The Microcirculation of Vaginal Tissue in Women with Obstetric Vesicovaginal Fistula and Short-Term Effects of Surgical Repair on Microvascular Parameters

Worldwide, an estimated 2.0–3.5 million women are currently living with an obstetric fistula. An obstetric fistula is an abnormal opening between two hollow organs in the female pelvis [1‒4]. Depending on the location, obstetric fistulas may result in urinary or fecal incontinence, both severely affecting the physical, mental, social, and economic well-being of patients [5, 6]. An obstetric fistula is the direct result of obstructed labor, which is a complication during childbirth where the fetus cannot pass through the vaginal birth canal. The continuous pressure of the fetus on the bladder, vagina, and rectum causes tissue ischemia, necrosis, and eventually fistula formation [1].

Ischemic trauma results in a lack of viable tissue that can be used for surgical closure of the fistula [5]. This makes obstetric fistula repair challenging, and it may also negatively affect the healing process after surgical closure. In an effort to understand the extent of the ischemic trauma, we aimed to assess the microvasculature of the tissue surrounding vesicovaginal fistulas (VVFs) and to investigate the effect of surgical repair on the vaginal microvasculature. We used handheld vital microscopy (HVM) with incident dark-field (IDF) imaging to non-invasively visualize the microvasculature before, during, and 2 weeks after VVF repair surgery [7]. This technique has been used to assess the vaginal microcirculation before, but it has not yet been used to assess the microvasculature in women with obstetric VVF [8‒13].

We hypothesized that the vaginal microcirculation surrounding fistulas was compromised beyond the extent of the fistula itself. Insights may help to better understand the viability of tissue surrounding fistulas, the regenerative capacity of vaginal tissue after extensive ischemic trauma, and the effects of surgery in an ischemic environment. Ultimately, this will lead to improved treatment strategies.

This prospective, observational pilot study was conducted in the Fistula Care Centre in Lilongwe, Malawi. Ethical approval was granted by the Malawian National Health Sciences Commission (NHSRC #2371). Women scheduled for surgical repair of a VVF were eligible if they were above the age of 18 and were excluded if they had only a rectovaginal fistula or a ureteric injury. Patient characteristics including age, months with VVF, extent of scar tissue, surgical history, human immunodeficiency virus (HIV) status, and the type of fistula according to the Goh classification were collected [14]. The sample size of this study was based on convenience. All participants provided oral and written consent before enrollment.

The surgical procedures were performed with participants in lithotomy position under spinal anesthesia. As part of standard care, all participants received antibiotics before surgery. Standard surgical closure techniques with or without the use of vascularized surgical flaps were used for VVF repair using a transvaginal approach [15, 16]. After surgery, participants received vaginal packing for 24 h and an indwelling urinary catheter for 10–14 days.

The vaginal microcirculation was assessed using the CytoCam (Braedius Medical, Huizen, the Netherlands), which is a HVM device that incorporates IDF imaging [7, 8]. IDF imaging visualizes the microvasculature by emitting green light in short illumination pulses of 2 ms. The light is absorbed by hemoglobin in erythrocytes and scattered back by surrounding tissue. This allows real-time visualization of individual erythrocytes flowing through the microvasculature, representing the functional microcirculation. The CytoCam is connected to a laptop that shows the real-time footage of the microcirculation and allows storage of short image sequences that can be analyzed later.

All measurements were performed by a researcher with extensive experience in performing microcirculation measurements. Imaging was performed at three time points in all participants: before surgery (T0), at the time of fistula closure (T1), and at the time of catheter removal post-operatively (T2, typically 10–14 days after surgery). The tip of the CytoCam was covered with a sterile disposable cap and the camera was adjusted for optimal focus and contrast before recording. The CytoCam was placed on the vaginal tissue perpendicularly. Image sequences of 3 s were recorded from the vaginal tissue surrounding the fistula in a clock-wise fashion at four locations, i.e., 12 (anterior wall), 3 (left lateral wall), 6 (posterior wall), and 9 (right lateral wall) o’clock in a 1-centimeter-wide concentric circle. To prevent pressure artifacts (closing the microvasculature by too much pressure of the device on the tissue), the tip of the CytoCam was placed gently on the tissue of interest.

The primary outcome of this study was microvascular flow in the tissue surrounding the fistula. Microvascular flow of all individual image sequences was scored as present or absent and percentages of measurements with present microvascular flow were calculated. Images sequences were assessed by two individual researchers trained for the assessment of the microcirculation. Differences between the two assessors were resolved by consensus among the two.

Secondary outcomes were (1) vaginal angioarchitecture in the tissue surrounding the fistula, (2) fistula closure, and (3) urinary continence.

Normal vaginal angioarchitecture has been described before. Weber et al. [10] introduced a classification system for the assessment of the vaginal microcirculation, incorporating score 1 (capillary loops), score 2 (capillary loops and vascular networks), and score 3 (vascular networks, no capillary loops) (Fig. 1). The angioarchitecture score relates to the condition of the vaginal epithelium; healthy, non-atrophic vaginal epithelium is characterized by capillary loops, and thin, atrophic vaginal epithelium is characterized by vascular networks. In our study, we used the Weber classification to assess all individual image sequences. If an image sequence could not be scored according to this classification, it was classified as abnormal. The percentage of measurements with normal angioarchitecture was calculated. In addition, the percentage of measurements with capillary loops (scores 1 and 2 combined) was calculated, since capillary loops indicate healthy, non-atrophic epithelium, which would be expected in normoestrogenic premenopausal women.

Fistula closure was evaluated by vaginal examination and a postoperative dye test. The bladder was infused with methylene blue and leakage into the vagina was assessed during vaginal examination. When no leakage was observed, the fistula was considered “closed.” The physician that performed the postoperative dye test was blinded for the microvasculature findings at the time of catheter removal. Urinary continence was evaluated using an objective pad weight test: participants received a pre-weighted pad and were asked to continue with their daily activities. After 1 hour, the pad was weighed and the difference was calculated to quantify urinary incontinence [17].

Descriptive statistics were used to present the demographic variables. Analysis was performed with the use of IBM SPSS Statistics for Windows, version 26.0 (IBM Corp., Armonk, NY, USA). All data were assessed for normality and were reported with appropriate measures of central tendency. The Wilcoxon signed-rank test was used to compare measurements before and after surgery. To assess whether microvascular parameters (flow and angioarchitecture) relate to clinical outcomes, we evaluated these microvascular parameters separately for participants with failed and successful fistula repair (based on fistula closure and urinary continence). A p value <0.05 was considered statistically significant.

Seventeen women undergoing VVF repair were included in this study. Baseline and fistula characteristics of all included participants are shown in Table 1. The mean age was 37.8 (standard deviation 15.3) years. Women had a fistula for a median duration of 8.5 [interquartile range 4–36] months before undergoing surgery. Two participants were HIV positive with a CD4 count >400 cells/mm³. Two participants had a prior attempted VVF repair. One participant had a concomitant rectovaginal fistula, the other participants solely had a VVF. Two participants with severe scar tissue underwent surgical repair with the use of a Singapore fasciocutaneous flap, whereas the other participants received VVF repair without the use of vascularized surgical flaps.

Before dissection (T0), the average percentage of measurements with present microvascular flow was 83.8% (Table 2). At fistula closure (T1), microvascular flow was present in 83.9% of the measurements, which increased to 93.8% at catheter removal (T2). There was a trend of improved microvascular flow between T0 and T2, although this was not significant (Table 2). The measurements per patient per time point are shown in online supplementary Table 1 (for all online suppl. material, see https://doi.org/10.1159/000534066).

Before surgery (T0), the vaginal angioarchitecture was normal in 75.8% and characterized by capillary loops (angioarchitecture score 1 and 2) in 59.6% of measurements (Table 2). At fistula closure (T1), 69.4% of measurements had a normal angioarchitecture with 51.9% demonstrating capillary loops. At catheter removal (T2), 89.1% of measurements showed normal vaginal angioarchitecture with capillary loops in 82.3%. Compared to measurements before surgery (T0), the presence of capillary loops was significantly higher at catheter removal (p = 0.04). The image sequences that were scored as abnormal were either “hypovascular” or showed hematomas (Fig. 2). Hypovascular images showed little to no functional microvasculature (i.e., no flow of erythrocytes) (Fig. 2a, b). Hematomas were observed as black areas without flow (Fig. 2c, d).

In 6 cases, we were able to visualize the microcirculation of the bladder through the fistula. Compared to vaginal microvasculature, the bladder showed a different angioarchitecture with a circular organization of capillaries (Fig. 3).

Fourteen (82.3%) participants had a closed fistula after surgical repair. There was one participant with severe urethral incontinence despite a closed VVF.

Microvascular parameters did not relate to clinical outcomes; participants with absent microvascular flow or abnormal angioarchitecture had a healed fistula and, in turn, participants with failed repairs did not necessarily have abnormal microvasculature at any given time point. Both women that had undergone prior fistula repair had normal microvascular flow and vaginal angioarchitecture before and after surgery. However, for one of these participants the repair was unsuccessful again (no fistula closure). The two women that underwent Singapore fasciocutaneous flap repair because of extensive scarring both showed abnormal vaginal angioarchitecture 2 weeks after surgery. One of these participants had a closed fistula and one did not. Of those with moderate scar tissue (clinical observation, n = 7), two participants had abnormal angioarchitecture 2 weeks after surgery and one had a failed surgical repair.

Our study demonstrates that the microvasculature surrounding VVFs is functional in most cases before, during, and after surgical fistula repair. This suggests that ischemic trauma is present, but not as severe as presumed. Our findings may reflect the great regenerative capacity of vaginal vasculature in (young) women. Though not statistically significant, we observed an improvement of vaginal angioarchitecture and an increase in capillary loops, indicating healthy non-atrophic vaginal tissue. This suggests that surgery in a (post)-ischemic vaginal environment may improve the vitality of the vaginal tissue. Microvascular parameters did not relate to clinical outcomes.

When labor is obstructed for a prolonged time, the pressure of the fetal head on the maternal pelvis causes tissue ischemia, necrosis, and fistula formation in the damaged tissue of the bladder and vagina [1, 5, 18]. The fistula is the epicenter of the ischemic injury, but it has long been thought that the surrounding tissue was also damaged, hampering adequate surgical repair and increasing the risk of surgical failure. Previous studies have shown that the vaginal microcirculation of healthy women is adequately perfused and has normal angioarchitecture [10]. In the current study, microvascular flow was absent in 20% of measurements and vaginal angioarchitecture was abnormal in 25% of measurements before surgery. Since functional microcirculation represents functional tissue that receives oxygen and nutrients, the absence of normal microvasculature indicates damaged tissue in which case wound healing may also be compromised [8]. However, the majority of measurements showed normal microvascular flow and angioarchitecture. In our opinion, this indicates that ischemic trauma after obstructed labor causes permanent local damage (i.e., the fistula), but the damage to the surrounding tissue is relatively mild and/or is restored by the great regenerative capacity of vaginal tissue. HVM measurements around “fresh” fistulas, shortly after obstructed labor, could provide more insight into the extent of damage to the tissue surrounding fistulas and the regenerative capacity of the damaged tissue.

Surgical repair of fistulas is based on the principle of removing (avascular, necrotic) tissue surrounding the fistula and approximating the fresh wound edges to allow wound healing of vital tissue in a tension-free way. However, directly after surgical closure of the wound, 20% of measurements still showed absent microvascular flow. This may indicate that not enough tissue was removed, or be a consequence of short-term surgically induced trauma to the microvasculature. A previous study on the effect of vaginal prolapse surgery on the microcirculation of the vaginal wall demonstrated that shortly after surgery, microvascular flow was absent or reduced as well [8]. In the current study, we also demonstrated that surgery of fistulas does not further deteriorate vascularization and even seems to improve it. We observed an increase in microvascular flow to almost normal levels, 2 weeks after fistula repair [10]. The vaginal angioarchitecture also improved over time, with a significant increase in the presence of capillary loops, indicating the presence of healthy epithelium. In our opinion, this demonstrates that surgical closure of the fistula improves the condition of the vaginal tissue. Potential reasons for this improvement may be the removal of the necrotic tissue by dissection, as well as less inflammation and irritation of vaginal tissue that was possibly caused by continuous leakage of urine through the fistula into the vagina.

This study reports on the microvasculature of obstetric VVFs and describes the microvasculature of the bladder and vascularized flaps. The study is limited by the lack of a control group and the relatively small sample size. However, previous studies on vaginal IDF imaging have demonstrated that reliable measurements can be performed in small samples with good inter-observer reliability and made the comparison to healthy volunteers possible [10]. Last, the focal depth of the CytoCam is limited to 300 μm, and consequently, we could not evaluate the microvasculature of the deeper layers of the vagina. Nevertheless, we consider the status of the more superficial vasculature to be representative of the vascularization of the deeper layers of the vaginal wall. In addition, there are currently no alternative techniques available that allow imaging of the microvasculature of deeper layers in this detail.

Microcirculatory assessment may benefit the evaluation of innovative approaches for VVF repair, such as angiogenic-inducing interventions. Multiple studies have evaluated the effect of platelet-rich plasma application in the surgical treatment of VVF [19‒21]. Medvedev et al. [19] assessed the neovascularization in the tissue surrounding VVFs after injection of platelet-rich plasma using biopsies to collect histological data on the microvascular density. Using HVM, these data could be obtained non-invasively and longitudinally without the need for traumatic tissue sampling that could negatively affect tissue healing. IDF imaging could also be useful in the evaluation of angiogenesis and tissue regeneration in other (innovative) approaches for VVF repair. Surgical innovation for VVF repair focusses on improving the healing conditions in a post-ischemic environment or providing new viable tissue in case of extensive tissue damage. Estrogens and estrogen-releasing hydrogels show promising effects on wound healing and this effect can be evaluated with IDF imaging [11, 22‒24]. However, in light of the results of this study, i.e., relatively good vasculature, it is questionable whether angiogenic-inducing strategies will be beneficial to surgical outcomes. Other potential promising treatment options for VVF repair include tissue-engineered biomaterials (e.g., amnion) or tissue sealants (e.g., fibrin glue, cyanoacrylate, autologous cell injection, or other bioglues), although most of these innovations are still in a preclinical phase [25‒30].

Although the vaginal microcirculation in women with VVF is compromised, extensive ischemic damage is not observed in the tissue surrounding the fistula. This may indicate resilience and great regenerative capacity of the vaginal vasculature in young women, or less extensive damage than has previously been presumed. Surgical fistula repair seems to further improve the vasculature rather than negatively affect it, which suggests surgery is often a good option for these patients. Future studies have to determine whether clinical and surgical outcomes are related to microvascular status and whether further improvement of the vascularization generates better outcomes.

We thank all women who participated in this study. In addition, we thank the Fistula Care Center in Malawi and its employees for facilitating us to perform this study.

Ethical approval was granted by the Malawian National Health Sciences Commission (NHSRC #2371). All participants provided oral and written informed consent before enrollment in this study.

The authors have no conflict of interest to declare.

No funding was received for this research.

L.P.M.: formal analysis, investigation, writing – original draft, and visualization. A.W.K.: conceptualization, writing – original draft, and supervision. Y.P.L.: formal analysis, visualization, and writing – editing and review. G.N. and E.C.: data collection and writing – editing and review. J.-P.W.R.R. and R.J.P.: conceptualization, writing – editing and review, and supervision.

Lennart P. Maljaars and Arnoud W. Kastelein contributed equally to this work.

Archived data shall be made available according to AMC policy on an individual basis. The AMC promotes to openly share the research data by using AMC Figshare. Interested third parties can contact the lead investigator and request data for research purposes conducted according to international and national guidelines. AMC Figshare is a trusted repository for the accessibility of the metadata and has a scheme based on international standards. Further inquiries can be directed to the corresponding author.