductive-age years, the vaginal microbiome is typically dominated by one or few Lactobacillus species that function to preserve acidity (vaginal pH < 4.5), conferring protection from pathogenic microorganisms [ 14 , 15 ]. The vaginal microbiome during and after menopause typically shifts in composition to lower proportions of Lactobacillus , higher pH, and increased bacterial diversity [ 16 ]. Post-menopausal patients frequently experience vaginal and vulvar atrophy, dyspareunia, and decreased libido secondary to hypoestrogenism [ 17 ]. Several reports have implicated vaginal Porphyromonas and Fannyhessea in pathogenesis of EC [ 18 – 22 ]. Few studies have examined the vaginal microbes involved in uterine fibroid, endometriosis, and adenomyosis pathogenesis or QOL [ 23 – 26 ]. Alterations in gut microbiomes have also been associated with poor QOL [ 27 – 30 ] and gynecological disease [ 31 – 33 ]. Research on gastrointestinal conditions demonstrated a strong link between gut dysbiosis and diarrhea, constipation, and abdominal bloating/pain [ 34 – 36 ]. Additionally, gut microbiota and their metabolites were found to directly modulate the central nervous system in settings of chronic pain [ 37 , 38 ]. Certain microbiota have been found to alter estrogen levels, and hypothesized to contribute to conditions like polycystic ovarian syndrome, endometriosis, EC, and uterine fibroids [ 39 ]. Gut microbial composition impacts both QOL and gynecological processes, though the directionality of the brain-gut-vagina interaction is still unknown. In this study, we investigated the relationship between preoperative patient-reported QOL and vaginal and rectal microbiomes amongst patients with EC and benign gynecological conditions undergoing hysterectomy. Previous analysis in this population by Chase et al. [ 40 ] revealed significant differences in symptoms experienced by patients with benign conditions and EC prior to hysterectomy. Integration of survey and microbiome data will help evaluate the directionality of these processes and whether microbiome modulation is a reasonable therapeutic area of research for improving patient QOL.

Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2