DENONVILLIERS’ FASCIA AND LATERAL LIGAMENTS OF RECTUM

DENONVILLIERS’ FASCIA

Nano et al., Hepato-Gastroenterology. 45(21):717-26, 1998 May-Jun.

These investigators studied the surgical anatomy of the main structures involved in the mobilisation of the rectum. These structures are – Denonvilliers’ fascia, mesorectum, middle rectal artery, superior and inferior hypogastric plexuses.

Denonvilliers’ fascia originates from the fusion of the primary pouch of Douglas. The two leaves which make up this structure are easily separated. The mesorectum surrounds the rectum on three sides; it is easily separated from the presacral fascia and its connective tissue is bordered by a thin membrane. Variant anatomy of the middle rectal artery is the rule.

Interestingly, Kourambas J et al., British Journal of Urology. 1998 Sept, 82(3):408-10, found that Denonvilliers fascia has no clearly defined layers and no defined lateral edge. No distinct neurovascular bundle was seen but nerves were scattered throughout the fasciae, including medially towards the midline.

The study by Nano et al on Denonvilliers’ fascia was performed on fetuses and not on adult cadavers. The latter study was performed on adult cadaver. Hence the difference.

 

The entire rectum and the other pelvic organs are surrounded by the endopelvic fascia. Muntean V. The surgical anatomy of the fasciae and the fascial spaces related to the rectum, Surgical and Radiological Anatomy. 21(5):319-24, 1999 have defined the anatomy of this based on 46 fresh cadaveric dissections. The rectum is attatched to the urogenital and presacral fasciae by rectal stalks, which take a spiral course around the rectum, being posterolateral in the upper rectum, lateral in the mid rectum and anterolateral in the lower rectum. Denonvilliers’ fascia is therefore angled forwards and downwards.

My thoughts

During surgical mobilisation of the rectum, after the posterior plane is defined, the rectum must be pulled up and pressed posteriorly, while the St.Mark’s retractor retracts the seminal vesicles, bladder and the prostate out of the way. This way, the fascia becomes taut and is now in a horizontal plane, easily amenable to incision. If this is not followed, then the Denonvilliers’ fascia gets stripped off the rectum with haemorrhagic consequences.

LATERAL LIGAMENTS

Jones et al., ‘Lateral ligaments of the rectum: an anatomical study’; British Journal of Surgery 1999, 86, 487-489.

Authors performed a dissection in the mesorectal plane in 28 cadaveric pelves. In 10 no connective tissue structure crossed from the pelvic side wall to the rectum. The remaining 18 had only very insubstantial connective tissue strands crossing this space. A total of 17 middle rectal arteries were found, all of them unilateral. 14 of these vessels crossed the mesorectum independent of any structures, while the remainder were part of a neurovascular bundle with a connective tissue element.

Authors proposed that the ‘lateral ligaments’ of the rectum do not exist and that the term should be dropped from surgical texts. When present, the middle rectal artery is a small vessel, close to the pelvic floor. The entire rectum may be mobilised by sharp dissection without the need for clamping or ligation of any significant structure.

Discussion

Little original work exists on the anatomy of the lateral ligaments. Gray’s Anatomy states that ‘around the middle rectal vessels, fascia extends from the posterolateral pelvic wall (level with the third sacral vertebra) to the rectum as the lateral rectal ligaments’. From a surgical standpoint, Goligher and Duthie describe these structures, as seen from above after ‘a certain amount of definition’, as having ‘a roughly triangular shape with the base on the pelvic side wall and the apex joining the side of the rectum’. By contrast, in the present study no significant structures crossing the mesorectal plane were identified. Insubstantial thin strands of connective tissue were sometimes found, but could not be thought to represent the lateral ligaments for a number of reasons. Most importantly, these strands of connective tissue were no different from those one would expect to find in any areolar plane. Furthermore, their extreme variabiligy both in terms of their height above the pelvic floor and direction of travel suggested that they were not constant structures. Indeed, they were often absent altogether.

There is significant variation in published data on the incidence of the middle rectal artery. The reange is from 12% to 100% of pelves. This probably reflects variation in the concept of what constitutes a vessel and differences in the method used to trace them. The present finding of a unilateral middle rectal artery in alittle over half of the pelves is supported by the work of DiDio et al. (57%) and Vogel and Klosterhalfen (50%). It is interesting to note that this vessel, when present, never exceeds 2 mm in diameter. In some respects the present findings are similar to those of Boxall et al. who conducted a very detailed study of the middle rectal artery. The middle rectal artery is generally small, and at or just above the pelvic floor. However, the authors found no evidence of any ‘accesory middle rectal artery’ nor were any structures found that could be termed a lateral ligament.

The relationship of the middle rectal artery to the lateral ligaments is also the subject of some controversy. The majority view is that the middle rectal vessels run in the lateral ligaments. Boxall and colleagues take a different view, claiming the lateral ligaments contain only accessory branches of the middle rectal artery, found in about 25% of pelves. They describe the middle rectal vessels as crossing the mesorectum in condensations of fascia, although quite separately from the lateral ligaments. In this study, the middle rectal artery crossed the mesorectal plane independently of any connective tissue structure.

Tiny branches from the autonomic pelvic plexi were seen crossing to the rectum. These structures were almost universally found to be distinct from the middle rectal vessels throughout their length and to have no association with connective tissue. They therefore cannot be confused with any ‘ligamentous’ structure and certainly do not require ligation and clamping as described in surgical texts. The authors do not believe therefore that they in themselves represent a lateral ligament or could considered a structure a surgeon could ‘hook’ a finger around. These results are in accord with the postulate that the lateral ligament is nothing more than a surgical artefact that results from injudicious dissection.

Taken together, these findings suggest that the lateral ligaments of the rectum do not exist. There is no anatomical argument against sharp dissection in the mesorectal plane and as a rule there is no vessel that requires clamping. By ‘hooking’ the finger into the tissue lateral to the rectum it may that the surgeon encounters mesorectal vessels and creates an artefactual ligament. This obviously raises the grave concern that such blunt dissection results in mesorectal tissue being left behind and increases the risk of local recurrence and severe autonomic nerve injury.

My comments

This is a botched study. The authors have not used any dye injection method. Whereas Michels et al in their DCR paper have injected latex with India ink into all rectal arteries to define the anatomy accurately in 400 dissection, the authors of the present paper have ‘looked’ for the vessel in a cadaver, where it is obvious that vessels are collapsed especially ones which are 2 mm in diameter. I would suggest that they have missed the Middle rectal artery in their dissections and come to some seriously fallacious conclusions.

 

This is a nice schematic diagram (taken from Delancey’s paper in Neurobiology of incontinence) of the female pelvis showing the paraurethral and paravaginal attatchments to the levator ani.

The dotted line defines the endopelvic fascia around the vagina. In the absence of the vagina (in the male)this fascia is the denonvilliers’ fascia which embryologically is in two layers.