Surgical Outcomes in Laparoscopic vs Open Diverting Colostomy for Radiation-Related Rectovaginal Fistula

Introduction

Rectovaginal fistula (RVF) refers to an abnormal connection between the rectum and vagina, which can cause bowel contents or even purulent discharge to leak out of the vagina. In patients with refractory symptoms, hematochezia and tenesmus can occur. Significantly higher healthcare resource utilization and costs were observed in patients with RVF, due to repeated interventions in the management of fistulas [1]. RVF results from a variety of conditions, such as obstetric trauma, radiation for malignancies in the pelvic area, and Crohn disease. RVF leads to significant social, emotional, and physical consequences, which negatively affecting women’s quality of life.

Radiation-induced rectovaginal fistula (RI-RVF) occurs in patients who received radiotherapy for gynecological malignancies, such as cervical or endometrial cancer, as well as for locally advanced rectal cancer [2]. It is estimated that up to 6% of patients receiving pelvic irradiation will develop RVF. The clinical presentation of RI-RVF is similar to that of RVF secondary to obstetric trauma, which is the most common cause of RVF. However, the inflammatory response and tissue fibrosis are much more severe in patients with RI-RVF.

RI-RVF differs significantly from RVF caused by obstetric trauma, Crohn disease, or surgical injury, primarily due to the unique pathophysiology of radiation-associated tissue damage. Radiation therapy triggers chronic, progressive injury characterized by fibrosis, microvascular obliteration, and ischemic necrosis, creating a hypovascular, hypocellular, and hypoxic tissue environment. This contrasts with non-radiation etiologies, in which acute trauma or inflammation often spares surrounding tissue viability. Consequently, RI-RVF exhibits markedly delayed healing, due to impaired angiogenesis and cellular regeneration, leading to higher surgical failure rates than with obstetric or inflammatory fistulas, in which healthier adjacent tissues often permit successful primary repair. Clinically, RI-RVF poses greater challenges due to extensive fibrosis and tissue fragility, necessitating complex reconstructive techniques, such as vascularized flaps, to address deficient wound healing. Additionally, RI-RVF frequently presents years after radiation, compounded by comorbidities such as chronic pain, strictures, or bowel dysfunction, further complicating management. These factors collectively result in poorer prognoses and elevated recurrence risks, underscoring the need for more multidisciplinary, tailored approaches in RI-RVF care than with the more localized and surgically amenable non-radiation-related fistulas.

Fistulas secondary to radiation injury usually open into the high or mid vagina. According to the connection point between the rectum and vagina, RI-RVF can be classified into high rectovaginal fistulas and midzone fistulas [3] (Figure 1). This distinction determines surgical approach; for example, high fistulas can require transabdominal or pull-through procedures, while low fistulas might be managed transvaginally or transanally. Fistulas located >7 cm from the anal verge correlate with better healing outcomes after fecal diversion, such as loop ileostomy. Conversely, low fistulas can necessitate sphincter-preserving techniques [4]. Additionally, fistulas are classified as small (<0.5 cm), medium (0.5–2.5 cm), or large (>2.5 cm). Larger fistulas often require advanced reconstruction, such as myocutaneous flaps or pull-through procedures, rather than primary closure.

Controlling local symptoms and improving patient quality of life is of great importance, since life expectancy is favorable in patients without local recurrence and distant metastasis. A multitude of approaches, including simple fistulotomy, inversion of the fistula, or sliding flap repair, can be used for midzone fistulas. Transabdominal repair has achieved a high success rate for high fistulas. Although promising results have been achieved after the use of the Martius flap or endorectal advancement flap, with or without sphincteroplasty, for the treatment of RVF, the recurrence risk of RI-RVF remains high, due to the post-radiation inflammatory response and tissue fibrosis [4]. With palliation of symptoms, mitigation of regional inflammation and improvement of tissue healing, a diverting colostomy is an effective procedure implemented prior to curative fistula repair or restorative resection [5,6]. In some cases, the fistulas were spontaneously healed after stool diversion [7].

Traditionally, open enterostomy is preferred for patients who previously underwent abdominal surgery or radiotherapy [8]. With the advantages of small abdominal incision and limited use of postoperative analgesics, the minimally invasive approach has been widely accepted in a variety of gastrointestinal surgeries [9,10]. However, laparoscopic colostomy has not been widely applied in patients with RI-RVF, given the intractable abdominal adhesions caused by previous surgery and radiation. Nevertheless, the evidence on the safety and feasibility of laparoscopic colostomy in patients with RI-RVF is lacking. In patients who underwent colorectal surgery and colostomy, the postoperative length of stay and rate of incision site infection were significantly higher in the open procedure group than in the laparoscopic group [11,12]. In this retrospective study, we evaluated the surgical outcomes of laparoscopic and open colostomy in patients from a single center with a diagnosis of RI-RVF.

Material and Methods

PATIENTS:

The data from 74 patients from January 2014 to December 2023 at the Affiliated Hospital of Qingdao University with a diagnosis of RI-RVF were retrospectively analyzed. The primary diseases prior to rectovaginal fistula were cervical cancer, endometrial cancer, and rectal cancer. The inclusion criteria were age ≥18 years, treatment with or without combined chemotherapy prior to surgery, fistula development at least 3 months after completion of radiotherapy, and patient survival at least 6 months after stoma creation. Patients with vesico-rectal-vaginal fistulas (n=1) and intestinal-rectal-vaginal fistulas (n=1) were excluded. Patients received diverting colostomy with simultaneous bowel resection (n=4), and with missing data (n=2) were also excluded.

High RI-RVF is located from the posterior fornix to the middle third of the rectum. Midzone RI-RVF refers to fistulas between the lower third of the rectum and the mid-portion of the vagina. The diameter of the fistula was measured as the distance between the 2 points of the opening along its longer axis in the rectum by rectoscopy or in a CT scan. The size of an RI-RVF is stratified into 3 groups: small (<0.5 cm in diameter), medium (0.5–2.5 cm) and large (>2.5 cm) [13].

Patients in the present study with cervical and endometrial cancer received low-dose rate to the 54 Gy/3 fraction and external beam radiotherapy to the 22 fractions of 2 to 44 Gy, during which 4 doses of cisplatin 50 mg/m2 were added. or intracavitary brachytherapy as a high-dose rate was added to the 35 Gy/5 fraction [4]. A cumulative radiation dose of 84 Gy was administered for less than 8 weeks [14]. Patients with locally advanced rectal cancer received either short-course radiotherapy (2500 cGy in 500 cGy per fraction) or standard fractionated chemoradiation (5000–5400 cGy in 180–200 cGy per fraction) [15]. Patients with different fistula size, location, and radiation doses were randomly assigned to either the open surgery group or the laparoscopic surgery group. According to the approach used for stoma creation, patients were allocated into 2 groups as follows: laparoscopic (n=28) and open group (n=38). The baseline characteristics, surgery data, postoperative recovery, and complications were analyzed for each group.

Surgery time was measured from the start (knife to skin) to the end (last skin suture) of the operation. Postoperative ileus was defined as absence of flatus by day 5 postoperatively and/or placement of a nasogastric tube for decompression [16]. Length of incision was measured in centimeters at skin level. For laparoscopic surgery, incision length was calculated as the sum of each port length. The length of hospital stay was calculated from admission to discharge. Patients meeting the following criteria were allowed for discharge: tolerance of liquid diet, well-functioned stoma, without complications, and independent ambulation, regardless of the healing of fistula. According to Centers for Disease Control and Prevention guidelines [17], surgical site infection was categorized into 3 groups: superficial-incisional, deep-incisional, and spaces/organ deep to the incision. We assessed patients 30 days after surgery for surgical site infection occurrence.

Symptoms of RI-RVF were graded by the modified radiation proctopathy system assessments scale (RPSAS) [18]. Patients scored 1 point if they were free of a specific symptom, in accordance with this scale. Pain score was assessed 24 h after surgery by the numerical pain scale, with 0 indicating “pain free” and 10 indicating “the worst pain” [19]. Patient-controlled intravenous analgesia with pethidine was applied to all patients in both groups. None received epidural anesthesia prior to surgery. Aside from emergency surgeries, the desired stoma sites were marked preoperatively by an attending physician and an experienced enterostomy therapist. The stoma was placed within the lateral half of rectus abdominal muscle. The stoma was checked daily by the enterostomy therapist during patient hospitalization. After discharge, stoma function and any related complications [20] were examined in an outpatient visit every 2 to 4 weeks.

SURGICAL TECHNIQUES:

Eight patients in the open group received the Trephine stoma technique [21], in which the stoma was created via a circular incision at the planned site, without laparotomy. For patients with previous surgeries (n=30), median incision laparotomy with colostomy was used, due to underlying peritoneal adhesion. Twenty-eight patients received laparoscopic colostomy. A 10-mm port was placed in the periumbilical area for the digital camera. The peritoneal cavity was inspected for findings, including adhesions, ascites, and carcinomatosis. Subsequently, two 5-mm ports were inserted in the right upper and lower quadrants separately to identify and mobilize the desired bowel segment for loop stoma creation. For an end sigmoid colon stoma, a 5-mm port was placed in the right upper quadrant of the abdomen, and a 12-mm port was placed opposite to the planned stoma site in the right lower abdomen. A laparoscopic stapler (Echelon Flex Endopath 60-mm standard stapler, USA) was used to transect the sigmoid colon through the 12-mm port. This type of stapler was used across all procedures. The mesentery was mobilized, and the proximal part of sigmoid colon was raised for stoma creation. Transverse colostomy was performed in patients with intractable pelvic adhesion or immobilized sigmoid mesentery; the greater omentum was dissociated from the transverse colon to allow for tension free exteriorization.

In all procedures, a circular skin disk was excised at a previously marked site. Then, a cruciate incision in the anterior rectus fascia was made after division of the subcutaneous tissues. Subsequently, a vertical incision in the posterior rectus fascia was created, following the split of the rectus abdominis muscle. When exteriorizing the divided bowel segment, a “pushing” method from within the abdominal cavity is preferred. The stoma usually extends 2 cm after pulling out and protrudes 0.5 to 1 cm above the skin surface after maturation. It was confirmed that the exteriorized bowel was well-vascularized without twist. The seromuscular layer of the extracted bowel was interruptedly sutured and fixed to the peritoneum with 4-0 Vicryl absorbable sutures (VCP711D 4/0 Johnson & Johnson, USA). Shortly after abdomen closure, the stoma was matured with eversion. The enterocutaneous anastomosis was performed circumferentially with previously described sutures that take a full-thickness bite of the terminal end of the colon and dermal layer of the skin. In the case of loop colostomy, the proximal end of the colon was matured 1 cm above the skin with eversion, while the distal part was matured flush with the skin. No bridge or rod was used, since the loop stoma was externalized without tension [22].

All patients received colostomy only without any type of repair either before or after colostomy creation, or in the same surgery. The reason is that our center has no experience for RVF repair. Neither contrast enema nor CT scan were performed postoperatively to assess fistula closure.

The study protocol was approved by the Ethics Committee of the Affiliated Hospital of Qingdao University and was conducted in line with the Declaration of Helsinki.

STATISTICAL ANALYSIS:

Descriptive analysis with a 95% confidence interval was performed. The distribution of quantitative variables was analyzed. Continuous variables are expressed as means and standard deviation (SD). Categorical data are expressed as frequency and percentage. The data was analyzed by IBM SPSS Statistics 29 for Mac. The differences between the open colostomy group and laparoscopic colostomy group were analyzed by the chi-square test or Fisher exact test for categorical data, and by the t test for continuous data. The Wilcoxon signed-rank test was used for analyzing non-parametric data, such as pain score and RPSAS symptom scores. P values of <0.05 were considered statistically significant.

Results

A total of 66 patients were included in this study: 28 received laparoscopic diverting colostomy and 38 underwent open surgery based on the surgeon’s discretion. The demographic characteristics of the 2 groups of patients are summarized in Table 1. No significant difference was found regarding age, body mass index, smoking status, ASA classification, comorbidities, primary disease for stoma creation, number of previous operations, and RPSAS symptom scores. Fistula location, size, sphincter involvement, and type of radiation exposure were comparable between the 2 groups.

The surgery data are shown in Table 2. All patients received colostomy alone, without combination with fistula repair. The average surgery time and estimated blood loss in the laparoscopic group were comparable to those of the open group. There was no statistically significant difference in blood transfusion rate between the 2 groups. Twenty-three patients in the laparoscopic group received sigmoid colostomy and 5 patients received transverse colostomy. In the open group, sigmoid colostomy was performed in 16 patients and transverse colostomy was performed in 22 patients. The length of incision in the laparoscopic group was significantly shorter than that in the open group (5 cm vs 13 cm, P<0.01).

As for postoperative recovery and complications (Table 3, Figure 2), there were no significant differences between the 2 groups in time to pass first flatus or defecation, time to resume liquid or normal diet, ICU admission, and reoperation. However, the postoperative length of stay was 2 days shorter in patients who underwent laparoscopic stoma creation (6 days vs 8 days, P=0.0165). Patients in the laparoscopic group had much lower pain scores [3] 24 h after surgery and fewer days requiring analgesics (1.5±0.58), P<0.01. A higher incidence of surgical site infection was found in the open group (5/38), P=0.04. Other perioperative complications, such as ileus, hemorrhage or hematoma, and perforation were comparable between the 2 groups. There were no significant differences in stoma-related complications.

Discussion

RVF is a rare but debilitating condition that severely influences patients’ quality of life. An increasing number of patients with gynecological and anorectal malignancies have received radiation therapy during the past decades, of which at least 4% develop RI-RVF [4]. When combining radiation with chemotherapy and other oncological treatment, a remarkable increase in disease-free survival was observed. Hence, the number of patients who could develop RI-RVF has grown over the past years [23]. The incidence of RI-RVF in patients with cervical and endometrial cancer treated with radiotherapy is estimated to be 3% [24]. It is generally accepted that patients receiving doses above 70 Gy are more likely to sustain long-lasting damage to surrounding tissue [25]. In contrast to brachytherapy, external beam radiation tends to increase the likelihood of developing radiation proctitis or even rectovaginal fistula [26]. Conditions that presumably cause microvascular injury and tissue ischemia, such as smoking, atherosclerosis, and diabetes mellitus, also contribute to post-radiation inflammation.

Ionized radiation can directly break double-strand DNA and compromise the integrity of the phospholipid bilayer. Furthermore, the free radicals generated from the ionization process result in oxidative stress injury [25]. The mucosal stem cells lining the gastrointestinal tract are highly susceptible to radiation injury, followed by desquamation, ischemia, fibrosis, and necrosis [27]. Subsequently, full-thickness bowel ischemia and ulceration occur, leading to fistula formation between the rectum and vagina [28]. Radiation-induced endothelial damage and microvascular thrombosis create chronic ischemia, while persistent inflammation further impedes tissue regeneration. This dual insult predisposes repaired tissues to necrosis, infection, and fistula reformation. RVF generally develops spontaneously 2 years after radiotherapy [4]. The radiosensitive epithelial and mucosal endothelial cells lose the ability to divide. Chronic radiation injury triggers excessive collagen deposition, leading to rigid, non-pliable scar tissue. Fibrosis reduces tissue elasticity, compromises blood supply, and creates a hostile microenvironment for surgical repair. This stiffness also increases tension on suture lines, raising the risk of breakdown. The passage of stool through the rectum and vagina further impairs the regeneration capacity of crypt cells. Additionally, the level of lactobacillus in the vagina decreases as fecal contamination persists. This alternation in vaginal microbiome can also play a crucial rule in impaired RVF healing [29]. Radiation causes obliteration of normal anatomical boundaries, through adhesions and fibrosis, distorting dissection planes. Surgeons face difficulty identifying and separating rectal, vaginal, and pelvic structures, elevating the risk of intraoperative injury (eg, bowel, bladder) or incomplete fistula closure. In summary, these pathophysiologic changes in irradiated tissues significantly reduce the success rate of surgical repair for RI-RVF [30].

RVF is well characterized by passage of stool or flatus through the vagina, making the diagnosis not that difficult. For simple fistula in patients with mild symptoms, bimanual palpation or enema can facilitate the diagnosis [31]. Patients with minimal symptoms can benefit from conservative treatment, such as bathing, wound care, and antibiotics, as needed [2] Otherwise, surgical repair should be considered in those with severe symptoms, such as anorectal pain, bleeding, and tenesmus. Low fistulas can be reconstructed via vaginal, perineal, or anal approaches, with advancement flaps, sphincteroplasty, and tissue interposition being the choice of treatment, while high fistulas require transabdominal repair [32]. Depending on the procedures used, the success rate for surgical repair varies. Overall, the management of RI-RVF remains challenging due to the high risk of recurrence. According to a study by Corte et al [33], early diverting stoma can significantly increase the likelihood of fistula healing after local procedures. This agrees with the study conducted by Barugola et al, in which a diverting stoma was shown to balance vaginal microbiota and therefore ameliorate repair outcomes [34]. Because of reducing the pressure gradient between the vagina and rectum, fecal diversion is still recommended as the first step to relieve symptoms after fistula formation [2]. Fecal diversion seems to improve women’s daily hygiene and quality of life. However, whether these benefits overshadow the influence of having a stoma on their psychological status remains controversial. In brief, the introduction of a diverting stoma has a positive impact not only on patients’ comfort but also on fistula healing.

Laparoscopy can be used for stoma creation, with detailed intra-abdominal inspection and flexible bowel selection [35]. However, this approach is not feasible in patients with complete bowel obstruction. Additionally, the open approach is more likely to be performed in emergency cases. Selection bias is inevitable in studies that enroll patients under those conditions. Patients diagnosed with RI-RVF rarely present with complete bowel obstruction. All cases in the present study underwent elective surgery. Generally, the surgeon can consider the open approach for patients who have undergone more than one previous surgery. Nevertheless, in our subgroup of patients that received at least one surgery, there was no difference between the laparoscopic and open group. In the laparoscopic group, all operations were performed successfully by highly skilled laparoscopic specialists, without conversion to laparotomy. The major reason for conversion to open surgery is intractable intestinal adhesion. Patients with RI-RVF receive surgery and radiotherapy in the pelvic area. The adhesion within the peritoneal cavity is relatively light, which explains the low conversion rate achieved in our study.

As for fecal stream diversion, the stoma can theoretically be created anywhere along the gastrointestinal tract, with the distal ileum and colon the most common sites. Diverting ileostomy was not included in this analysis for the following reasons. First, the perioperative management of a colon stoma is different from that of high-output ileostomy. The incidence of dehydration, small bowel obstruction, and skin irritation was higher with ileostomy than with colostomy [36]. Second, in patients with a competent ileocecal valve, loop ileostomy raises the potential risk of developing closed-loop obstruction [11]. Diverting loop ileostomy is the appropriate procedure for ileorectal vaginal fistula. However, it was excluded from this study. Loop colon stomas have the operative benefits of reduced blood loss, avoidance of mesentery dissection, and colon transection. However, this type of stoma does not divert feces completely, which can affect the healing of a fistula.

In our study, laparoscopic diverting colostomy was associated with a shorter hospital stay and lower rate of incision-related complications than the open procedure, which is in line with a propensity-matched cohort study conducted by Ivatury et al [11]. The increased length of incision needed for open colostomy poses a higher risk for wound infection, which in turn lengthens hospital stays. Other factors contributing to shorter hospital stays include a shorter length of incision and lower requirement for analgesics. Trephine colostomies accounted for only a small percentage of cases in the open group. Although the laparoscopic approach takes more time to set up, it indeed shortens the postoperative stay by 2 days, which results in a saving in costs and hospital resources to some extent. First, fewer inpatient days free up beds for other patients and cut costs (eg, nursing, utilities). Second, lower opioid use reduces pharmacy expenses and avoids costs of managing opioid-related adverse events. Third, fewer complications mean fewer interventions, such as imaging, wound care, and reoperations. Additionally, shorter length of stay could accelerate patient turnover, which improves surgical scheduling and resource allocation. Stoma creation has been viewed as an independent risk factor for surgical site infection [37]. Obviously, patients who underwent open surgery were more prone to incisional infection than those with minimally invasive procedure on account of wound length, as shown in this study.

Prior studies have shown no significant differences in operative time and morbidity rates between laparoscopic and open stoma creation. However, patients who underwent laparoscopic surgery began functioning earlier and had shorter postoperative hospitalization [38]. In patients with RI-RVF, restorative resection with pull-through coloanal anastomosis and prophylactic colostomy demonstrated superior symptom relief, compared with simple colostomy. However, stoma reversal was only successful in 30% of the resection group, highlighting challenges in restoring bowel continuity after radiation damage [5]. Simple colostomy remains the safest option for RI-RVF in patients with poor general health or extensive radiation fibrosis. A study of 50 patients with RI-RVF showed that 96% underwent fecal diversion, with healing rates influenced by factors such as fistula distance from the anal verge (>7 cm correlated with better outcomes) [4].

There are some potential limitations in our study. First, the non-randomized, retrospective design introduces risks of selection bias and residual confounding. The choice between laparoscopic and open stoma creation was based on surgeon discretion rather than on standardized criteria, potentially leading to systematic differences between groups. For example, surgeons may have preferentially selected laparoscopy for patients with less complex anatomy or lower comorbidity burdens, while reserving open approaches for patients at higher risk. Such confounding by indication could distort comparisons of outcomes, such as complication rates and recovery times, between approaches. Additionally, unmeasured confounders, such as variations in preoperative radiation dose, prior pelvic surgeries, and nutritional status, can further bias results. Second, the short follow-up period of 6 months limits the ability to capture delayed complications, such as parastomal hernias or stoma-related mechanical issues, which often manifest beyond 1 to 2 years after surgery [39]. These risks underestimate long-term morbidity and misrepresent the true safety profile of either technique. Furthermore, critical outcomes, such as stoma reversal success rates and quality-of-life metrics, require extended observation to be assessed fully. Third, the single-center, retrospective nature of the study restricts generalizability. Institutional biases, such as standardized surgical protocols or homogeneous patient demographics, may not reflect broader clinical practices or diverse populations. The reliance on retrospective data also raises concerns about information bias, including incomplete documentation of intraoperative variables, such as adhesion severity and radiation fibrosis extent, or subjective interpretation of outcomes. Finally, while all surgeons were experienced in colorectal surgery, interoperator variability in technique, such as port placement and mesh use for hernia prevention, could influence outcomes. The small sample size precluded propensity-score matching or subgroup analyses to adjust for baseline imbalances, limiting statistical power to detect clinically meaningful differences. These limitations underscore the need for prospective, multicenter randomized trials with standardized protocols, long-term follow-up, and rigorous adjustment for confounders to validate these findings.

Conclusions

For patients with a diagnosis of RI-RVF, laparoscopic diverting colostomy was feasible and did not increase stoma-related complications and postoperative morbidities in the present study. Moreover, laparoscopic surgery was associated with shorter length of hospital stay and lower incidence of surgical site infections than was the open procedure. A prospective study with larger sample size and longer follow-up period should be conducted to verify our findings.