Multivisceral Resection of Advanced Pelvic Tumors: From Planning to Implementation

Abstract

Pelvic exenteration involves radical multivisceral resection for locally advanced and recurrent pelvic tumors. Advances in tumor staging, oncological therapies, preoperative patient optimization, surgical techniques, and critical care medicine have permitted the safe expansion of pelvic exenterative surgery at specialist units. It is now understood that in carefully selected patients, 5-year survival can exceed 60% following pelvic exenteration, and that very low mortality figures and an optimum postexenteration quality of life are possible. In the present review, we provide a contemporary summary of the current state of the art in pelvic exenterative surgery following all key phases of the treatment pipeline from patient staging and tumor assessment, to treatment planning and surgery.

Multivisceral Resection of Pelvic Tumors

The technique of pelvic exenteration was first described by Brunschwig in 1948 ¹ as a palliative procedure for advanced cervical cancer, which at the time conferred a dismal prognosis, with treatment primarily limited to radium irradiation and deep X-ray therapy. Brunschwig was firmly of the opinion that this radical treatment was purely palliative, even though he was careful, as far as was possible at the time, to select only patients with disease confined to the pelvis. His first published series describing outcomes in 22 patients undergoing the procedure reported significant procedure-related blood loss with a combined intraoperative and early postoperative mortality of 23%. ¹ Early scepticism and reservations regarding morbidity were offset by the fact that these procedures were being offered exclusively to those with no hope for extending survival otherwise.

The decades that followed Brunschwig's early descriptions have witnessed a coalescence of multiple paradigm shifts and incremental innovations in associated medical specialties that have permitted expansion, widespread adoption, and implementation of pelvic exenteration. First, the fundamental notion that “big surgery” equates to unsafe surgery has been countered by advances in surgical hemostasis, anesthesia, critical care, antisepsis, and blood transfusion. Simultaneously, growing evidence has amassed that firmly supports the concept of exenteration as a curative, rather than palliative technique in advanced pelvic malignancy—“big” does not necessarily mean “bad” in terms of disease biology and disease course, and patients with cancers confined to the pelvis can be offered treatment with curative intent.

In the present day, pelvic exenteration can be defined as the radical multivisceral surgical resection of locally advanced or recurrent malignant tumors of the pelvis. The primary aim of this approach in modern day practice is to achieve complete oncological tumor clearance (R0 resection). This frequently necessitates en bloc extirpation of a “composite” surgical resection specimen, consisting of tumor mass and all contiguously involved anatomical structures, including adjacent viscera, as well as, to varying extents, pelvic vasculature, nerves, soft tissue, and bone. It is acknowledged that complete tumor clearance is the most significant factor influencing survival and recurrence after exenterative surgery, with recent data indicating that 5-year survival of more than 60% is possible in carefully selected patients undergoing R0 resection. ^{2 3}

In the present review, we provide a comprehensive outline of the process of multivisceral resection for pelvic tumors, covering all components of the clinical pathway from the initial planning stage to the conduct of surgery.

Indications for Pelvic Multivisceral Resection and Types of Exenterative Procedure

Locally advanced, persistent or recurrent tumors of the pelvis without evidence of distant metastatic spread are not only life-limiting but can also lead to significantly impaired quality of life by causing locoregional disturbances, such as constant opiate refractory pelvic and lower limb pain, chronic sepsis and fistulae, gait alteration, and problems with voiding of urine and defecation. Pelvic exenteration is recommended as the most assured means of achieving R0 resection (no cancer cells seen at the resection margin or cancer cells present but >1 mm from resection margin) and long-term survival under these circumstances. The advanced nature of these tumors means that they adhere less consistently to visceral boundaries and anatomical planes and frequently follow a more nebulous pattern of local extension and invasion. Hence, to be sure of removing all tumors in this context, it is recommended that radical surgery involve removing all tumor contiguous structures, en bloc.

The indications for such multivisceral resection are broad, such that the primary or recurrent tumor of origin may be of urological, gynecological, colorectal, or soft tissue/skeletal origin. ⁴ Correspondingly, the exenterative procedure needs to be tailored specifically for a given patient and their radiologically predicted disease distribution, typically requiring simultaneous removal of a compilation of pelvic structures including the prostate and seminal vesicles (in the male), the uterus, fallopian tubes, ovaries, and vagina (in the female), as well as the bladder, rectum, pelvic veins and arteries, nerves, musculofascial structures, and bone, to varying extents. Definitions of pelvic exenteration are not always uniform and are heterogeneously applied in the literature. Broadly, four types of pelvic exenterative procedure have been described in the literature: anterior, posterior, total, and extended. In all cases, the procedure may be limited to the supralevator pelvis with preservation of the perineum or may have an infralevator component and require a perineal phase to the operation also.

Anterior pelvic exenteration involves resection of the central pelvic organs and anatomical structures in the anterior portion of the pelvis, including reproductive viscera, bladder, urethra, and distal ureters, while preserving the rectum. Posterior pelvic exenteration is performed to treat rectal cancers that invade the uterus, cervix, or vagina, or gynecological tumors that involve the rectum or anus. This procedure spares the anterior vagina, bladder, ureters, and urethra. Total pelvic exenteration involves en bloc resection of all the internal pelvic viscera including the bladder, urethra, lower ureters, and rectum (with or without the anus). This is the most commonly performed type of multivisceral resection in pelvic malignancy, as it offers the most reliable chance of achieving R0 resection. Extended exenteration is indicated in circumstances where tumor is noted to involve/abut structures beyond the “conventional” plane of pelvic multivisceral resection, for example, where the pelvic sidewall or anterior/posterior bony tissues are involved. Here, an extended exenterative approach involving en bloc side wall or bony resection should be considered if the prospects for achieving R0 resection seem favorable. A key limitation of existing classification and definitions of exenterative approaches is that they fail to adequately catalog the magnitude of surgery required at times. As an example, a laterally extended exenteration approach makes no distinction on whether the operative intervention to achieve an R0 requires resection and reconstruction of the ureter alone, or resection of the internal iliac vessels, or resection and/or reconstruction of the external iliac vessels, or lateral bony resection of the ilium or ischium.

Staging in Advanced Pelvic Disease

Treatment decisions in advanced pelvic malignancy are strongly influenced by radiological interpretation of tumor extent and distribution. Preoperative imaging in this context aims to determine locoregional resectability and the anatomical tissue needing to be excised to enable the greatest chance of an R0 resection, and to assess for the presence of distant metastatic disease.

It is widely accepted that all patients should undergo contrast-enhanced multidetector staging computed tomography (CT) scans of the chest, abdomen, and pelvis to assess for evidence of extrapelvic disease. Our own practice involves image acquisition after administration of intravenous contrast, with images in arterial and portal venous phases at slice thickness of 1 mm and reconstruction of 0.8 mm.

In a recent consensus statement, the Beyond TME Collaborative, a group comprised invited experts from surgery, radiology, oncology, and pathology, recommended that the optimal modality for imaging the pelvis is a high-resolution thin-section magnetic resonance imaging (MRI) scan, and that this should be performed within 6 weeks of planned surgical resection. ⁵ MRI has been shown to provide a clear appreciation of the extent of tumor involvement in the four main compartments of the pelvis (anterior, axial/central, posterior, and lateral; Fig. 1A ), and of the predicted surgical planes of dissection, with better delineation of soft tissue and bony structures. Recent analysis by Dresen et al found that MRI was capable of highly accurate preoperative evaluation of recurrent rectal cancers with a negative predictive value of 93 to 100% for invasion into critical structures. ⁶ However, it is also important to note that the authors observed a propensity for MRI to overstage recurrent tumors, which was attributed mainly to the misinterpretation of regions of diffuse fibrosis postchemoradiotherapy, especially at the lateral pelvic sidewall. ⁶ Similarly, Georgiou et al investigated the diagnostic accuracy of MRI in detecting pelvic tumor invasion according to intrapelvic compartments for planning exenterative surgery. ⁷ Here, the authors defined seven intrapelvic compartments—peritoneal reflection (PR), anterior (above PR), anterior (below PR), central, posterior, lateral, and inferior. A 3- to 7-mm section T2-weighted technique was employed using a surface pelvic phased array coil and a small field of view 1.5 T MRI system. Reported sensitivity of MRI was >93% in all but the lateral compartments, where it was 89%. The authors conclude that radiologically subdividing the pelvis into seven compartments is of value in treatment planning and provides the surgeon with clearer information regarding the extent and position of disease while also enabling the radiologist to improve image interpretation and to standardize reporting. ⁷ Chew et al undertook a Delphi-based study to evaluate clinical and radiological criteria used for surgeons to determine suitability for pelvic exenteration surgery in recurrent rectal cancer. ⁸ The overwhelming consensus from the expert panel, slightly at odds with the findings of Georgiou et al, was that tumor involvement of the central and anterior compartments (with the exception of bony involvement) was associated with a more favorable outcome and higher likelihood of R0 resection. As with several previous studies, the authors suggest that lateral pelvic sidewall involvement, as noted on preoperative MRI, in particular iliac vessel and ureteric involvement, are poor prognostic indicators, associated with a reduced likelihood of achieving R0 resection. The authors also point out the potential limitation of MRI in terms of distinguishing fibrosis from active disease. ⁸ In this regard, the use of whole-body positron emission tomography (PET)-CT is currently being investigated by several ongoing clinical trials.

Fig. 1.

Imaging and staging . ( A ) Sagittal high-resolution thin-section pelvic MRI showing advanced rectosigmoid cancer with involvement of the bladder trigone and extension close to the surface of S1; ( B ) cross-sectional MRI pelvis showing advanced rectosigmoid cancer with sepsis and indistinct tumor boundaries; and ( C ) PET-CT images of same patient in ( B ), showing diffuse avidity, as well as a pelvic sidewall avid nodal deposit. CT, computed tomography; MRI, magnetic resonance imaging; PET, positron emission tomography.

A further key area of limitation in radiological staging and surgical planning in the authors experience is the presence of sepsis and the radiological changes that can accompany this. We and others have noted that anastomotic failure after primary rectal cancer surgery (and the sepsis that can accompany it) is associated with recurrence of rectal cancer ^{9 10 11} ; however, this finding also makes distinguishing the precise margins of the recurrent cancer—in a field of sepsis-associated radiological change—much more challenging ( Fig. 1B ).

The advantage with PET-CT may lie in the fact that it provides useful information with regard to the presence of metastatic disease, while simultaneously providing hybrid anatomical and metabolic imaging data with which to assess local tumor extent and to differentiate scar/fibrosis from viable tumor ( Fig. 1C ). These potential advantages notwithstanding, it is acknowledged that false-positive interpretations of FDG-avidity on PET-CT caused by displaced pelvic viscera, for example, bladder, seminal vesicles, uterus, and small bowel loops, as well as sepsis or radiation-associated inflammation can result in reduced specificity. For these reasons, and given uncertainty regarding precise utility, the Beyond TME Collaborative recommendations do not, at the present time, advocate routine use of PET-CT imaging in the work up of patients for pelvic exenteration. ⁵

In addition to imaging, a full endoscopic evaluation of the colorectum is also needed to confirm or refute the presence of any synchronous lesions which may subsequently influence polypectomy or extent and/or type of surgical intervention.

The Multidisciplinary Team and Initial Patient Assessment

The authors firmly believe that patients with complex locally advanced and recurrent pelvic malignancy should be referred to specialist units that possess the necessary surgical, oncological, radiological, pathological, and critical care expertise to manage them through all steps of the clinical pathway, ideally from diagnosis and initial assessment, to surgery and after care where deemed appropriate. The complex cancer multidisciplinary team (MDT) should ideally exist as a distinct entity in its own right, rather than be undertaken in conjunction with the standard cancer MDT. Irrespective, there are two essential requirements: (1) to have members of the MDT in attendance with expertise in the aforementioned areas, ideally with internally audited data demonstrating sound decision making as substantiated by quality of histopathological resection specimens and oncological outcomes; (2) to have all relevant patient-specific information available for discussion at the MDT. In regard to the second point, it is important to appreciate that patients are frequently referred from other institutions, may have had multiple previous treatments and procedures, and the comprehensiveness of accompanying information can be highly variable. A pro forma-based approach is essential to ensure that all aspects of care are covered, in transparent fashion, beginning with a summary of patient demographics and relevant comorbidities, including performance status. A complete outline of previous treatment, where applicable, ideally with operative notes, should be provided for the MDT, as this has relevance for the surgeon, in terms of operative planning, for the oncologist, to aid in determining suitability for further neoadjuvant chemo- and/or radiotherapy or reirradiation, and in particular, with respect to previous to determine the potential for any further boosts, and for the radiologist, in terms of interpreting imaging.

Patients who are of good general health, with acceptable performance status are considered for radical multivisceral resection where there is no evidence of metastatic disease and where preoperative imaging suggests R0 resection is feasible. Increasingly, this approach may also be suitable for patients with resectable oligometastatic disease also.

MDT discussions need to outline and document the radiological findings as precisely as possible, to define the role of oncological therapy (if required), and to define the plan for surgical resection to enable an R0. The surgical strategy will require detailed planning in terms of the type of exenteration, for example, whether there will be both fecal and urinary diversion, whether flap reconstruction of the perineum is anticipated, and which other teams may need to be recruited to assist in the achievement of an R0 resection such as vascular or neuro/spinal surgeons. Importantly, this package of information will then need to be relayed back to the patient, before final decisions are made regarding how to proceed. The importance of this discussion and correctly calibrating patient and family expectations of what is achievable, and at what functional cost, cannot be overstated, and is often regarded as the most important step. The possible options and plan, as set out by the MDT, should ideally be discussed with patients and relatives in a dedicated complex cancer clinical appointment. Our experience is that these encounters work best where consultation is provided in a multidisciplinary manner, with appropriate support services and cancer nurse specialists, often in combined clinics with oncologists and/or palliative care specialists.

The initial patient consultation will also involve a thorough abdominal, pelvic and perineal, and lower limb clinical examination, at which time further useful information is obtained which may influence treatment decisions. Digital rectal examination, where possible, may allow assessment of height, orientation and fixity of tumors to adjacent structures, and advise if a formal examination under anesthesia (EUA) is required. In our experience, imaging modalities are often limited in resolution in the low pelvis and perineum, and especially, where there is involvement of the vagina, an EUA may better define the precise extent of vaginal involvement, determine margins of surgery, and help distinguish the method of reconstruction from a simple vaginaplasty, to biological mesh replacement, or soft tissue flap reconstructions ( Fig. 2A–C ). Evaluation of scars and prior procedures on the abdomen or thighs may also indicate the need for CT angiography to outline the patency, caliber, and utility of feeding and perforator vessels, and to assist in optimal flap design during any reconstruction phase ( Fig. 3 ).

Fig. 2.

Different methods of vaginal reconstruction . ( A ) Medial thigh flap vaginal reconstruction; ( B ) bilateral gluteal artery flaps to reconstruct vagina; and ( C ) rectus abdominus muscle flap reconstruction of near-total vaginectomy at outpatient review.

Fig. 3.

CT angiogram in patient with locally recurrent rectal cancer and multiple previous operations showing patent good caliber left Inferior epigastric artery which could be used for the reconstruction of any perineal/sacral defect, but absent right inferior epigastric artery. CT, computed tomography.

Where exenteration is an option, patients need to be made aware of the potential functional and body image changes that may result from this intervention, possible complications of surgery, likelihood of R0 resection, and anticipated length of postoperative stay in hospital. It goes without saying that informed consenting of patients (and their family) for such major life-altering interventions must be approached differently to standard practice for more reserved surgical procedures. It is important also that patients have sufficient time to digest the information provided and to carefully consider their decision, and as a result, some patients may require future supplementary appointments.

Following this, additional referrals to the stoma team, orthotics services, and prehabilitation services are frequently required in those patients wishing to go ahead with surgical interventions. More broadly, the impact of advanced cancer requiring such radical surgery on the mental well-being of patients is slowly being explored in the medical literature. For example, the interaction of clinical and psychological variables in patients undergoing pelvic exenteration was recently evaluated by Arnaboldi et al, who assessed the feasibility of a psychological intervention package administered to 49 pelvic exenteration candidates. They noted that although suffering with treatment and condition-related anxiety, pelvic exenteration candidates demonstrate a range of adaptive coping mechanisms, which can be enhanced through psychological prehabilitation. ¹²

Neoadjuvant Therapy

Chemotherapy, and radiotherapy or reirradiation, may be used to improve outcomes further in patients being considered for pelvic exenteration. Their application is tumor specific and needs to be guided by the appropriate tumor board; however broadly, several goals have been described in the available literature.

First, neoadjuvant therapy may provide early treatment of occult micrometastatic disease, and thereby improve long-term cancer-specific outcomes. In the setting of the high postoperative morbidity associated with pelvic exenteration surgery, many patients may not be able to tolerate adjuvant chemotherapy following pelvic exenteration, and consequently, the prospect of total neoadjuvant therapy prior to radical surgery is an attractive one. Second, neoadjuvant therapy may also be used to locally downstage what is initially perceived to be a poorly resectable or unresectable tumor, and thereby increase operability. A key area of contention however is the planning of surgical margins following neoadjuvant therapy and this is discussed further under “Surgical Planning and Operative Approach to Multivisceral Resection” heading. Finally, “neoadjuvant” treatment may also be employed in the setting of metastatic disease with the intention of facilitating systemic control, and subsequent sequential or synchronous multisite metastasis and exenteration surgery, depending on response. While neoadjuvant therapy seems to have oncological advantages, it also has not insignificant shortcomings which need to be considered. These include multiorgan deconditioning, and the possibility of disease progression (locally or systemically) during the course of treatment. In addition, regression-related fibrosis and inflammation can make surgical dissection more hazardous. ¹³ In this setting, the recent findings from a multinational cohort of 1,291 pelvic exenteration patients revealed that while 78% underwent some form of neoadjuvant therapy (chemoradiotherapy 64%; radiotherapy alone 10%; chemotherapy alone 3%; treatment details unknown 23%), this did not translate to a statistically significant benefit in survival. ¹⁴ In addition, patients who received neoadjuvant therapy were 1.7 times more likely to experience a major complication in the first 30 days following exenteration. ¹⁴

It remains an area of contention as to whether subsequent surgical resection should be aimed at the more modified postneoadjuvant treatment margins or preneoadjuvant therapy margins. In light of the fact that cancer cells are heterogenous, and their response to treatment is similarly heterogenous, it seems likely that modifications of initially planned surgical margins to a more limited margin may risk leaving residual disease. In keeping with this, we and others have noted the presence of nests of viable cancer cells among a zone of regression, which might have been left in situ if exenterative margins had been modified after the initial neoadjuvant therapy, potentially contributing to early recurrence.

Perioperative Medicine and Prehabilitation

The preoperative anesthetic work-up for patients being planned for exenteration is complex, and all patients should be assessed by a consultant anesthetist experienced in this type of radical surgery, to assist with the decision to operate. Risk stratification allows better informing of patients and relatives about the likely risks of prolonged anesthetic and major surgery. Increasingly, this process is aided by cardiopulmonary exercise testing (CPET), and further supplementary tests such as cardiac echocardiography or lung function testing where deemed appropriate. Recent data indicate that one of the principle drivers behind the high treatment costs associated with exenterative surgery is the extended length of hospital stay that patients undergoing exenteration typically require, which is often aggravated by postoperative complications. ^{15 16} Preoperative exercise programs designed to improve patients' fitness before surgery have emerged as one relatively cost-effective way to potentially enhance the postoperative recovery of patients undergoing major surgery. A recent systematic review of preoperative exercise interventions for patients undergoing cancer surgery presented data from 17 studies investigating the effectiveness of this measure in six different cancer cohorts. ¹⁷ While not directly attributable to patients undergoing pelvic exenteration, the authors noted a 48% reduction in postoperative complications in some patients receiving preoperative exercise programs, with a corresponding reduction in average length of stay of 3 days in the intervention group was noted. ¹⁷ These encouraging figures have stimulated interest for the use of this approach in patients undergoing exenterative surgery.

In our experience, all patients being considered for pelvic exenterative surgery should undergo CPET testing and be provided with prehabilitation guidance to aid decision making. Given that the work-up for these patients can be several weeks and may include neoadjuvant treatment, there appears to be an opportunity over this period to introduce a program of preoperative conditioning without delaying treatment. This and other elements of prehabilitation are the subject of investigation in the currently active Pre-operative Physical Activity (PEPA) (ACTRN12617001129370) and Wessex Fit-4-Cancer Surgery (NCT03509428) trials.

Surgical Planning and Operative Approach to Multivisceral Resection

Planning for Exenterative Surgery

The decision to offer surgery and the planning of exenterative surgical procedures are complex, and require experienced multidisciplinary teams, working in concert with motivated and carefully counseled patients, for optimal results to be achieved.

As the anatomy of disease extent is heterogenous, so the surgery required to facilitate a multivisceral R0 resection is heterogenous. Consequently, the surgical team consists of surgeons with multidisciplinary skills, including colorectal surgery, together with one or more of the following specialties, depending on tumor specifics: urology, vascular surgery, orthopaedic surgery, neuro/spinal surgery, gynecology, and plastic surgery. An additional consideration will be whether adjuncts to surgery such as intraoperative radiotherapy (IORT) or intraperitoneal chemotherapy may be required also.

A key area of debate focuses on how to plan surgical margins in the era of neoadjuvant treatment, and whether any eventual surgical resection should be planned based on postneoadjuvant treatment imaging in patients with significant radiological regression or based on more extensive pretreatment margins. As previously highlighted, the concern here is that a more limited margin of resection might risk leaving residual disease, in particular, if preoperative treatment has resulted in tumor discontinuity. Fig. 4 illustrates this point, demonstrating clusters of viable tumor cells encased within a zone of chemoradiotherapy-induced fibrosis. Accordingly, our practice, and most recommendations, such as the Beyond TME Collaborative and the UK ACPGBI (Association of Coloproctology of Great Britain and Ireland) IMPACT initiative, suggest that surgery should be based on imaging when the tumor is at its maximum extent. ^{5 18} Careful discussions and planning with radiology colleagues are needed to formulate a clear operative plan and planes of dissection therefore to maximize achieving an R0 resection.

Fig. 4.

Nest of viable tumor cells (arrow) in zone of regression, with dense collagen deposition and fibrosis.

Broadly speaking, the operation may be said to have an abdominal phase, a perineal phase, and a reconstructive phase. Where a significant sacrectomy is needed, then a prone phase may also be required.

General Principles

For the abdominal and perineal phases, the patient is placed in Lloyd-Davies position. Depending on the type and method of any plastics reconstruction, and the need to harvest any vein for vascular reconstruction, the thighs may need to be bilaterally prepped and draped in a way to enable access during the reconstruction phase.

The procedure typically commences with a full-length midline laparotomy. If surgery is being conducted for recurrent disease, then a significant adhesiolysis may at times be needed. The peritoneal surface and liver should be carefully inspected for occult metastatic disease. In the absence of any suspicious abnormalities, the procedure should be continued as planned. Where subtle abnormality or pathological uncertainty is found, consideration should be given to frozen section analysis before proceeding further. Attempts at definitive surgery should be avoided in the case of the unexpected finding of widespread peritoneal surface disease.

The cecum and small bowel are mobilized and packed into the upper abdomen assisted by an abdominal self-retainer. In cases of recurrent pelvic malignancy, the small bowel may be adherent in the pelvis or to the area of tumor regrowth, which would necessitate en bloc resection with the specimen.

The next step usually involves identification of the ureters; these may be more medially situated in patients who have undergone previous pelvic surgery. We employ a selective policy of preoperative stenting. The course of both ureters should be defined with sharp dissection, and at the same time, an assessment is performed to determine if involved.

If the ovaries and/or gynecological organs are present and require resection, an initial step at this stage would also involve the isolation, ligation, and transection of the infundibulopelvic ligaments.

Once the decision to proceed is clear, the subsequent steps are tailored to the individual case, and will differ depending on the compartments that require resection. At all times, care must be taken to ensure the greatest chance of a negative resection margin.

Anterior Compartment

The bladder is initially dissected free from the retropubic space or the space of Retzius. The dissection is next extended laterally taking care to avoid injury to the external iliac vessels. The vas deferens or round ligament is ligated, followed by the identification and subsequent ligation and transection of the superior and inferior vesical vessels to fully mobilize and devascularize the bladder. Subsequently, the endopelvic fascia must be opened bilaterally, and the dorsal venous complex ligated to enable access to the urethra, which is cut at the level of the pelvic floor. Where additional anterior extensions are required due to concern regarding involvement of the symphysis or pubic rami, the plane of dissection may be extended anteriorly through a combination of abdominal and perineal phases, with the en bloc resection of any bony structures of concern.

Posterior Compartment

Where the sacrum or other posterior structures such as the presacral fascia, piriformis, or nerve roots deep to the mesorectum are not involved, then the plane of dissection posteriorly remains the TME plane and begins with the identification of the areolar tissue in the plane between the visceral mesorectal fascia and the parietal fascia. This plane may then be followed down to the pelvic floor. The rectum should be washed before transection and divided at a point sufficient to provide the necessary margin for cancer control.

Where the posterior structures are involved, then the plane of dissection needs to encompass the involved field. Varying degrees of posterior extension can be achieved, including dissection onto and including the sacral periosteum, or resection of the anterior sacral bony cortex, or partial sacrectomy itself. As a result of the vascular supply to this area, a lateral compartment dissection bilaterally may be required. The sacral level of transection for anterior osteotomies is determined by marking the area with screws and intraoperative fluoroscopy ( Fig. 5A ). Once the planned bony level of resection is reached, our practice involves placing a small gauze swab into the pelvic cavity, anterior to the sacrum, and posterior to the pelvic vessels and soft tissue structures, to prevent inadvertent injury when performing a sacral osteotomy. During this process, the surgeon must take care not to cause injury to the thecal sac. In the posterior approach, the screw marks the level of the osteotomy (as visualized with lateral fluoroscopy) to facilitate joining to the anterior unicortical osteotomy ( Fig. 5B ). If lateral (vertical) osteotomies are necessary through the ilium, these are performed from the anterior approach. The roof of the sciatic notch is identified (with the sciatic nerve and gluteal vessels protected), and a high speed burr used to perform the osteotomy. Care must be taken in orienting this correctly, given the oblique nature of the sacroiliac joints.

Fig. 5.

Sacrectomy . ( A ) Lateral fluoroscopy showing planned level of bony resection at the marked metal pin placed at lower border of S2; ( B ) bony division in prone patient position; and ( C ) off-midline skin incision for sacrectomy.

Where it is necessary to turn the patient prone for sacrectomy, the end colostomy and/or ileal conduit are fashioned, exteriorized, and matured at the premarked sites. The patient is then moved into the prone position. A posterior midline incision is made extending and encircling the anus as needed. In our practice, curved off-midline incisions sometimes offer better healing ( Fig. 5C ). The gluteus maximus muscles are subsequently dissected away from the sacral attachments, and the sacrospinous and sacrotuberous ligaments are divided to access the pelvic cavity posteriorly ( Fig. 5B ). Lateral fluoroscopy is used to visualize the level of the anterior osteotomy and to guide the level of bony transection ( Fig. 5A ). Laminectomy, dural sac ligation, and sacral resection are then performed at the required level. The resultant soft tissue defect then requires reconstruction by the plastic surgery team, and this most commonly takes the form of either a pedicled myocutaneous rectus abdominus flap, gluteal flaps, or biological meshes, depending on anatomy, preservation of the gluteal vessels at abdominal phase, local practice, and experience ( Fig. 6A–C ). Our preference is to employ a rectus abdominus flap for all sacral resections above the level of S3, wherever possible.

Fig. 6.

Perineal reconstruction after sacrectomy . ( A ) Patient in supine position with RAM flap harvested and lying in the inferior part of the wound prior to being passed into the perineum; ( B ) RAM flap being delivered into sacrectomy defect in prone position; and ( C ) biological mesh (4-mm-thick bovine ADM) used to reconstruct distal sacrectomy wound with bone division below S3 level. ADM, acellular dermal matrix; RAM, rectus abdominis muscle.

Lateral Compartment

This phase is generally considered to pose the greatest challenge in terms of resectability, owing to the major vascular and neural structures located therein. The key operative steps here involve dissection of the ureters, and circumferential mobilization and skeletonization of the common, internal and external iliac arteries and veins. For the ureters, the dissection will dictate whether the patient will need ureteric resection and reimplantation, for example, with a Boari flap, or whether an en bloc cystectomy/cystoprostatectomy with urinary conduit is required. In circumstances where no ureteric involvement is noted, then full ureterolysis should be performed and rubber vessel slings may be placed around each ureter for ease of identification throughout the dissection that will follow ( Fig. 7A ).

Fig. 7.

( A ) Vascular and ureteric vessel loops to enable identification of key anatomical structures; ( B ) aberrant venous anatomy with right internal iliac vein originating from the left common iliac vein; and ( C ) en bloc resection of the internal iliac vascular system with a recurrent tumor mass showing the underlying sciatic nerve (S), obturator nerve (white arrowhead), and piriformis muscle (P).

From a vascular perspective, planning of anticipated vascular dissection and resection can at times be aided significantly by preoperative CT or magnetic resonance angiography which can enable the surgical team to better define the extent of arteriovenous tumor involvement. Importantly, where there is suspicion, these scans may assist in identifying whether there is disease at the level of the iliac bifurcation or involving the external iliac vessels, and to plan accordingly for vascular reconstruction(s) if appropriate.

Typically, dissection commences by obtaining proximal and distal control by vessel looping of the common, external, and internal iliac vessels. This step also helps identify aberrant venous anatomy which is always a concern ( Fig. 7A, B ). Access to the internal iliac vein (IIV) is best provided following ligation and division of the internal iliac artery (IIA). If the decision is to proceed with resection of the sidewall, the next step is usually to divide the IIA and IIV and to dissect in the space lateral to these vessels.

Sciatic nerve roots, the obturator nerve, and piriformis are encountered in this field ( Fig. 7C ) and again depending on the anatomy of the disease, may need to be resected to ensure clear margins. Great care must be taken in the lateral dissection due to the ever-present threat of hemorrhage form thin-walled venous tributaries of the IIV. Where possible, the IIA may be divided inferior to the origin of the superior gluteal artery as the presence of this vessel may be required later for perineal reconstructive purposes if a gluteal flap is employed.

Perineal Phase

Where there is anticipated tumor extension distal to the levator sling, an infralevator resection and perineal approach will also be required. The incision in the perineum is tailored to the anatomy of disease, but typically, a skin-preserving approach may be used. Once through the skin and subcutaneous fat, the direction of dissection and degree of extension into the urogenital or anal triangles will typically depend on the a priori determined work-up. Where extensive, the dissection can be extended posteriorly to the sacrum, laterally to the ischial tuberosities, and anteriorly to the pubic rami and/or symphysis.

Intraoperative Radiotherapy

IORT is an adjunctive technique in which a single high fraction of radiation treatment is delivered during a surgical procedure on borderline resectable tumors, where one anticipates a close or involved cancer margin ( Fig. 8A, B ).

Fig. 8.

IORT in exenterative surgery . Recurrent tumor mass involving left pelvic sidewall on ( A ) MRI and ( B ) PET imaging. Imaging clearly suggests that even after an extended sidewall resection the pathological margin will still be closed. IORT, intraoperative radiotherapy.

The rationale for this approach is simple; radiotherapy has an established place in the treatment of radiation sensitive pelvic tumors, and historically, the greater the dose that has been delivered, the more tumoricidal effect and hence oncological benefit achieved. However, beyond a certain threshold, the radiation tolerance of healthy adjacent tissue can be exceeded, a fact that is particularly true in deep-seated pelvic tumors.

As a result of these limitations, IORT was developed to work synergistically with radical surgery, by enabling the delivery of very high single fraction treatment, escalating preoperative radiotherapy doses to levels not conventionally feasible, while simultaneously delivering treatment precisely to the most at-risk field, while simultaneously allowing shielding of radiation sensitive structures out of the targeted treatment zone. ^{19 20}

IORT is internationally delivered by two methods, high-dose rate brachytherapy or intraoperative electron beam radiotherapy (IOERT; Fig. 9A–D ) and is utilized in most international comprehensive cancer centers. ²¹ Evidence for the utility of IORT is supported by three key observations. Positive prospective long-term data from internationally renowned centers of excellence; low/absent tumor recurrence rates in IORT fields compared with non-IORT fields; and the convergence of survival figures between patients undergoing R0 surgery without IORT and those undergoing R1 surgery with IORT. In an international collaborative effort to synthesize the evidence in the field, we and others noted that despite study quality limitations and heterogeneity in design and methodology, a significant effect favoring improved local control (odds ratio: 0.22; 95% confidence interval [CI] = 0.05–0.86; p  = 0.03), overall survival (hazard ratio [HR]: 0.33; 95% CI = 0.2–0.54; p  = 0.001), and disease-free survival (HR: 0.51; 95% CI = 0.31–0.85; p  = 0.009) was noted with IORT treatment. Importantly, this oncological benefit seemed to take place without an increase in total, urologic, or anastomotic complications, although greater wound issues were noted. ²¹ As a result of this and other studies, IOERT is now part of the recommended treatment modalities by the National Comprehensive Cancer Network and to be used for locally advanced or recurrent rectal cancer.

Fig. 9.

Following resection of the involved sidewall, ( A ) the resection margin is predicted to be closed or involved (1.4 mm in this case). ( B ) As a result, an IOERT applicator is positioned by the clinical oncologist and surgical teams. ( C ) The margin of concern is visualized down the applicator. ( D ) The patient and applicator are aligned with a mobile self-shielding theater-based linear accelerator (IntraOp Medical Corp, Sunnyvale, CA) and up to 20 Gy boost may be provided. IOERT, intraoperative electron beam radiotherapy.

Conclusion

Multivisceral exenterative pelvic surgery has evolved and been refined significantly since its original description, through a combination of incremental advances in several fields including imaging, neoadjuvant therapies, resectional and reconstructive surgery, prehabilitation, and anesthetic and critical care medicine.

Despite the complexities of treatment, these advances have allowed the safe expansion of this field, with good oncological and quality of life outcomes being increasingly noted. Nevertheless, management decisions are complex, and treatment requires specialist multimodality multispecialty approaches. To date, high-quality prospective and randomized studies have been lacking in the field; however, as treatment becomes progressively centralized to specialist units able to provide the breadth of expertise and options needed for this patient group, this will also increasingly enable collaborative approaches for larger and higher quality prospective studies.

Funding Statement

Funding The authors did not receive any financial support for this manuscript.