Article Text

Download PDFPDF

Trends and current aspects of reconstructive surgery for gynecological cancers
  1. Matteo Loverro1,
  2. Alessia Aloisi2,
  3. Lucia Tortorella3,
  4. Giovanni Damiano Aletti4,5 and
  5. Amanika Kumar6
    1. 1Department of Gynecology and Obstetrics, Policlinico Universitario Agostino Gemelli, Rome, Italy
    2. 2Department of Gynecology, European Institute of Oncology, IEO, IRCCS, Milan, Italy
    3. 3Department of Women, Child and Public Health Sciences, Fondazione Policlinico Universitario A. Gemelli IRCCS, Rome, Italy
    4. 4Department of Gynecology, European Institute of Oncology, Milan, Italy
    5. 5Department of Oncology and Hemato-Oncology, University of Milan, Milan, Italy
    6. 6Mayo Clinic Minnesota, Rochester, New York, USA
    1. Correspondence to Dr Amanika Kumar, Mayo Clinic Minnesota, Rochester, New York, USA; kumar.amanika{at}


    Gynecologic cancers can lead to gynecologic tract destruction with extension into both the gastrointestinal and urinary tracts. Recurrent disease can also affect the surrounding bony pelvis and pelvic musculature. As opposed to advanced ovarian cancer, where cytoreduction is the goal, in these scenarios, an oncologic approach to achieve negative margins is critical for benefit. Surgeries aimed at achieving a R0 resection in gynecologic oncology can have a significant impact on pelvic anatomy, and require reconstruction. Overall, it appears that these types of radical surgery are less frequently performed; however, when required, multidisciplinary teams at high-volume centers can potentially improve short-term morbidity. There are few data to examine the long-term, quality-of-life outcomes after reconstruction following oncologic resection in advanced and recurrent gynecologic cancers. In this review we outline considerations and approaches for reconstruction after surgery for gynecologic cancers. We also discuss areas of innovation, including minimally invasive surgery and the use of 3D surgical anatomy models for improved surgical planning.

    In the era of ‘less is more’, pelvic exenteration in gynecologic oncology is still indicated when there are no other curative-intent alternatives in persistent or recurrent gynecological malignancies confined to the pelvis or with otherwise unmanageable symptoms from fistula or radiation necrosis. Pelvic exenteration is one of the most destructive procedures performed on an elective basis, which inevitably carries a significant psychologic, sexual, physical, and emotional burden for the patient and caregivers. Such complex ultraradical surgery, which requires removal of the vagina, vulva, urinary tract, and/or gastrointestinal tract, subsequently needs creative and complex reconstructive procedures. The additional removal of sidewall or perineal structures, like pelvic floor muscles/vulva, or portions of the musculoskeletal pelvis, and the inclusion of intra-operative radiation further complicates reconstruction. This review paper will focus on the reconstruction aspects following pelvic exenteration, including options for urinary tract restoration, reconstruction of the vulva and vagina, as well as how to fill large empty spaces in the pelvis. While the predominant gastrointestinal outcome after exenteration in gynecologic oncology is an end colostomy, we also present some novel new options for gastrointestinal tract reconstruction at the end.

    • Pelvic Bones
    • Pelvic Floor
    • Surgical Flaps
    • Cervical Cancer

    Statistics from

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Scope of the Problem

    With the expansion of systemic treatments, such as immunotherapy or targeted therapies, and improved delivery of less toxic radiotherapy, we have experienced a significant increase in cancers survivors after primary treatment, with subsequent decreased need for large destructive oncologic surgeries. In fact, it appears that the rate of pelvic exenteration for gynecologic malignancy is decreasing (exenterations for gynecologic cancer decreasing from n=99 in 2005–2010 to only n=10 in the 2011–2018 period in the National Surgical Quality Improvement Program (NSQIP)), with overall only about 15% of exenterations within the NSQIP database being performed for gynecologic cancers.1 2

    However, the characteristics of patients who undergo pelvic exenteration has changed over time. The population worldwide, especially in developed countries, has progressively become older and more obese, adding to the challenges in peri-operative management. Second, the selection criteria for patients who would benefit from pelvic exenteration have expanded, now including those with central lesions and also those who would require a laterally extended endopelvic resection. Pelvic exenteration is still indicated at times for palliative purposes, when disease-related symptoms are not manageable with other therapeutic options.

    While the procedure has undergone technical modifications with the use of new surgical devices and better peri-operative care which may be associated with improved surgical outcomes, this radical surgery is still associated with significant morbidity. Multiple studies reported higher morbidity and mortality rates with more complex surgery, including the need for perineal and infralevator resection, musculoskeletal resection with large space defects, and urinary reconstruction.3 Especially concerning is that most pelvic exenterations are still performed in low-volume centers with less than one exenteration per year.4 In order to optimize peri-operative, short-term and long-term outcomes and quality of life, it is of utmost importance to treat these patients in high-volume centers in a highly specialized, dedicated multidisciplinary setting.

    Considering that, undoubtedly, one of the most difficult portion and long-term life-changing aspect of pelvic exenteration and laterally extended endopelvic resections is represented by the reconstruction, several large-volume centers, including those of the authors, have developed dedicated multidisciplinary services to meet the needs of these patients. At Mayo Clinic, for example, surgical services such as orthopedic surgery, plastic surgery, urology, and vascular surgery are included as part of the surgical team. In addition, there is a broad supportive service team, including wound care nurses, hyperbaric oxygen, radiation oncology, psycho-oncology, palliative medicine, menopause management, social work, financial counselors, and patient navigators. This extensive team serves to address the full scope of needs for the patient and their caregivers, recognizing the support needed during the pre-surgical and post-surgical period, and not only in the direct peri-operative stay. The colorectal group in Southern Australia have published data on a similar dedicated service, demonstrating positive short-term outcomes, including increased R0 resection, when surgery is performed by specialized multidisciplinary teams at high-volume centers.5

    In addition to specialized high-volume centers with multidisciplinary dedicated care teams, the application of Enhanced Recovery After Surgery principles and minimally invasive surgery approaches also has the potential to improve peri-operative outcomes and quality of life. In fact, a pooled analysis of several retrospective studies demonstrated less blood loss and less 30-day morbidity and, not surprisingly shorter hospital length of stay in the group receiving minimally invasive surgery.6 Fewer specific data in gynecologic oncology are available, but existing evidence suggests the feasibility of minimally invasive surgery with low conversion rates while acknowledging that patient selection is critical. Specifically, patients with laterally extended disease, muscle and bone involvement, or need for a large myocutaneous flap may not be good candidates for minimally invasive surgery.7 8

    When comparing different reconstructive approaches after pelvic exenteration, especially in gynecologic oncology, there is very limited evidence, from either the surgical or quality-of-life standpoint. Moreover, given that conducting a randomized controlled trial or even retrospective matched studies in this setting is extremely difficult due to multiple factors, such as tumor type and spread, patient morbidities, previous oncologic and surgical treatments that may limit treatment options, we lack high quality randomized data and must rely on non-randomized, mostly retrospective, institutional and database series. The aim of this review is to give a critical analysis of the current literature, in order to help gynecologic oncologists in the decision-making process between different reconstructive approaches, including patient selection.

    Urinary Tract Reconstruction

    Urinary tract reconstruction represents a fundamental part of pelvic exenteration. Over the last century, many techniques have been described all over the world, which can be mainly divided into either incontinent and continent reconstructions, or urinary diversions that can be directly exteriorized to the skin or indirectly through a digestive segment such as wet colostomy or ureterosigmoidostomy (see Gastrointestinal section).

    Incontinent Conduit

    Incontinent conduit is a urinary diversion with low surgical complexity, comparatively shorter operative time, and low complication rate compared with other urinary reconstructions.9 10 Ileal conduit, first described by Bricker in 1950,11 continues to be the most commonly performed urinary reconstruction in gynecologic oncology.9 10 The conduit is composed of a 15–20 cm segment of isoperistaltic ileum, which is harvested at least 15 cm away from the ileocecal valve, to prevent vitamin B12 and bile salt malabsorption. One end of the segment is obliterated and the ureters are separately anastomosed to its antimesenteric side; the other end of the conduit is exteriorized to the skin as a urostomy. In order to try to prevent anastomotic strictures (which occur in up to 14%),9 10 Wallace proposed two alternative techniques for performing the uretero-enteric anastomoses. He proposed to join and anastomose the ureters either head-to-head (Wallace I)12 or head-to-tail (Wallace II)13 (figure 1). The superiority of this technique remains to be demonstrated; in a meta-analysis that compared the outcomes of Bricker’s technique with those of Wallace’s technique, uretero-enteric stricture rate was similar (3% vs 2%, respectively).14 Another benefit of the Wallace technique is that there is only one uretero-enteric anastomosis. Besides stricture, potential complications of incontinent conduits include urinary tract infection (16–23%), lithiasis (4.5–15%), peristomal skin complications, or stoma issues, such as stenosis, retraction, prolapse, or parastomal hernia (up to 50%).9 10 14

    Figure 1

    Options for ureteral anastomosis to a bowel conduit. (A) The traditional Bricker technique with two anastomoses on the conduit, one for each ureter. (B) Wallace I technique with first anastomosis of the ureters to each other with subsequent anastomosis to the bottom (peritoneal end) of the conduit limb.

    Continent Cutaneous Diversion

    Continent cutaneous diversion, first described as Miami pouch by Bejany in 1988,15 is a pouch made by a 20–25 cm detubularized segment of the right colon, which works as a low-pressure reservoir, and 10–15 cm segment of the ileum, which will be exteriorized to the skin as a stoma (frequently placed for cosmetic reasons at the umbilicus). Various variants of this technique have been described over the years, using different segments and lengths of bowel, such as the Indiana16 or the Mainz pouches.17 In general, the presence of the ileocecal valve and the difference in pressure between the colonic reservoir and the narrow and contractile ileum allow excellent continence rates (72–98%)9 10 not requiring any urostomy bag. The stoma can be constructed either with the ileum or the appendix. Although the use of the appendix has a lower incidence of incontinence than the ileum, it has higher risk of stenosis (up to 30%).10

    A most common long-term disadvantage of this particular urinary reconstruction is that patients will need to learn how perform intermittent self-catheterization and irrigation, which sometimes can be difficult. In the first several post-reconstruction weeks, this is particularly burdensome, as irrigation and catheterization may be needed frequently (usually every 3–6 hours). The most common complications of continent conduit include ureteral stricture (up to 27%), lithiasis (5–24%), urinary infection (15–50%), metabolic acidosis and vitamin B12 deficiency, pouch torsion or pouchitis (20–23%).9 10 18 Pouchitis is a rare but severe complication that may be prevented with regular pouch irrigations. Asymptomatic bacteriuria may occur in more than half of these patients, but these cases do not usually require antibiotics.19

    Orthotopic Ileal Neobladder

    Orthotopic ileal neobladder, first described by Lilien et al,20 is a urinary reconstruction that can be considered only when the urethra and the bladder neck can be spared. This reconstruction most closely resembles the original bladder with no stoma required. The basic principle of this technique is to create a low-pressure reservoir, using about 40–50 cm of ileum, that will be anastomosed to the bladder neck allowing good continence rates (77–90%).9 10 18 The ureters are separately anastomosed to the neobladder. In the Le Duc technique, the terminal ureter is tunneled in the ileal mucosa to prevent urine reflux. However, the direct ileo-ureteral anastomosis is easier, faster, with lower risk of stenosis than the Le Duc technique.20 21 Various variants of this technique have been described over the years, using different segments and lengths of bowel, such as the Budapest pouch22 that is created by using the right colon as reservoir (anastomosed to the bladder neck), and the terminal ileum, which is employed for the uretero-enteric anastomoses. Complications of neobladders include urinary retention requiring self-intermittent catheterization (5–35%), lithiasis (8–25%), urinary infection (5–50%), metabolic acidosis, and vitamin B12 deficiency (if terminal ileum is used).9 10 19 23

    Unfortunately, while there is a large experience on orthotopic neobladder within the field of urology (it is the second most frequently performed reconstruction),9 10 18 there is limited experience in the field of gynecologic oncology due to the oncologic resection required in the majority of cases with cystectomy. Except for extremely rare circumstances, the entirety of the bladder neck and proximal urethra are excised in supralevator pelvic exenteration while the entire urethra is removed in any exenteration requiring cystectomy and perineal resection.24 Therefore, the utility of orthotopic neobladder in the setting of gynecologic oncology surgical reconstruction is rare, but when the opportunity does arise, it behoves the team to involve reconstructive urology.

    Most publications and meta-analyses evaluating surgical outcomes of different urinary diversions, merged gynecological, colorectal, or urological cancers in both male and female patients, with obvious consequent biases. Regarding gynecologic cancers, most papers are from single institutions and retrospective with no randomized trials. For these reasons data are difficult to interpret and a critical analysis of comparisons between different techniques is extremely hard to make. A Cochrane collaboration study that included five randomized trials failed to find the ‘ideal’ diversion.25 No statistical differences in the frequency of surgical complications for the various methods of diversion were reported (with a slight favorable trend for incontinent conduits), including anastomotic urinary leaks. For this reason, many variables, such as age, co-morbidities, oncologic history, previous treatments (especially radiotherapy), patient’s functional preference, availability of healthy bowel, surgeon experience, and intra-operative complexity, all play an important role and need to be taken into consideration when choosing the type of diversion. Even from the quality-of-life point of view, no significant differences were found between continent and incontinent diversions,26–28 even if we recommend continent diversions in young and fit patients.

    We may therefore conclude that after pelvic exenteration in gynecologic oncology, the literature does not support any single urinary reconstruction option, but emphasizes the need for personalization. Incontinent diversion may represent a good option for patients with multiple co-morbidities, those who may benefit from shorter operative times, or for those unable to perform intermittent self-catheterization or decline the constraints entailed by continent diversions. Furthermore, having no urine dwell time in the pouch, patients have very low risk of developing metabolic acidosis due to urine absorption. Instead, for healthy patients who do not accept or tolerate an external stoma, due to potential urine leaks that may limit social activities or severely compromised body image, a continent reconstruction may be the preferred technique. Regardless, we advise that all patients understand that intra-operative evaluation of the overall case, anatomy, and health of the bowel may mean conversion to an incontinent rather than continent conduit.

    As experience with urologic reconstruction decreases within gynecologic oncology, and the advantage of collaboration with urology in some or all cases, we believe that patients requiring urinary reconstruction will benefit from high-volume referral centers3 9 26 29 30 in order to ensure an approach by a multidisciplinary team focusing on pre-operative evaluation, procedure selection, patient education, and surgery, including specialized reconstructive approaches that are required for optimal results.

    Reconstruction of Pelvic Floor Defects

    The extensive pelvic-perineal defects resulting from ultra-radical resections, which involve both the pelvic floor and skin, entail significant functional and esthetic challenges for both patients and surgeons. When a primary closure is not able to collapse dead spaces now void of previously space filling organs, reconstructive techniques with flaps are needed. Flap reconstruction is defined as the transposition of a healthy tissue from a donor site to a recipient site. This allows a tension-free skin closure, improves the wound blood supply, and fills the pelvic dead space. Various techniques for creating flaps are available and can be classified based on the type of tissue composition (skin, muscle, musculocutaneous, fasciocutaneous), on the harvesting site (local, regional, or distant), or on the blood supply (random or axial).

    Local and Regional Flaps

    A local flap involves tissues adjacent to the perineal defect that can be advanced, rotated, or transposed while maintaining their own blood supply (ie, V-Y flap). Local flaps are quick and relatively simple to perform, but are limited in size, and their blood supply relies on a random pattern. Regional flaps offer more volume and robust vascularity compared with local flaps. However, they are relatively small and may need to be performed bilaterally for larger defects. Two key regional flaps, the Singapore and Lotus petal flaps, receive their blood supply from perforating vessels originating from the internal pudendal artery.

    Distant Flaps

    Distant pedicled flaps are transferred from separate anatomical regions to the perineum while preserving their vascular connections to those regions. Fasciocutaneous flaps (composed by skin, subcutaneous tissue, and the underlying fascia) are employed for covering extensive superficial defects, while larger and deeper defects are addressed with myocutaneous flaps where the inclusion of the muscular tissue contributes to fill the pelvic dead space. Larger distant fasciocutaneous flaps (ie, the anterolateral thigh and the deep inferior epigastric perforator) can also be used but will not provide filling for empty spaces. These provide a large amount of skin coverage. Larger defects with empty pelvic spaces are better served with myocutaneous flaps (ie, vertical rectus abdominus myocutaneous, vastus lateralis anterolateral thigh, and gracilis).31 These are described further below. Research has shown that myocutaneous flap closure significantly reduces overall and major complication rates in comparison with primary closure.32 We agree and recommend extreme caution when attempting primary closure without flaps, especially in the infralevator exenteration. However, in the setting where skin is mostly spared and radical surgery was confined to the supralevator compartment, an omental flap may be sufficient for filling of empty pelvic space with primary closure of the skin defects. An omental flap is a distant pedicle flap often used to fill dead space in the pelvis. The omentum has its own blood supply from the gastroepiploic vessels. It is dissected from the transverse colon up to the greater curvature of the stomach, then both right and left vessels can be used as pedicle to obtain the required length and reach the pelvis. Limitations of an omental flap include (1) lack tissue in underweight patients and (2) necrosis of the omentum in cases of reduced blood supply.33

    Cutaneous and Fasciocutaneous Flap

    Perforator-based flaps have emerged as a trend in perineal reconstruction. Compared with traditional myocutaneous flaps, perforator flaps offer the advantage of thinner skin flaps (with or without fascia), resulting in more favorable outcomes and reduced donor site complications. The medial or inner thigh, rich in perforators, provides greater flexibility in designing flaps tailored to specific defects.34

    The anterolateral thigh flap is derived from the lateral thigh and relies on perforators originating from the transverse branch of the lateral circumflex femoral artery. The donor site defect can generally be directly closed, but grafting closure can be necessary in larger transpositions. While robust, harvesting the anterolateral thigh flap presents technical challenges. If necessary, it can be harvested as a myocutaneous flap alongside the vastus lateralis muscle (vastus lateralis anterolateral thigh). The anterolateral thigh is large and so can occasionally add too much bulk to the perineum, which has long term consequences for a patient’s ability to sit and walk.

    Among perforator flaps, the pedicled deep inferior epigastric perforator relies on the periumbilical rectus abdominis perforators and provides a vertical flap composed of skin and subcutaneous tissue while preserving the muscle and its aponeurosis.35

    The superficial circumflex iliac perforator flap is very versatile for perineal reconstruction, characterized by a long pedicle and by the possibility to obtain large flaps, which can eventually be harvested along with lymphatic tissue, to prevent lower limb lymphedema after inguinal lymphadenectomy.31

    Myocutaneous Flap

    The rectus abdominis flap provides a vertical (vertical rectus abdominus myocutaneous) or oblique or transverse skin paddle and a robust muscular support to fill the pelvic dead space and reconstruct large perineal defects. It is the primary abdominal flap routinely employed in any radical surgery requiring perineal procedures. This flap comprises the rectus abdominis muscle containing the deep inferior epigastric vessels, which emit perforating branches through the muscle to supply the skin of the abdomen.

    The gracilis flap consists of a section of skin and fat located at the junction of the perineum and thigh along with the underlying gracilis muscle which provides bulk for reconstruction. Owing to the unpredictable viability of the skin paddle, some surgeons prefer to use a muscle-only gracilis flap and achieve tension-free skin closure through a local flap if necessary.36 For moderate to large defects, a unilateral gracilis is rarely sufficient and bilateral gracilis can be used alternatively. Additionally, a lotus flap with a gracilis flap can also be employed to add both muscle and skin coverage.

    A recent meta-analysis has shown that both vertical rectus abdominus myocutaneous and gracilis flaps, the two most frequently employed muscular flaps, are safe options for perineal reconstruction. The former offers the advantage of robustness and a reliable skin paddle. Nevertheless, it is associated with a higher rate of donor site morbidity.37 While a vertical rectus abdominus myocutaneous flap may be preferable for the filling of perineal defects created with infralevator exenteration, there are challenges with stoma placement when both a urostomy and colostomy are needed. In a patient with a long torso, stacked stomas that go through the fascia and rectus muscle on one side with donor vertical rectus abdominus myocutaneous flap from the other side may be possible. However, when a patient has a shorter torso and is unable to have stacked stomas, the vertical rectus abdominus myocutaneous flap presents a challenge for proper stoma placement (figure 2). An alternative to stacked stomas or bilateral stomas is to have both urinary and gastrointestinal waste into a single stoma (wet colostomy) usually placed on the patients left and described below. In patients with prior Maylard incisions, which requires ligation of inferior epigastric vessels, the use of a vertical rectus abdominus myocutaneous technique, as well as the deep inferior epigastric perforator, is contraindicated. In this particular type of patient, a gracilis flap or other thigh donor site reconstruction may represent an excellent alternative.

    Figure 2

    Stoma placement options and challenges after total exenteration when urinary and gastrointestinal stoma are required. On the right, patients can have one stoma on each side of the abdomen. This allows for better long-term management of stomas and appliances for the patient but will not allow for a vertical rectus abdominus myocutaneous flap to be used for pelvic reconstruction without the cost of parastomal hernia risk. On the left, stomas are stacked with the left upper stoma being an end colostomy and the left lower stoma being a urostomy. This allows for contralateral vertical rectus abdominus myocutaneous flap harvest for management of perineal defect; however, the need for a long torso with adequate spacing will complicate this approach.

    In conclusion, abdominal-based flaps offer several advantages, including bulky tissues, ease of dissection, and the use of non-irradiated tissues. However, these advantages come with certain drawbacks, including an increased risk of donor site morbidities, such as abdominal hernias, donor site infections, and serious strategical problems in surgeries requiring one or two ostomies.

    Conversely, perineal-based or thigh-based flaps, such as a gracilis, can be harvested unilaterally or bilaterally and are suitable for smaller perineal defects. Moreover, they are associated with relatively low donor site morbidity, although with a less reliable skin paddle. An anterolateral thigh offers a large skin paddle and can also use muscle, but is associated with some donor site challenges including, sometimes, the need for skin grafting of the donor site. In general, we have an approach of assessing and being prepared for all options at the time of surgery, recognizing that specific intra-operative findings may require a change in course for pelvic reconstruction.

    Pelvic Space Filling

    In the case of pelvic exenteration creating a large defect in the pelvic floor, various surgical strategies have been employed to reduce complications due to the empty pelvis syndrome, such as migration of the small bowel in the pelvis, bowel obstructions, fistula, pelvic abscess, and wound dehiscence.

    As described above, a myocutaneous flap (vertical rectus abdominus myocutaneous and vastus lateralis anterolateral thigh) are the best choice. A vertical rectus abdominus myocutaneous flap can fill the empty space, close the pelvic floor, and provide a substantial skin paddle. However, double stomas can complicate the use of abdominal muscle. Moreover in some cases, when a pelvic exenteration is performed entirely using minimally invasive surgery, the abdominal wall incision can be avoided and a vastus lateralis anterolateral thigh flap may be considered the preferred option. The use of other pelvic fillers, such as obstetric balloon or silicone expanders, remains controversial, showing efficacy in case reports and small cohorts only.38

    The choice of a reconstructive strategy is a complex decision that must consider various factors.While previous algorithms primarily evaluate the defect size only, more recent approaches suggest considering the three-dimensional nature of the defect, the anatomical subunits involved in vulvoperineal resection, co-morbidities, prior surgeries/irradiation that may have altered the anatomy, tissue laxity, vascularization, and the surgical approach (eg, minimally invasive or open exenteration).31 39 A logical algorithmic approach, coupled with implementation of post-operative care protocol, the integration of experimental techniques such as 3D bioprinting, will collectively contribute to improve reconstructive outcomes while minimizing the occurrence of complications. Specifically, 3D model printing prior to surgery can improve the evaluation of space, anatomic configuration, and help with surgical planning. Such 3D models can be useful for all parts of the surgical plan, and are available at some centers. An example of a model is provided in online supplemental figures 1-2.

    Supplemental material

    Supplemental material

    Vaginal Reconstruction and Neovaginas

    Primary vaginal and pelvic floor reconstruction after exenteration is currently considered a fundamental part of the reconstructive phase and its impact on patients’ quality of life, sexual function, and body image has been emphasized in several publications.40–48 However, the reconstruction of the vaginal defect is a challenge, as it must succeed in restoring both the form and the function of the original structure. Over the last century, many techniques have been described for neovagina construction, such as the Singapore (or pudendal) technique, the gracilis muscle flap reconstruction, the deep inferior epigastric perforator flap, the gluteal thigh flap, the rectus abdominis myocutaneous procedure (either vertical or transverse), or the intestinal vaginal reconstruction. Given that data are very scant, heterogeneous, and mostly retrospective, it is difficult to make comparisons between the different techniques and to identify a hypothetical ideal reconstruction.

    Overall, it appears there is a preference towards rectus abdominus myocutaneous flaps with regards to long-term functional outcomes. Soper et al compared flap-specific complications of gracilis myocutaneous and rectus abdominis myocutaneous neovaginal reconstructions after radical pelvic surgery, reporting lower overall incidence of flap-specific complications and significantly lower incidence of flap loss in the transverse rectus abdominis myocutaneous group47 (figures 3 and 4). Jurado et al in 2009, presented a series of 29 patients, over a 10-year period, comparing the transverse rectus abdominis myocutaneous flap and the Singapore fasciocutaneous flap. Vaginal stenosis, necrosis, and shortness occurred less frequently for the transverse rectus abdominis myocutaneous flap than for the Singapore flap (19.0% vs 28.6%, 14.5% vs 50%, and 0% versus 100%, respectively), concluding that the Singapore flap carries a higher rate of complications along with worse sexual function.49 The decision to perform either a transverse (transverse rectus abdominis myocutaneous) or a vertical (vertical rectus abdominus myocutaneous) approach is purely based on the surgeon’s personal preference and there are no prospective randomized data available; to date no significant differences have been reported between procedures, although some authors support their preference for a vertical rectus abdominus myocutaneous approach based on the ease of abdominal wall closure due to a smaller fascia defect. It is important to note that the above literature cited were not confined to reconstruction in gynecologic malignancy, and therefore may not be applicable to a gynecologic oncology setting. Also, as stated in the previous section, the choice of vaginal reconstruction procedure must be contextualized to the overall procedure, including issues with abdominal wall management, such as hernia treatment prevention and ostomy placement.

    Figure 3

    Technique for creation of neovagina made from bilateral gracilis myocutaneous flaps.

    Figure 4

    Technique for creation of neovagina made from vertical rectus abdominus muscle flap.

    Stenosis occurs in up to 85% of neovaginas in some series,49 50 preventing normal vaginal coitus with subsequent low sexual function, and therefore must be considered. An alternative to the rectus abdominis myocutaneous procedure with lower rates of stenosis is an intestinal vaginal reconstruction. Intestinal neovagina reconstruction uses isolated bowel segments, most commonly sigmoid colon, cecum, or ileum, to create a neovagina, and is associated with overall low rates of complication. This type of procedure is mostly performed for non-oncologic reasons, such as congenital disorders or gender affirmation, and the experience in the gynecologic oncology field is extremely limited.51 The potential advantages of using the bowel for a neovagina include adequate caliber and length, lower risk of stenosis, with evident psychosexual benefits and reported sexual activity of 86–100% over the first year after surgery. An additional benefit is that the neovagina fills the ‘empty’ pelvis and that the cecum, in patients who previously underwent pelvic radiotherapy, is less likely to have radiation-related issues. Moreover, even after resection of the rectosigmoid colon at the time of an exenteration, using the cecum and portion of the ascending colon to reconstruct the vagina leaves a reasonable segment of intact colon avoiding potential functional issues. On the other hand, if the patient desires a continent urostomy as a urinary reconstruction, it may be better to consider a different method of reconstruction. One frequently reported disadvantage of using the bowel for neovagina include discomfort due to heavy mucous discharge from the neovagina. However, the amount of discharge usually decreases over time. Again, data are limited in gynecologic oncology.

    Patient selection is critical in neovagina construction. Vaginal reconstruction requires longer operative times (average additional 2 hours) compared with pelvic exenteration alone. There is additional complication risk related to the vaginal reconstruction (overall increase of about 20%), such as pelvic abscess (10%), fistulas (up to 20%), delayed wound healing (10–46%), flap necrosis49 50 or skin graft ulceration, all of which can be worse in a previously irradiated field. Moreover, patients need to be appropriately counseled about sexual function and values and preferences. Data concerning sexual function after vaginal reconstruction are scarce and interpretation of results controversial. The most common problem noted by patients in adjusting to sexual activity can be their self-consciousness about urostomy or colostomy, vaginal dryness (33%) or discharge (28%), and dyspareunia (18%) due to vaginal stenosis or inadequate sensation of the neovagina. According to a literature review, that merged data from 199 patients, there is a great variability (14–84.6%) in sexual activity after vaginal reconstruction.52 For these reasons it is of utmost importance to select and identify those who will benefit from this procedure, and those for whom the neovagina will only be a burden with increased peri-operative morbidity. Similar to the urinary and pelvic space reconstructions, we approach vaginal reconstruction in the context of the overall procedure, knowing that intra-operative findings may alter the outcome. For this reason, extensive and honest counseling with patients about potential outcomes and value-concordant priorities is essential pre-operatively.

    In conclusion we may conclude that vaginal reconstruction is feasible and can be proposed to selected patients, after extensive counseling. Rectus abdominis myocutaneous flap seems to be the preferable option for reconstructing pelvic floor and vagina, without any significant additional morbidity to pelvic exenteration. Intestinal vaginal reconstruction may be of great potential, especially from the sexual function point of view.

    Gastrointestinal Tract Reconstruction

    Although the vast majority of patients undergoing radical pelvic resection will undergo a standard terminal end colostomy, several alternatives are worth noting. As is the case for almost all data heretofore presented, the data in gynecologic oncology are limited, and are mostly based on single-institution series.

    Wet Colostomy

    In the setting of a total pelvic exenteration with removal of both the urinary and gastrointestinal tract, one possibility is a wet colostomy. In this procedure, the ureters are anastomosed proximal to the terminal gastrointestinal colostomy, usually the sigmoid (ureterosigmoidostomy). Stool and urine will be relieved via a single stoma and single device. Complications ensuing from reflux of gastrointestinal contents to the ureters include frequent urinary tract infections, specifically pyelonephritis, metabolic abnormalities (hyperchloremic and hypocalcemic acidosis), kidney disease, and large volumes of watery malodorous stool which can be difficult to manage.

    An alternative to a wet colostomy with a single stoma and device is a double-barrel wet colostomy. In this a loop colostomy is performed with the distal limb of the colon acting as a urinary reservoir. The ureters are anastomosed to the distal limb that is exposed less to stool, and the loop colostomy is mature in a single defect in the left anterior abdominal wall.53–55 An advantage of this technique besides the single stoma is the fewer number of abdominal anastomoses. A small but important series from the Ohio State showed that overall there is no difference in complications in patients with a double-barrel wet colostomy versus separate diversions; however, conclusions are limited given the inevitable variability in each case with regards to risk factors and influence on reconstruction choice.56 There has been concern about malignancy in the colon in the setting of urine exposure to the functioning colon, but the rate is unclear. This risk, however, should be contextualized to the overall benefits and the need for revision in the future if malignancy occurs.57 Other risks include necrosis of the distal colon urinary reservoir, stoma retraction, and fistula.55

    In anal and low rectal cancer, the technique of a perineal colostomy58 and coloanal anastomosis has been well-described. Generally speaking these techniques allow patients to not have a fecal diversion, but continence rates are somewhat limited. While seemingly safe in colorectal and coloanal cancers, there are no data for oncologic safety in gynecologic cancers. Additionally, the presence of lower vaginal disease or vulvar disease would not allow for coloanal anastomosis. In a perineal colostomy, the end of the colon is sewn to the open perineal wound. Several techniques have been described to achieve continence with variable outcomes.58 59


    Pelvic exenterative surgery in gynecology oncology brings the obvious challenge of reconstructing the gastrointestinal, urologic, and gynecologic tracts. With the improvement of systemic treatments, thankfully the need for these ultra-radical surgeries has become less frequent over time. However, when required, the benefit of a high-volume center with dedicated services seems clear. Careful patient selection, meticulous operative techniques, efforts to minimize operative times and blood loss, multidisciplinary team effort, intensive post-operative management with the availability of significant physiological support remain the cornerstones to reduce the complication rate and maximize the success of such radical surgery. Expansion of enhanced recovery programs after surgery and minimally invasive techniques being applied to pelvic exenteration can potentially allow further improved peri-operative outcomes. There is a continued need for high-quality, person-centered research around quality of life for patients and their caregivers, particularly as it comes to benefits and preferences for reconstructive options. We believe there remains work to be done and larger multicentric prospective studies are warranted to identify optimal reconstructive options and to identify which patients may benefit the most from each approach.

    Ethics statements

    Patient consent for publication

    Ethics approval

    Not applicable.



    • ML and AA are joint first authors.

    • Twitter @matteoloverro

    • Correction notice This article has been corrected since it was first published. The second affiliation has been updated.

    • Contributors All authors (ML, AA, LT, GDA, AK) served as writers and provided input to manuscript aims and design. AK (contributing author) compiled all edits and documents. GDA, AA, and ML all contributed as reviewers of the final draft.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests None declared.

    • Provenance and peer review Commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.