Shifts in gut and vaginal microbiomes are associated with cancer recurrence time in women with ovarian cancer

Many studies investigating the human microbiome-cancer interface have focused on the gut microbiome and gastrointestinal cancers. Outside of human papillomavirus driving cervical cancer, little is known about the relationship between the vaginal microbiome and other gynecological cancers, such as ovarian cancer. In this retrospective study, we investigated the relationship between ovarian cancer, platinum-free interval (PFI) length, and vaginal and gut microbiomes. We observed that Lactobacillus-dominated vaginal communities were less common in women with ovarian cancer, as compared to existing datasets of similarly aged women without cancer. Primary platinum-resistance (PPR) disease is strongly associated with survivability under one year, and we found over one-third of patients with PPR (PFI < 6 months, n = 17) to have a vaginal microbiome dominated by Escherichia (>20% relative abundance), while only one platinum super-sensitive (PFI > 24 months, n = 23) patient had an Escherichia-dominated microbiome. Additionally, L. iners was associated with little, or no, gross residual disease, while other Lactobacillus species were dominant in women with >1 cm gross residual disease. In the gut microbiome, we found patients with PPR disease to have lower phylogenetic diversity than platinum-sensitive patients. The trends we observe in women with ovarian cancer and PPR disease, such as the absence of Lactobacillus and presence of Escherichia in the vaginal microbiome as well as low gut microbiome phylogenetic diversity have all been linked to other diseases and/or pro-inflammatory states, including bacterial vaginosis and autoimmune disorders. Future prospective studies are necessary to explore the translational potential and underlying mechanisms driving these associations.

133 provided in Table 1. This study was approved by the University of Oklahoma Health Sciences 134 Center Institutional Review Board (February 22 nd , 2016, reference #6458) and all participants 135 gave written informed consent for their participation in the study.  246 microbiomes dominated by Escherichia had higher odds of occurring in PFI < 6 months 247 compared to PFI > 24 months (log OR = 2.812, 95% CI 0.267 -5.62, p-value = 0.024, Fig. 2A).
248 Additionally, one of the patients with PFI < 6 months and one of the patients with benign 249 pathology identified with a 'diverse' vaginal microbiome had Escherichia at greater than 20% 250 relative abundance, while no other patients with PFI > 24 months had Escherichia relative 251 abundance above 5% (Fig. 1) Fig. S4). Consumption of antibiotics within 273 the past month was associated with a lack of Lactobacillus-dominance (Fig. 2B); however, this 274 relationship was not significant (log OR = -2.12, 95% CI -5.54 -0.84, p-value = 0.0515).   (Table S4); ultimately there was not sufficient data to assess the relationship 287 between most of these Lactobacillus species and clinical variables. Nevertheless, in individuals 288 with high Lactobacillus abundance (n = 11), we found L. iners at significantly higher abundance 289 in patients with either no gross residual disease or residual disease < 1cm (n = 7), compared to 290 patients with residual disease > 1cm (n = 4, p-value = 0.0359, Fig. 3).
294 However, Prevotella was less common in those with residual disease > 1cm and those with a 295 history of hormonal disease (Fig. 2C), while a highly diverse vaginal microbiome was more 296 common in women over 60 years old and in patients with a history of hormonal disorders, such 297 as thyroid disease (Fig. 2D); but once again, none of these associations were statistically 298 significant (Table S3, p-value > 0.05).   353 inflammatory cytokines IL-1 and IL-6 (55). The low Lactobacillus levels we observed may be 354 related to glycogen availability -nearly 75% of women in this study were post-menopause, and 355 vaginally produced glycogen is known to decrease after menopause; likewise, chemotherapy can 356 inhibit ovarian estrogen production and result in lower glycogen levels (44, 54, 56, 57). More 357 importantly, many women in this study have had at least one ovary surgically removed during 358 initial cancer treatment. Ovary removal leads to decreased estrogen production and thus, a likely 359 decrease in vaginal glycogen levels; however, more research is needed to fully explain the 360 ovarian-estrogen-glycogen dynamic (50). Glycogen abundance may also help explain the 361 relationship between vaginal microbiome cell density and Lactobacillus, as widely available 362 glycogen may encourage a densely colonized Lactobacillus vaginal community due to high 363 nutrient availability (57). Hormone replacement therapy (HRT) can be used to replace estrogen 364 production that is stopped after menopause, and four women (two pre-menopause and two post-365 menopause) were on HRT. Women on HRT or pre-menopause were more likely to have 366 Lactobacillus-dominated communities, but this was not significant (p-value = 0.0744). While we 367 did not document glycogen levels in our study, the positive relationship between likely estrogen 368 presence and Lactobacillus abundance lends credence to the idea that the low proportion of 369 women with Lactobacillus abundance in our study is due to low estrogen, and thus glycogen.
370 The retrospective nature of our study means that we were unable to assess Lactobacillus levels in 371 women with ovarian cancer before they progressed to stage III/IV, or prior to chemotherapy.
372 Therefore, we could not investigate anti-gynecological cancer properties of vaginal 373 Lactobacillus; nevertheless, by comparing patients in this study to similarly aged women without 374 ovarian cancer, we present further evidence that low Lactobacillus levels are more common in 375 women with ovarian cancer (p-value = 0.037).

377
The presence, and size, of residual disease is strongly correlated with decreased survivability in 378 ovarian cancer (7) and therefore our finding that L. iners was at significantly higher abundance 379 (p = 0.0359) in patients with either no gross residual disease or residual disease < 1cm after 380 treatment may point to L. iners as a potential path toward a biomarker. L. iners is a common 381 vaginal bacterium (46, 58, 59) but its role in health and disease is sometimes contradictory (21); 382 L. iners has been found at high relative abundance in low-grade squamous intraepithelial lesions 383 in the cervix but at low abundance in high-grade squamous intraepithelial lesions (26). Yet, 384 another study found L. iners at high abundance in women with normal cytology when compared 385 to women with squamous intraepithelial lesions (60). Other studies have found L. iners to be 386 positively associated with cervical cancer (61, 62) but L. iners is also linked to clearance of HPV 387 (63), which is a causative agent of cervical cancer. Further research is necessary to better 388 understand the role of L. iners in gynecological cancers in general, and in the potential inhibition 389 of gross residual disease in ovarian cancer. 397 months showed greater than 20% relative abundance of Escherichia, as compared to 4.34% of 398 PFI > 24 months and only one of 5 benign cases. The explanation for why Escherichia was 399 significantly more common in patients with platinum-resistant tumors is unclear, and because 400 this study was retrospective, we were unable to track Escherichia abundance before we knew 401 each patient's platinum-sensitivity. One possible pathway is via interactions between the 402 microbiome, immune system, and how platinum-based chemotherapies induce cancer cell death.           with PFI < 6 months had lower phylogenetic diversity than benign and platinum-sensitive patients but this was not a significant result (p-value = 0.18).