Introduction

Sacrectomy as part of pelvic exenterative surgery, for locally advanced primary or recurrent pelvic malignancy, was first described in 1981 [1]. Despite a 35-year experience, the resection of tumours involving the sacrum remains a surgical challenge, associated with significant morbidity and not insignificant mortality [2, 3]. Following assessment of the patient, and staging via a positron emission tomography (PET) scan, a determination of the resectability of the tumour with magnetic resonance imaging (MRI) is performed. Where sacrectomy above the level of S3 is required, it is performed in the prone position [4]. This poses challenges in gaining control of any haemorrhage that may occur during sacrectomy once the patient has been placed prone and necessitates proactive control of potential bleeding. We describe a technique of pre-emptive dissection and ligation of the internal iliac venous plexus, and the three named venous tributaries are visceral, gluteal and presacral.

Materials and methods

A retrospective descriptive analysis was performed on 25 patients from January 2005 to December 2010, who underwent internal iliac vein triple tributary venous ligation, either unilaterally or bilaterally as part of a pelvic exenterative surgery and sacrectomy from January 2005 to December 2010 at the Royal Prince Alfred Hospital Sydney, Australia. Information on demographics, operative details, morbidity and mortality was collected.

Our approach to patient assessment is to perform a full clinical assessment, pelvic magnetic resonance imaging (MRI), a positron emission tomogram (PET) scan, and discuss at our multidisciplinary team (MDT) meeting. Patients with an American Society of Anesthesiologists (ASA) score greater than 3 would be deemed unfit for such radical surgery. An examination under anaesthesia (EUA) aids in determining the degree of en bloc visceral resection necessary and allows for further tissue biopsies to be performed, if necessary. MRI imaging in conjunction with EUA determines resectability, whereas PET scan assesses for incurable distant metastatic disease which would be a contraindication to surgery. Presacral fascial involvement or bony infiltration determines the level of sacral transection. Traditionally sacrectomy at S3 and below is performed in the abdomino-lithotomy position, while sacrectomy of S3 and above is performed in the prone position. Two illustrative MRI-road mapping views of patients requiring sacrectomy are shown in Figs. 1 and 2.

Fig. 1
figure 1

Illustrates an extensive presacral recurrent rectal cancer abutting the sacrum up to Mid-S1. The dotted lines demonstrate the lines of transection required to achieve an R0 margin. This scenario requires bilateral triple ligation of the vessels

Full size image
Fig. 2
figure 2

Illustrates a recurrent rectal cancer mass involving the left sacroiliac joint and lower S1/S2 vertebrae with encasement of the S2 nerve root. The dotted line demonstrates the lines of transection to achieve an R0 margin. This scenario required only a left-sided triple ligation of the vessels as a left hemisacrectomy was performed with preservation of the right sacrum

Full size image

Perineal and sacral wound complications are not common but tend to be of higher incidence in patients who have had radiotherapy, ligation of both superior gluteal vessels, a myocutaneous flap reconstruction or a complete soft tissue exenteration with en bloc sacrectomy due to the large empty space. As a result, when possible we try to preserve the rectum, for example, in patients with sacral chordomas or use the omentum to fill the defect to reduce the incidence of postoperative fluid collections, which ultimately discharge via the path of least resistance. Additionally, these patients undergo specialised nursing care until mobile to prevent them from developing pressure sores on their wounds that can result in dehiscence.

Surgical technique

The patient is placed in a modified Lloyd-Davies position, whereupon a thorough abdominal exploration for metastatic disease is performed. After adequate exposure of the operative field is established, a lymph node dissection is performed from the aortic bifurcation down to the origin of the internal iliac vessels.

The technique of triple tributary ligation proper then begins, with the previously described mobilisation and isolation of the common and external iliac vessels to the inguinal ligament [3]. The relationship or involvement of these vessels to the tumour is assessed, and a decision is made on whether a partial or complete vascular resection, with patch or bypass, is necessary. Vessel loops are placed at proximal and distal ends of the external iliac vessels, to enable vascular control during further dissection. As highlighted in a paper by Austin and Solomon [3], critical in the exposure of the internal iliac vein (IIV) is the identification and prior ligation of the internal iliac artery (IIA) (Fig. 3).

Fig. 3
figure 3

Anatomical relation of the left internal iliac vein posterior to the artery and the left ureter

Full size image

Once the IIV is exposed, it is ligated 0.5–1.0 cm from its junction with the external iliac vein (EIV). Dissection is then followed along the course of the IIV medially and inferiorly into the pelvis for approximately 5 cm. Three groups of veins are encountered during this step: gluteal, visceral and sacral. Large S1 and S2 foraminal veins are a feature of this dissection. Each of these groups of veins is ligated in situ (Fig. 4). With each branch ligated, the IIV collapses and can be safely divided (Fig. 5). If sacrectomy is required from the mid-body of S1–S3, it is necessary to further ligate distal branches or tributaries of the IIV to the anatomical level of transection. This may require two or three ligations of the visceral, gluteal and sacral foraminal tributaries. All vessels passing posteriorly are suture-ligated using 5/0 prolene. Additional gluteal tributaries in the plane of dissection when encountered are divided and suture-ligated. It is important to note that the most delicate part of this technique is taking the posterior and lateral veins off the posterior aspect of the IIV as they pierce the posterior muscle, enter the sacral foramina and enter the lesser sciatic foramina.

Fig. 4
figure 4

Schematic illustration demonstrating the three named branches of the internal iliac venous plexus and the triple tributary ligation performed

Full size image
Fig. 5
figure 5

Three named branches of the internal iliac vein following ligation

Full size image

Results

Triple tributary ligation during sacrectomy was performed in 25 patients (12 male, 13 female) by a single surgeon (AA E). The ages ranged from 20 to 80 years, with a mean age of 46.2 years. ASA grades for the cohort were: ASA 1 (n = 9), ASA 2 (n = 15) and ASA 3 (n = 1). The cancers resected included 19 (76%) primary tumours and 6 (24%) secondary tumours. The origins of the cancer included chordoma in 7 patients (28%) and recurrent colorectal cancer in 5 (20%). The remaining 52% of the cancers were a heterogenous group of predominantly primary cancers, with 1 case of metastatic melanoma (Table 1).

Table 1 Cancer Pathology in 25 patients undergoing triple tributary ligation during sacrectomy
Full size table

Table 2 outlines the procedural details of the sacrectomies performed. There were 13 unilateral and 12 bilateral sacrectomies. Nineteen operations were completed in a single stage, with 6 requiring two-stage procedures. Only 1 of the two-stage procedures was planned at the outset as two stages.

Table 2 Procedural details in 25 patients undergoing triple tributary ligation during sacrectomy
Full size table

Median operating time was 8.5 h (range of 2.32–19.67 h). Median blood loss in the cohort was 5500 mL (range of 1600–18000 mL). It should be noted that estimated blood loss for 4 patients was not recorded. Median transfusion of packed red blood cells was 9 units (range of 0–34 units). The average number of days in the intensive care unit was 1 day (range 0–10 days), with a median length of stay in hospital of 18 days (range 5–148 days) (Table 3).

Table 3 Operative and admission details in 25 patients undergoing triple tributary ligation during sacrectomy
Full size table

There was no intraoperative mortality in the cohort. There was one death at 30 days secondary to gram-negative septicaemia. Postoperative morbidity occurred in 17 (68%) patients (Table 4).

Table 4 Triple tributary ligation with sacrectomy in pelvic exenteration: morbidity and mortality details
Full size table

Discussion

Significant haemorrhage is a well-recognised problem in pelvic exenterative surgery requiring sacrectomy [2, 3], with increasing blood loss associated with more rostral sacrectomy [5]. It is well established that massive blood transfusion is independently associated with adverse outcomes, including death [6]. Thus, in pelvic exenterative surgery requiring sacrectomy at S3 or above, it is essential to take pre-emptive measures to minimise blood loss prior to positioning the patient prone for sacrectomy, whereupon haemorrhage control is compromised.

The connection between the internal iliac veins and the epidural venous plexus was described by Batson in the 1940s [7, 8]. Crucially in the context of haemostasis, this communication between the presacral plexus and the paravertebral plexus is valveless [9]. Experimental studies have demonstrated that blood loss from presacral veins can reach 1000 mL/min, with an increase in the vein diameter of only 1 mm leading to a threefold increase in blood loss [10]. Therefore, unlike arterial bleeding which is readily controlled in most circumstances by suture ligation or electrocauterisation, venous bleeding can be both torrential and endless [11].

Some authors have advocated preoperative embolization as an alternative pre-emptive modality, as is seen in the emergency setting for pelvic haemorrhage related to trauma. However, its use in sacral resection during pelvic exenteration may lead to more haemorrhage with the formation of pelvic collaterals [12]. Similarly, isolated ligation of the internal iliac vein without ligation of the tributaries has been demonstrated to increase pressure in the paravertebral venous plexus in an animal model and postulated as a significant factor in major haemorrhage during sacrectomy [13]. Our technique differs in that each tributary is ligated separately to minimise significant pressure shunts within the pelvic venous plexus.

Conclusions

It is not possible to draw statistically significant conclusions from this small, heterogenous cohort of a single institution, and further investigation is required. We have, however, demonstrated the technical feasibility of triple tributary ligation. Its use has not been associated with any appreciable increase in adverse sequelae or consequences. This technique has broader relevance for visceral resections where the pelvic side wall fascia, obturator internus or piriformis muscles are also involved.