Objective
The objective of the study was to further characterize the vascular and ureteral anatomy relative to the midsacral promontory, a landmark often used during sacrocolpopexy, and suggest strategies to avoid complications.
Study Design
Distances between the right ureter, aortic bifurcation, and iliac vessels to the midsacral promontory were examined in 25 unembalmed female cadavers and 100 computed tomography (CT) studies. Data were analyzed using Pearson χ 2 , unpaired Student t test, and analysis of covariance.
Results
The average distance between the midsacral promontory and right ureter was 2.7 cm (range, 1.6–3.8 cm) in cadavers and 2.9 cm (range, 1.7–5.0 cm) on CT ( P = .209). The closest cephalad vessel to the promontory was the left common iliac vein, the average distance being 2.7 cm (range, 0.95–4.75 cm) in cadavers and 3.0 cm (range, 1.0–6.1 cm) on CT ( P = .289). The closest vessel to the right of the promontory was the internal iliac artery, with the average distance of 2.5 cm (range, 1.4–3.9 cm) in cadavers and 2.2 cm (range, 1.2–3.9 cm) on CT ( P = .015). The average distance from the promontory to the aortic bifurcation was 5.3 cm (range, 2.8–9.7 cm) in cadavers and 6.6 cm (range, 3.1–10.1 cm) on CT ( P < .001). The average distance from the aortic bifurcation to the inferior margin of the left common iliac vein was 2.3 cm (range, 1.2–3.9 cm) in cadavers and 3.5 cm (range, 1.7–5.6 cm) on CT ( P < .001).
Conclusion
The right ureter, right common iliac artery, and left common iliac vein are found within 3 cm from the midsacral promontory. A thorough understanding of the extensive variability in vascular and ureteral anatomy relative to the midsacral promontory should help avoid serious intraoperative complications during sacrocolpopexy.
Hemorrhage and ureteral injury are infrequent but serious complications of sacrocolpopexy, with reported rates of 4.4% and 1%, respectively. These complications generally occur during dissection of the presacral space to expose the anterior longitudinal ligament and during suture placement.
Modifications of the sacrocolpopexy, including the location of the graft attachment have evolved as surgeons encountered and reported complications. For example, the report by Sutton et al of a life-threatening hemorrhage during graft fixation at the S3-S4 level led authors to recommend a graft fixation at the S1-S2 level to better visualize the middle sacral vessels. Current descriptions of the procedure advocate graft attachment to the anterior longitudinal ligament at the level of sacral promontory. However, surgeons often choose the sacral fixation site based on intraoperative findings because the vascular anatomy of the presacral space is highly variable.
The incorporation of the laparoscopic and robotic-assisted approach to sacrocolpopexy has introduced further challenges. Because dissection and suturing may be technically difficult due to the marked angle between the lower lumbar vertebrae and the anterior surface of the sacrum, surgeons may choose fixation points above the sacrum. Importantly, 2 previous cadaver studies have shown that the left common iliac vein may course less than 1 cm cephalad to the midsacral promontory (MSP). The proximity of the left common iliac vein (LCIV) to the MSP makes it especially vulnerable to injury when dissection and suture placement extends above the upper margin of the sacrum, potentially leading to catastrophic complications.
The lack of haptic feedback with the robot-assisted approach adds to the challenges of identifying important anatomical structures, such as the aortic bifurcation and sacral promontory. In addition, electrosurgical energy is widely utilized for tissue dissection with the minimally invasive techniques. This may potentially lead to unrecognized vascular and ureteral injuries that manifest days to weeks postoperatively. Indeed, delayed diagnosis of ureteral injury has been reported with the laparoscopic approach.
Few studies have evaluated the vascular anatomy of the presacral space using the midpoint of the sacral promontory as a reference point, and even fewer have examined the relationship of the ureter to this important anatomic landmark. Two studies assessed the relationship of the great vessels to the MSP and correlated these findings to the sacrocolpopexy. Other studies from the orthopedic literature described measurements from the great vessels to the S1-S2 level and the distance from the aortic bifurcation to the inferior margin of L5. Additionally, only 2 studies evaluated the relationship of the ureters to the MSP, one in 10 cadavers and the other in 38 computed tomography (CT) urograms.
In our review of the literature, no studies concurrently evaluated the vascular and ureteral anatomy in cadavers and CT images using the midpoint of the sacral promontory as a reference point. In addition, no studies have assessed the distal extent of the left common iliac vein from the termination point of the abdominal aorta, a landmark that surgeons may be able to see or palpate intraoperatively. Thus, the objectives of this study were to further characterize the vascular and ureteral anatomy using the midsacral promontory as a reference point and to correlate the findings with a safe dissection and sacral mesh attachment during sacrocolpopexy.
Materials and Methods
Presacral space anatomy was examined in 25 unembalmed female cadavers and 100 CT studies. The cadavers were obtained from the Willed Body Program at the University of Texas Southwestern Medical Center in Dallas. Unembalmed female cadavers used for educational courses not involving pelvic dissection and without evidence of pelvic malignancy were examined, with permission from the Willed Body Program and the course directors.
All available cadavers that met the above criteria from August 2011 to August 2012 were evaluated. All CTs obtained for the evaluation of hematuria or recurrent urinary tract infections in women from February 2011, the time at which this radiological technique was introduced in our institution, to May 2012 were eligible for inclusion. Studies were excluded if evidence of urogenital malignancy, bony metastasis, or previous pelvic surgery was identified because these conditions are known to significantly alter anatomy.
The cadaveric portion of this study was considered exempt by the University of Texas Southwestern Institutional Review Board in accordance with the Code of Federal Regulations. Review of radiographic images and demographic data was approved by University of Texas Southwestern Medical Center and Parkland Hospital Institutional Review Boards. Age, race, height and weight, and cause of death were collected.
Cadaver data
Gross support of the uterus and vaginal walls was examined prior to the dissection in all cadavers. This was accomplished by placing a single tooth tenaculum on the cervix and vaginal walls and applying gentle downward traction. In cadavers, the abdomen was entered and the posterior abdominal wall peritoneum was incised vertically from the level of the aortic bifurcation to the posterior cul de sac and pelvic floor. The incision was kept medial to the right ureter and lateral to the right border of the rectosigmoid, just medial to the right uterosacral ligament. Care was taken to not disrupt the anatomic location of the right ureter. Ureteral stents were placed in the right ureter to help with the identification of its medial surface. The fat and loose connective tissue deep to the peritoneum was sharply and bluntly dissected to expose the sacral promontory, which was defined as the most superior point on the anterior surface of the first sacral (S1) vertebra.
The MSP was marked with a metal pin and used as the reference point for all of the measurements ( Figure 1 ). The MSP was determined as the midpoint between the junctions of the body of the fist sacral vertebra (S1) with the ala of the sacrum. Distances measured from the MSP included the following: (1) the medial aspect of the right ureter; (2) the closest great vessel on the right; (3) the closest great vessel on the left; (4) the closest cephalad vessel; and (5) aortic bifurcation. In addition, the vertical distance from the aortic bifurcation to the inferior margin of the LCIV was measured ( Figure 1 ). Measurements were obtained using the same rotating caliper and plastic ruler. All measurements were taken twice by female pelvic medicine and reconstructive surgeons (M.M.G. and M.M.C.), and photographs were taken of all of the dissections.
Computed tomography data
The Parkland Hospital radiological database was queried for all women who underwent a CT urogram for the diagnosis of hematuria or recurrent urinary tract infections since implementation of this diagnostic modality at this institution. Subjects with a diagnosis of malignancy or previous pelvic or urological surgery were excluded because these conditions are known to significantly alter anatomy.
All examinations were performed on a 64-section multidetector CT scanner (Aquillion 64; Toshiba Medical Systems) using the standard 3-phase protocol (unenhanced, nephrographic, and urographic) used at our institution. Patients were positioned supine on the CT couch, and an unenhanced acquisition was first obtained from the superior aspect of the kidneys through the pubic symphysis. Iohexol (Omnipaque 300; GE Health Care, Indianapolis, IN), 100 mL was injected intravenously at 3 mL/sec, and a nephrographic phase was obtained after 120 seconds. The urographic phase was acquired 10 minutes after contrast injection from the superior aspect of the kidneys through the pubic symphysis. Prone imaging was then performed, as needed, to opacify the entirety of both ureters. The unenhanced and nephrographic phases were reconstructed as 5 mm thick images, and the urographic phase was reconstructed as 2 mm thick images.
For optimal accuracy of the measurements, the 2 mm images in the urographic phase of the examination were evaluated. A 3-dimensional workstation (Aquarius; TeraRecon, San Mateo, CA) was used to reconstruct the images. Image evaluation began by determining the axial, coronal, and sagittal midpoint of the MSP, which served as reference point for all measurements. In the transverse plane, the distance between the medial border of the right ureter and the MSP was measured ( Figure 2 ). In the same transverse plane, the vascular structure nearest the MSP was identified on the right and left, and the closest distance between this vascular structure and the MSP was measured. An oblique sagittal plane was then used to measure the distances between the MSP and aortic bifurcation and MSP and inferior margin of the LCIV. In the same oblique plane, distances between the aortic bifurcation and the inferior margin of the LCIV were measured. All measurements were obtained twice, once by the participating radiologist (T.A.A.) and once by the female pelvic medicine and reconstructive surgeon (M.M.G.); the average of the 2 measurements was used in the data analysis.
Statistical analysis
Demographics were compared between the 2 groups using an unpaired Student t test; race/ethnicity were compared using a Pearson χ 2 test. All other measures are continuous and were compared using the unpaired Student t test. Ultimately the association between the outcome measures and graph assignment was adjusted for the demographic differences using an analysis of covariance. Statistical analysis was accomplished using SAS, version 9.2 (SAS Institute, Cary, NC).
Results
Twenty-five unembalmed female cadavers and 100 CT urograms were evaluated. Of 161 patients identified from the radiological database, 61 were excluded because of a diagnosis of malignancy or previous pelvic or urological surgery. Limited demographic characteristics of both study groups were available and are provided in Table 1 . The most common cause of death in the cadaver group was nongynecological cancer; the time of death ranged from January 2011 to July 2012. None of the cadavers had descent of the vaginal walls or cervix to or beyond the hymen. None of the CT urogram subjects had ureteral stents. Pelvic organ support in the CT group was not known.
| Characteristics | Cadavers (n = 25) | CT urograms (n = 100) | P value |
|---|---|---|---|
| Age, y | 71.9 ± 18.1 | 51.4 ± 14.4 | < .001 |
| Race | < .001 | ||
| White | 23 (92) | 15 (15) | |
| Black | 1 (4) | 25 (25) | |
| Hispanic | 0 (0) | 56 (56) | |
| Other | 1 (4) | 4 (4) | |
| BMI, kg/m 2 | 23.7 ± 5.6 | 31.9 ± 7.49 | < .001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
