Available online at www.sciencedirect.com
ScienceDirect EJSO 41 (2015) 738e745
www.ejso.com
Whole mount microscopic sections reveal that Denonvilliers’ fascia is one entity and adherent to the mesorectal fascia; implications for the anterior plane in total mesorectal excision? A.C. Kraima a,b, N.P. West b, D. Treanor b, D.R. Magee b, H.J. Rutten c, P. Quirke b, M.C. DeRuiter a, C.J.H. van de Velde d,* a
Department of Anatomy and Embryology, Leiden University Medical Center, P.O. Box 9600, 2300 ZC Leiden, The Netherlands b Pathology and Tumour Biology, Leeds Institute of Cancer and Pathology, University of Leeds, St. James’s University Hospital, Beckett Street, Leeds LS9 7TF, United Kingdom c Department of Surgery, Catherina Hospital Eindhoven, P.O. Box 1350, 5602 ZA Eindhoven, The Netherlands d Department of Surgery, Leiden University Medical Center, P.O. Box 9600, 2300 RC Leiden, The Netherlands Accepted 25 March 2015 Available online 2 April 2015
Abstract Background: Excellent anatomical knowledge of the rectum and surrounding structures is essential for total mesorectal excision (TME). Denonviliers’ fascia (DVF) has been frequently studied, though the optimal anterior plane in TME is still disputed. The relationship of the lateral edges of DVF to the autonomic nerves and mesorectal fascia is unclear. We studied whole mout microscopic sections of en-bloc cadaveric pelvic exenteration and describe implications for TME. Methods: Four donated human adult cadaveric specimens (two males, two females) were obtained from the Leeds GIFT Research Tissue Programme. Paraffin-embedded mega blocks were produced and serially sectioned at 50 and 250 mm intervals. Sections were stained with haematoxylin & eosin, Masson’s trichrome and Millers’ elastin. Additionally, a series of eleven human fetal specimens (embryonic age of 9e20 weeks) were studied. Results: DVF consisted of multiple fascial condensations of collagen and smooth muscle fibres and was indistinguishable from the anterior mesorectal fascia and the prostatic fascia or posterior vaginal wall. The lateral edges of DVF appeared fan-shaped and the most posterior part was continuous with the mesorectal fascia. Fasciae were not identified in fetal specimens. Conclusion: DVF is adherent to and continuous with the mesorectal fascia. Optimal surgical dissection during TME should be carried out anterior to DVF to ensure radical removal, particularly for anterior tumours. Autonomic nerves are at risk, but can be preserved by closely following the mesorectal fascia along the anterolateral mesorectum. The lack of evident fasciae in fetal specimens suggested that these might be formed in later developmental stages. Ó 2015 Elsevier Ltd. All rights reserved.
Keywords: Total mesorectal excision; Denonvilliers’ fascia; Rectal cancer; Mesorectal fascia; Autonomic nerves
Introduction The introduction of total mesorectal excision (TME) radically improved the surgical treatment and outcome of * Corresponding author. Leiden University Medical Center, Department of Surgery, K6-R, P.O. Box 9600, 2300 RC Leiden, The Netherlands. Tel.: þ31 71 5262309; fax: þ31 71 5266750. E-mail address: [email protected] (C.J.H. van de Velde). http://dx.doi.org/10.1016/j.ejso.2015.03.224 0748-7983/Ó 2015 Elsevier Ltd. All rights reserved.
rectal cancer. The TME principle involves en-bloc removal of the diseased rectum and surrounding mesorectum within an intact mesorectal fascia. Dissection in the ‘holy plane’ along the mesorectal fascia is said to enable preservation of the autonomic nerves.1 Tumour involvement of the circumferential resection margin and incomplete mesorectal excision are the most important predictors for recurrent disease, emphasizing the importance of en-bloc removal of
A.C. Kraima et al. / EJSO 41 (2015) 738e745
an intact mesorectal package with no disruptions, tears and/ or perforations.2,3 The Dutch TME trial showed that surgical damage to the autonomic nerves was the major cause of post-operative anorectal and urogenital dysfunction.4,5 As a consequence, posterior, lateral and anterior dissection planes were defined in TME to warrant complete removal of the mesorectum as well as identification and preservation of the autonomic nerves.6e9 For such an important area, it is perhaps surprising that the precise relationships between the autonomic nerves and peri-rectal fasciae, which form the dissection planes for rectal surgeons, are still debated.10 Denonvilliers’ fascia (DVF) has been frequently studied and so far there is no consensus on its embryological origin and topological anatomy. Some have argued that DVF was more closely related to the prostate,6,7 whilst others believed that DVF was situated closer to the rectum11 or was even adherent with the anterior mesorectum.12 Due to these contrary descriptions, the anterior plane in TME is frequently disputed. Some advocate dissection posterior to DVF,13,14 some argue to ‘split the layers’ of DVF,15,16 whereas others are convinced that optimal TME should be performed anterior to DVF.12,17e19 Contradictory views on the anterior dissection plane in TME risk suboptimal surgery and consequent poorer oncologic and functional outcome. Tumour involvement of the circumferential resection margin is most frequently reported in case of anterior tumours, which are anatomically closely related to DVF.20 In addition, incomplete mesorectal excisions are often reported.21 Patients requiring an abdominoperineal excision (APE) of the rectum and anus suffer from a poorer oncologic outcome as low rectal tumours are technically more difficult to resect. Specifically in advanced tumours, APE has an increased risk of tumour involvement of the circumferential resection margin in comparison with anterior resection specimen.22,23 Excellent knowledge of the complex pelvic anatomy is a prerequisite to optimize the oncological and functional outcome of rectal cancer. However, to what extent can we enhance the outcomes in TME if the anterior plane is still questioned? We aimed to study DVF in whole mount microscopic sections of adult pelvic exenteration specimens and concentrated on the morphological and topological anatomy, the anterior plane in TME and the lateral edges in relation to the autonomic nerves and mesorectal fascia. Additionally, a developmental series of human female and male fetuses was analysed to gain better insight in the development of endopelvic fasciae. Methods Adult cadaveric specimens Four human adult pelvic exenteration specimens were obtained through the University of Leeds GIFT Tissue Research Programme (www.gift.leeds.ac.uk) from consented donor bodies belonging to two males and two females. Ethical
739
approval was granted by the Northern and Yorkshire Regional Ethical Committee, Jarrow, UK (unique reference number 11/H0903/6). There was no history of pelvic surgery or pelvic pathology at post mortem examination. The specimens were retrieved during a tissue donation autopsy performed at St. James’s University Hospital in Leeds in with the body in the prone jack-knife position according to the extralevator APE technique as described by Holm et al.24 The specimens were essentially pelvic exenterations including the anal canal and rectum up to the recto-sigmoid junction, anal sphincters, perineal body, mesorectum with an intact mesorectal fascia, levator ani muscle, obturator internus muscle, vagina or prostate and penile bulb and posterior bladder wall. After fixation in 8% formaldehyde solution for seven days, the specimens were transversely sectioned at one centimetre. The slices were photographed and dissected to fit in Super Mega Cassettes measuring 74.8 52.5 16.5 mm (CellPath; Powys; UK). All tissues were dehydrated in graded ethanol and embedded in paraffin mega blocks. In addition, a developmental series of seven human female fetal pelvic specimens (embryonic age of 10, 12, 14, 15, 16, 19, and 20 weeks) and four human male fetal pelvic specimens (embryonic age of 9, 10, 12 and 20 weeks) were studied from collections in the Department of Anatomy & Embryology, Leiden University Medical Centre and the University of Warsaw, Poland. All fetuses were obtained with informed consent after miscarriage or legal abortion and were free of congenital pelvic malformations. The fetal specimens were dehydrated in graded ethanol and xylene, and embedded in paraffin blocks. Histological staining The mega blocks of the pelvic exenteration specimens were transversally cut in serial 5 mm sections. In one male and one female specimen every 10th section was collected onto glass slides and stained with haematoxylin and eosin (H&E), creating a series with a cross-sectional interval of 50 mm. Additional sections were collected from each mega block and stained with Masson’s trichrome (MT) and Millers’ elastin (ME).25 In the other male and female specimen every 48th, 49th and 50th section were collected, creating three series with a cross-sectional interval of 250 mm, of which one series was stained with H&E and one series with MT. The remaining series was kept for additional stains with ME. The paraffin blocks containing the fetal pelvic specimens were serially cut in transverse sections of 8 and 10 mm and alternately stained with H&E and azan to reveal collagen. A series was stained using antibodies against alpha-smooth muscle actin (SMA; Sigma-Aldrich, A2547) and S-100 (S100; DAKO, Z-031101) to detect smooth muscle fibres and the peripheral neural network, respectively. The protocol used can be explored online at: www.caskanatomy.info/research.
740
A.C. Kraima et al. / EJSO 41 (2015) 738e745
Image acquisition All stained glass slides from the adult pelvic exenteration specimens were digitally scanned in Leeds with an Aperio XT slide scanner (Aperio, San Diego, California, USA) at 20 magnification, creating a resolution of 0.46 microns per pixel. The digital images were compressed with JPEG2000 quality 70 and viewed in Aperio ImageScope version 10.2.2.2319. Selected glass slides from the fetal pelvic specimens were photographed using an Olympus AX70-microscope with an Olympus D12 camera. Results Adult pelvic exenteration specimens In the male exenteration specimens, DVF was located between the seminal vesicles, prostate and mesorectum. It
extended from the top of the seminal vesicles and terminated in the perineal body. DVF covered the posterior surface of the seminal vesicles and consisted of a fascial condensation of collagen and smooth muscle fibres. Here, DVF was indistinguishable from the mesorectal fascia (Fig. 1a). At a more inferior level, DVF was situated posteriorly and adjacent to the prostate. Even at high microscopic magnification DVF could neither be distinguished from the prostatic fascia nor from the mesorectal fascia, indicating that they were fused. Laterally, the edges of DVF appeared fan-shaped and curved along the anterolateral angle of the mesorectum. The posterior parts of DVF were continuous with the mesorectal fascia. A small intervening space was detected between these fasciae filled with adipose tissue and several small nerve fibres (Fig. 1b). At successive inferior levels, multiple small nerves and ganglionated nerve plexuses were located in this intervening space. The prostatic fascia merged with the endopelvic
Figure 1. This figure shows the relationship of DVF to the seminal vesicles (SV) and prostate (P). The inferior hypogastric plexus (arrows in window a) is located laterally to the mesorectal fascia (arrowheads in window a). The arrows in detail window a.I show the lateral edges of DVF at superior level. The lowest arrow points out the part that fuses with the mesorectal fascia (arrowhead). The arrows in detail window a.II show that DVF is adherent to the mesorectal fascia (arrowheads). At inferior levels, nerves (N) are located in the intervening space (star in detail window b.I) which is located between the prostatic fascia (lower arrow) and mesorectal fascia (arrowheads). The prostatic fascia fuses with the endopelvic fascia (upper arrow). The rightmost arrowheads show that DVF and the mesorectal fascia are continuous. The arrow in window b.II shows that DVF cannot be distinguished separately from the mesorectal fascia (arrowheads). ME: Millers’ elastin, MT: Masson’s trichrome, O: obturator internus muscle, A: apex of prostate, R: rectum, M: mesorectum. Scale bars in windows a and b: 8 mm, windows a.I and a.II: 800 mm, window b.I: 2 mm and window b.II: 1 mm.
A.C. Kraima et al. / EJSO 41 (2015) 738e745
741
Figure 2. This figure illustrates the lateral edges of DVF at the prostatic base (P). Detail window a.I shows the multi-layered lateral edges. Between these layers, small nerve fibres can be detected (arrows). Note the close relation of DVF with the outer longitudinal layer of the rectal wall (OLL). The arrow in detail window a.II shows that at high magnification DVF is as closely related to the prostatic fascia as to the mesorectal fascia (arrowhead). ME: Millers’ elastin, LAM: levator ani muscle, R: rectum, M: mesorectum, ICL: inner circular layer of the rectal wall. Scale bars in window a: 8 mm, window a.I: 2 mm, window a.II: 300 mm.
fascia posterolaterally. At the apex of the prostate, the lateral continuation of DVF appeared less organized as multiple fascial condensations reached the pelvic side wall. Small nerve fibres, blood vessels and adipose tissue were still detectable between these fascial layers (Fig. 2). Due to the natural pelvic tapering, the mesorectum shrank in volume inferiorly. The anterior mesorectum was less voluminous relative to the posterolateral mesorectum, resulting in a closer relationship between the anterior rectal wall and DVF. The outer longitudinal layer of the rectal muscular wall approximated DVF at certain points. DVF narrowed at the level of the internal urethral orifice before
anchoring in the perineal body. At high magnification, the parts of DVF located posterior to the prostate consisted of a multi-layered condensation of collagen and smooth muscle fibres with a paucity of elastin. Small blood vessels and nerve fibres ran between these layers over the whole horizontal axis of DVF but were more abundant at inferior levels. In the female exenteration specimens, DVF was situated between the cervix uteri, vagina and rectum. It extended from the pouch of Douglas’ to its termination in the perineal body. The morphological anatomy was identical to that of the male specimens, consisting of a multi-layered
Figure 3. This figure shows the fan-shaped lateral edges at the level of the vagina (V). The arrows in detail window a.I show small nerve fibres lateral to DVF. The mesorectal fascia at the anterolateral mesorectum in this female specimen is separate from the lateral edge of DVF (arrowheads detail window a.I). Detail window a.II shows that DVF and the mesorectal fascia (arrowheads) are fused in the midline. In certain places, DVF is torn from the mesorectum (arrow) demonstrating the delicateness of the whole mount microscopic sections and the adherent DVF and anterior mesorectum. At a more inferior level, nerves (arrowheads in detail window b.I) are located in the intervening space (star in detail window b.I) which formed by the lateral extension of DVF, the mesorectal fascia and the endopelvic fascia. DVF is laterally continuous with the mesorectal fascia (arrowheads detail window b.I). DVF is adherent to the mesorectal fascia (arrowheads detail window b.II) and the posterior vaginal wall (VW). ME: Millers’ elastin, MT: Masson’s trichrome, B: bladder, R: rectum, M: mesorectum, U: urethra, O: obturator internus muscle. Scale bar in window a: 7 mm, window a.I: 3 mm, window b: 8 mm, window b.I:2 mm, windows a.II and b.II: 800 mm.
742
A.C. Kraima et al. / EJSO 41 (2015) 738e745
condensation of collagen and smooth muscle fibres. At the level of the cervix uteri and upper vagina, the lateral continuations of DVF appeared fan-shaped as well. In the midline, the posterior part of DVF was fused with the mesorectal fascia and in a lateral direction it curved along the mesorectal fascia. No nerve fibres were detected between these laterally curving edges of DVF and the mesorectal fascia (Fig. 3a). The inferior hypogastric plexus was identified laterally to the mesorectal fascia and ran along the curved mesorectum in an anterior direction. At a more inferior level, an identical intervening space described in the male specimens was formed in the female specimens by an angle between the lateral vagina, pelvic side wall and fused DVF and mesorectal fascia (Fig. 3b). The part of DVF located posteriorly to the vagina was comprised of more elastin, although most elastin was detected in the anterior vaginal wall. DVF was fused with the posterior vaginal wall over its whole vertical axis (Fig. 4). At high magnification, the inner and outer muscular layers of the vaginal wall were closely related and were invested by extensive vascular plexuses.
Fetal pelvic specimens In all fetal specimens, the mesorectum was recognized as a peripheral layer of densely packed mesenchymal cells surrounding the rectum (Fig. 5). There was a clear difference in density of the mesenchymal cells forming the mesorectum and the mesenchymal cells giving rise to the extraperitoneal tissues outside the mesorectum. This border represented the future mesorectal fascia. Collagen depositions were seen on visceral and parietal surfaces in the female foetuses aged 14 weeks and older, and the male fetus aged 20 weeks, but no evident fasciae were detected that would correspond to the adult morphological characteristics of fasciae. In both female and male specimens, the anterior mesorectum reached the vagina and prostate, respectively. At this site, an area rich in densely-packed collagen was detected that was continuous with the future anterolateral mesorectal fascia. We identified this area as the future DVF. At successive inferior levels, the attachment of the anterior mesorectum to the vagina and prostate narrowed until the levator ani muscle. The SMA stained sections revealed the smooth musculature surrounding the cervix uteri and vagina. From the pouch of Douglas until the perineal body, smooth muscle fibres invested to some extent the layer of densely packed mesenchymal cells. Autonomic nerves were located anterolaterally to the mesorectum at the level of the cervix uteri and upper vagina. We did not observe major nerve fibres belonging to inferior hypogastric plexus within the mesorectum. Only a minority of small nerve fibres ran between the rectum and vagina or prostate. In none of the fetal specimens, the relation between autonomic nerves and the lateral continuations of DVF were examinable.
Figure 4. This figure illustrates the lateral edges of DVF at the vagina. Detail window a.I shows the small nerve fibres (arrows) that run between the fan-shaped edges of DVF, which is adherent to the mesorectal fascia (arrowhead). Note the extensive vascular plexuses in the posterior vaginal wall (VW). The arrow in detail window a.II shows that at high magnification DVF is as closely related to vagina as to the mesorectal fascia (arrowheads). Small nerve fibres (N) run in the anterior mesorectum. ME: Millers’ elastin, LAM: levator ani muscle, R: rectum, ACL: anococcygeal ligament, OLL: outer longitudinal layer of the rectal wall, ICL: inner circular layer of the rectal wall. Scale bars in window a: 7 mm, window a.I: 3 mm, window a.II: 300 mm.
Discussion Until now, there has been no consensus on the anatomy of DVF and its relationship to the anterior dissection plane in TME. We have studied DVF in whole mount microscopic sections of adult pelvic exenteration specimens
A.C. Kraima et al. / EJSO 41 (2015) 738e745
743
Figure 5. Windows a, d and g are successive inferior levels from the cervix uteri (C), upper vagina (UV) until the lower vagina (LV). The inferior hypogastric plexus (IHP) and neurovascular bundles (NVB) are located laterally to the mesorectum (M) and rectum (R). Note that tiny nerve fibres traverse between the cervix uteri, vagina and the rectum. Windows b, e and h are consecutive slides relative to window a, d, and g and show the densely packed mesenchymal cells that are located between the anterior mesorectum and the cervix uteri and vagina (collagen is depicted in blue). The future mesorectal fascia is demonstrated by the arrowheads in all windows showing azan stained sections. The anterior mesorectum is adherent to the cervix uteri and vagina (arrows in detail window c.I and f.I). This relation disappears caudally as the mesorectum shrinks (arrow in detail window i.I). The smooth muscle fibres are revealed in detail windows c.II, f.II and i.II and the arrows indicate the fibres that slightly invest the anterior part of the mesorectum. SMA: smooth muscle actin, LAM: levator ani muscle. Scale bars in windows aeh: 500 mm, details windows: 200 mm.
and confirm that the optimal anterior plane in TME is located anterior to DVF. Macroscopically, it is a single anatomical entity extending from the pouch of Douglas to the perineal body. DVF is adherent to the anterior mesorectal fascia and cannot be distinguished from it even at high microscopic magnification. Any dissection carried out posterior to DVF puts the oncologic outcome at risk as this may easily lead to incomplete resection, tumour perforation and consequent tumour involvement of the circumferential resection margin. Rectal surgeons should be aware that DVF is anteriorly fused with the prostatic fascia in males and the posterior vaginal wall in females, meaning that
dissection anterior to DVF carries the risk of damaging these organs. Dissection behind DVF has been advocated by some researchers from whom the studies of Lindsey et al.13,26 are most notable. Based on operative analysis and histological analysis of surgical specimens, they reported that DVF is more closely related to the prostate and, in case of nonanteriorly and non-circumferentially located tumours, dissection should take place in the mesorectal plane between the fascia propria recti and DVF. In their series, the presence of DVF on the anterior surface of the mesorectum was histologically confirmed in 12 of the 30
744
A.C. Kraima et al. / EJSO 41 (2015) 738e745
specimens, which would indicate that dissection took place in the optimal anterior plane leaving DVF on the posterior prostate. As the latter was not histologically confirmed, the lack of DVF in the other 18 specimens could have been the result of suboptimal removal of the anterior mesorectum. Two of the three anterior planes as they suggested (the close rectal and mesorectal planes),7 would expose patients to suboptimal oncologic surgery, specifically at the level of the low rectum. We agree that DVF has macroscopically no discernible layers, but contradict the presence of two distinct microscopic layers as DVF consisted of a multilayered condensations of collagen, smooth muscle fibres and little elastin at high magnification in all whole mount microscopic sections. Moreover, several researchers studied the lateral edges of DVF in relation to the autonomic nerves. The Y-shaped lateral extensions as described by Peschaud et al.27 could not be confirmed. Kinugasa et al.14 reported on a lateral division into two or three laminae near the posterolateral corner of the prostate merging into or ending at the neurovascular bundles. One lamina extended dorsolaterally and separated the mesorectum from the autonomic nerves. Our findings confirm that the lateral edges of DVF consist of multiple laminae, though the most posterior part of DVF was continuous with the mesorectal fascia. We observed a tiny space between the lateral edges of DVF and the mesorectal fascia in one female specimen, but in the midline these were fused. As such, a border formed by a posterior lamina of DVF between the mesorectum and autonomic nerves could not be confirmed and consequently, dissection posterior to DVF as a wish to preserve the autonomic nerves cannot be supported. Is preservation of the autonomic nerves feasible when dissecting in front of DVF? Heald et al.28 stated that if the principles of TME are realized by following the innermost surgically identifiable areolar plane in anterolateral direction, gentle backward pressure on the specimen would allow surgeons to enter the plane anterior to DVF. If extra care is paid at the lateral edges of DVF, damage of the autonomic nerves could be averted. Our microscopic findings fully support these surgical descriptions. The neurovascular bundles are always at risk during dissection of the anterolateral mesorectum, but (sharp) dissection directly on the mesorectal fascia should allow preservation of the autonomic nerves. As DVF is adherent to the mesorectal fascia, there is no need to shift planes. It must be emphasized that not all small nerve fibres can be preserved. Some small nerve fibres ran between the prostatic fascia/posterior vaginal wall and DVF. We observed a similar distribution of these small nerve fibres in all fetal specimens. The inferior hypogastric plexus and major efferent fibres were located outside the mesorectum and very little nerve fibres were situated between the rectum and prostate/vagina. A small series of patients, in whom dissection was performed in the extramesorectal plane as proposed by Lindsey et al.,7 showed that this resulted in a justifiable rate of postoperative erectile dysfunction.29
Furthermore, the embryological origin of DVF cannot be elucidated based on microscopic analysis of the fetal specimens. Neither DVF nor peri-rectal fasciae could be recognized in our series that would correspond to the adult anatomy. A fascia morphologically comprises a single- or multi-layered condensation of collagen and smooth muscle fibres. However, we identified an area of densely-packed collagen as the future DVF, which was topologically similar to the adult anatomy of DVF. This would indicate that DVF might be formed in early fetal development by local condensation of mesenchymal cells, which is in agreement with the study of Aigner et al.30 Additionally, the absence of evident fasciae could suggest that mechanical stress on and stretch of the expanding volumes of the mesorectum, rectum and internal genital organs is also of relevance in the formation of DVF and peri-rectal fasciae, as has been proposed recently.31 In conclusion, DVF is adherent to and continuous with the mesorectal fascia. The anterior plane in TME is consequently anterior to DVF. Rectal surgeons must realize that there is no need to change planes when dissecting the anterolateral mesorectum. Autonomic nerves can be preserved by (sharp) dissection directly on the mesorectal fascia, which leads to the plane anterior to DVF. The usage of robot-assisted dissection might be valuable in this delicate anatomical region. Acknowledgements We would like to thank the GIFT donors who made this study possible and without whom no progress could be made in this area. Martin Waterhouse, Dave Turner and Mike Hale are gratefully acknowledged for their help in scanning the slides and processing the images. Aidan Hindley is gratefully acknowledged for his help in retrieving and administrating the GIFT specimens. Adam Kolesnik from the University of Warsaw is gratefully acknowledged for providing us with four fetal specimens. This study is funded by Technology Foundation STW (grant number 10903). PQ is funded by Yorkshire Cancer Research and the Experimental Cancer Medicines Centre. NW is funded by the Pathological Society of Great Britain and Ireland, The Academy of Medical Sciences and the National Institute for Health Research. The Aperio scanners are supported by the MRC Bioinformatics Centre (grant number MR/L01629X/1). Conflict of interest We declare no conflict of interest. References 1. Heald RJ. The ‘Holy Plane’ of rectal surgery. J R Soc Med 1988;81(9): 503–8. 2. Quirke P, Durdey P, Dixon MF, Williams NS. Local recurrence of rectal adenocarcinoma due to inadequate surgical resection. Histopathological study of lateral tumour spread and surgical excision. Lancet 1986;2:996–9.
A.C. Kraima et al. / EJSO 41 (2015) 738e745 3. Meredith KL, Hoffe SE, Shibata D. The multidisciplinary management of rectal cancer. Surg Clin North Am 2009;89:177–215. 4. Lange MM, van de Velde CJ. Urinary and sexual dysfunction after rectal cancer treatment. Nat Rev Urol 2011;8:51–7. 5. Wallner C, Lange MM, Bonsing BA, et al. Causes of fecal and urinary incontinence after total mesorectal excision for rectal cancer based on cadaveric surgery: a study from the Cooperative Clinical Investigators of the Dutch total mesorectal excision trial. J Clin Oncol 2008;26: 4466–72. 6. Church JM, Raudkivi PJ, Hill GL. The surgical anatomy of the rectum e a review with particular relevance to the hazards of rectal mobilisation. Int J Colorectal Dis 1987;2(3):158–66. 7. Lindsey I, Guy RJ, Warren BF, Mortensen NJ. Anatomy of Denonvilliers’ fascia and pelvic nerves, impotence, and implications for the colorectal surgeon. Br J Surg 2000;87:1288–99. 8. Moszkowicz D, Alsaid B, Bessede T, et al. Where does pelvic nerve injury occur during rectal surgery for cancer? Colorectal Dis 2011; 13:1326–34. 9. Bertrand MM, Alsaid B, Droupy S, Benoit G, Prudhomme M. Optimal plane for nerve sparing total mesorectal excision, immunohistological study and 3D reconstruction: an embryological study. Colorectal Dis 2013;15:1521–8. 10. Kinugasa Y, Murakami G, Suzuki D, Sugihara K. Histological identification of fascial structures posterolateral to the rectum. Br J Surg 2007;94:620–6. 11. Golligher JC. Anterior resection. In: Golligher JC, editor. Operative surgery of the colon, rectum and anus. 3rd ed. London: Butterworhts; 1980;, p. 143–56. 12. Heald RJ, Moran BJ. Embryology and anatomy of the rectum. Semin Surg Oncol 1998;15:66–71. 13. Lindsey I, Warren B, Mortensen N. Optimal total mesorectal excision for rectal cancer is by dissection in front of Denonvilliers’ fascia (Br J Surg 2004; 91: 121e123). Br J Surg 2004;91:897. 14. Kinugasa Y, Murakami G, Uchimoto K, Takenaka A, Yajima T, Sugihara K. Operating behind Denonvilliers’ fascia for reliable preservation of urogenital autonomic nerves in total mesorectal excision: a histologic study using cadaveric specimens, including a surgical experiment using fresh cadaveric models. Dis Colon Rectum 2006;49:1024–32. 15. Zhang C, Ding ZH, Li GX, Yu J, Wang YN, Hu YF. Perirectal fascia and spaces: annular distribution pattern around the mesorectum. Dis Colon Rectum 2010;53:1315–22. 16. Nano M, Levi AC, Borghi F, et al. Observations on surgical anatomy for rectal cancer surgery. Hepatogastroenterology 1998;45:717–26. 17. Moriya Y, Sugihara K, Akasu T, Fujita S. Nerve-sparing surgery with lateral node dissection for advanced lower rectal cancer. Eur J Cancer 1995;31:1229–32.
745
18. Bissett IP, Hill GL. Extrafascial excision of the rectum for cancer: a technique for the avoidance of the complications of rectal mobilization. Semin Surg Oncol 2000;18:207–15. 19. Clausen N, Wolloscheck T, Konerding MA. How to optimize autonomic nerve preservation in total mesorectal excision: clinical topography and morphology of pelvic nerves and fasciae. World J Surg 2008;32:1768–75. 20. Quirke P, Steele R, Monson J, et al. Effect of the plane of surgery achieved on local recurrence in patients with operable rectal cancer: a prospective study using data from the MRC CR07 and NCIC-CTG CO16 randomised clinical trial. Lancet 2009;373:821–8. 21. Leonard D, Penninckx F, Fieuws S, et al. Factors predicting the quality of total mesorectal excision for rectal cancer. Ann Surg 2010;252: 982–8. 22. den Dulk M, Putter H, Collette L, et al. The abdominoperineal resection itself is associated with an adverse outcome: the European experience based on a pooled analysis of five European randomised clinical trials on rectal cancer. Eur J Cancer 2009;45:1175–83. 23. van Leersum N, Martijnse I, den Dulk M, et al. Differences in circumferential resection margin involvement after abdominoperineal excision and low anterior resection no longer significant. Ann Surg 2014;259:1150–5. 24. Holm T, Ljung A, Haggmark T, Jurell G, Lagergren J. Extended abdominoperineal resection with gluteus maximus flap reconstruction of the pelvic floor for rectal cancer. Br J Surg 2007;94:232–8. 25. Everett MM, Miller WA. The role of phosphotungstic and phosphomolybdic acids in connective tissue staining. Histochemical studies. Histochem J 1974;6:25–34. 26. Lindsey I, Warren BF, Mortensen NJ. Denonvilliers’ fascia lies anterior to the fascia propria and rectal dissection plane in total mesorectal excision. Dis Colon Rectum 2005;48:37–42. 27. Peschaud F, Moszkowicz D, Alsaid B, Bessede T, Penna C, Benoit G. Preservation of genital innervation in women during total mesorectal excision: which anterior plane? World J Surg 2011;36:201–7. 28. Heald RJ, Moran BJ, Brown G, Daniels IR. Optimal total mesorectal excision for rectal cancer is by dissection in front of Denonvilliers’ fascia. Br J Surg 2004;91:121–3. 29. Kneist W, Junginger T. Male urogenital function after confirmed nerve-sparing total mesorectal excision with dissection in front of Denonvilliers’ fascia. World J Surg 2007;31:1321–8. 30. Aigner F, Zbar AP, Ludwikowski B, Kreczy A, Kovacs P, Fritsch H. The rectogenital septum: morphology, function, and clinical relevance. Dis Colon Rectum 2004;47:131–40. 31. Kim JH, Kinugasa Y, Hwang SE, Murakami G, Rodriguez-Vazquez JF, Cho BH. Denonvilliers’ fascia revisited. Surg Radiol Anat 2015;37: 187–97.
ScienceDirect EJSO 41 (2015) 738e745
www.ejso.com
Whole mount microscopic sections reveal that Denonvilliers’ fascia is one entity and adherent to the mesorectal fascia; implications for the anterior plane in total mesorectal excision? A.C. Kraima a,b, N.P. West b, D. Treanor b, D.R. Magee b, H.J. Rutten c, P. Quirke b, M.C. DeRuiter a, C.J.H. van de Velde d,* a
Department of Anatomy and Embryology, Leiden University Medical Center, P.O. Box 9600, 2300 ZC Leiden, The Netherlands b Pathology and Tumour Biology, Leeds Institute of Cancer and Pathology, University of Leeds, St. James’s University Hospital, Beckett Street, Leeds LS9 7TF, United Kingdom c Department of Surgery, Catherina Hospital Eindhoven, P.O. Box 1350, 5602 ZA Eindhoven, The Netherlands d Department of Surgery, Leiden University Medical Center, P.O. Box 9600, 2300 RC Leiden, The Netherlands Accepted 25 March 2015 Available online 2 April 2015
Abstract Background: Excellent anatomical knowledge of the rectum and surrounding structures is essential for total mesorectal excision (TME). Denonviliers’ fascia (DVF) has been frequently studied, though the optimal anterior plane in TME is still disputed. The relationship of the lateral edges of DVF to the autonomic nerves and mesorectal fascia is unclear. We studied whole mout microscopic sections of en-bloc cadaveric pelvic exenteration and describe implications for TME. Methods: Four donated human adult cadaveric specimens (two males, two females) were obtained from the Leeds GIFT Research Tissue Programme. Paraffin-embedded mega blocks were produced and serially sectioned at 50 and 250 mm intervals. Sections were stained with haematoxylin & eosin, Masson’s trichrome and Millers’ elastin. Additionally, a series of eleven human fetal specimens (embryonic age of 9e20 weeks) were studied. Results: DVF consisted of multiple fascial condensations of collagen and smooth muscle fibres and was indistinguishable from the anterior mesorectal fascia and the prostatic fascia or posterior vaginal wall. The lateral edges of DVF appeared fan-shaped and the most posterior part was continuous with the mesorectal fascia. Fasciae were not identified in fetal specimens. Conclusion: DVF is adherent to and continuous with the mesorectal fascia. Optimal surgical dissection during TME should be carried out anterior to DVF to ensure radical removal, particularly for anterior tumours. Autonomic nerves are at risk, but can be preserved by closely following the mesorectal fascia along the anterolateral mesorectum. The lack of evident fasciae in fetal specimens suggested that these might be formed in later developmental stages. Ó 2015 Elsevier Ltd. All rights reserved.
Keywords: Total mesorectal excision; Denonvilliers’ fascia; Rectal cancer; Mesorectal fascia; Autonomic nerves
Introduction The introduction of total mesorectal excision (TME) radically improved the surgical treatment and outcome of * Corresponding author. Leiden University Medical Center, Department of Surgery, K6-R, P.O. Box 9600, 2300 RC Leiden, The Netherlands. Tel.: þ31 71 5262309; fax: þ31 71 5266750. E-mail address: [email protected] (C.J.H. van de Velde). http://dx.doi.org/10.1016/j.ejso.2015.03.224 0748-7983/Ó 2015 Elsevier Ltd. All rights reserved.
rectal cancer. The TME principle involves en-bloc removal of the diseased rectum and surrounding mesorectum within an intact mesorectal fascia. Dissection in the ‘holy plane’ along the mesorectal fascia is said to enable preservation of the autonomic nerves.1 Tumour involvement of the circumferential resection margin and incomplete mesorectal excision are the most important predictors for recurrent disease, emphasizing the importance of en-bloc removal of
A.C. Kraima et al. / EJSO 41 (2015) 738e745
an intact mesorectal package with no disruptions, tears and/ or perforations.2,3 The Dutch TME trial showed that surgical damage to the autonomic nerves was the major cause of post-operative anorectal and urogenital dysfunction.4,5 As a consequence, posterior, lateral and anterior dissection planes were defined in TME to warrant complete removal of the mesorectum as well as identification and preservation of the autonomic nerves.6e9 For such an important area, it is perhaps surprising that the precise relationships between the autonomic nerves and peri-rectal fasciae, which form the dissection planes for rectal surgeons, are still debated.10 Denonvilliers’ fascia (DVF) has been frequently studied and so far there is no consensus on its embryological origin and topological anatomy. Some have argued that DVF was more closely related to the prostate,6,7 whilst others believed that DVF was situated closer to the rectum11 or was even adherent with the anterior mesorectum.12 Due to these contrary descriptions, the anterior plane in TME is frequently disputed. Some advocate dissection posterior to DVF,13,14 some argue to ‘split the layers’ of DVF,15,16 whereas others are convinced that optimal TME should be performed anterior to DVF.12,17e19 Contradictory views on the anterior dissection plane in TME risk suboptimal surgery and consequent poorer oncologic and functional outcome. Tumour involvement of the circumferential resection margin is most frequently reported in case of anterior tumours, which are anatomically closely related to DVF.20 In addition, incomplete mesorectal excisions are often reported.21 Patients requiring an abdominoperineal excision (APE) of the rectum and anus suffer from a poorer oncologic outcome as low rectal tumours are technically more difficult to resect. Specifically in advanced tumours, APE has an increased risk of tumour involvement of the circumferential resection margin in comparison with anterior resection specimen.22,23 Excellent knowledge of the complex pelvic anatomy is a prerequisite to optimize the oncological and functional outcome of rectal cancer. However, to what extent can we enhance the outcomes in TME if the anterior plane is still questioned? We aimed to study DVF in whole mount microscopic sections of adult pelvic exenteration specimens and concentrated on the morphological and topological anatomy, the anterior plane in TME and the lateral edges in relation to the autonomic nerves and mesorectal fascia. Additionally, a developmental series of human female and male fetuses was analysed to gain better insight in the development of endopelvic fasciae. Methods Adult cadaveric specimens Four human adult pelvic exenteration specimens were obtained through the University of Leeds GIFT Tissue Research Programme (www.gift.leeds.ac.uk) from consented donor bodies belonging to two males and two females. Ethical
739
approval was granted by the Northern and Yorkshire Regional Ethical Committee, Jarrow, UK (unique reference number 11/H0903/6). There was no history of pelvic surgery or pelvic pathology at post mortem examination. The specimens were retrieved during a tissue donation autopsy performed at St. James’s University Hospital in Leeds in with the body in the prone jack-knife position according to the extralevator APE technique as described by Holm et al.24 The specimens were essentially pelvic exenterations including the anal canal and rectum up to the recto-sigmoid junction, anal sphincters, perineal body, mesorectum with an intact mesorectal fascia, levator ani muscle, obturator internus muscle, vagina or prostate and penile bulb and posterior bladder wall. After fixation in 8% formaldehyde solution for seven days, the specimens were transversely sectioned at one centimetre. The slices were photographed and dissected to fit in Super Mega Cassettes measuring 74.8 52.5 16.5 mm (CellPath; Powys; UK). All tissues were dehydrated in graded ethanol and embedded in paraffin mega blocks. In addition, a developmental series of seven human female fetal pelvic specimens (embryonic age of 10, 12, 14, 15, 16, 19, and 20 weeks) and four human male fetal pelvic specimens (embryonic age of 9, 10, 12 and 20 weeks) were studied from collections in the Department of Anatomy & Embryology, Leiden University Medical Centre and the University of Warsaw, Poland. All fetuses were obtained with informed consent after miscarriage or legal abortion and were free of congenital pelvic malformations. The fetal specimens were dehydrated in graded ethanol and xylene, and embedded in paraffin blocks. Histological staining The mega blocks of the pelvic exenteration specimens were transversally cut in serial 5 mm sections. In one male and one female specimen every 10th section was collected onto glass slides and stained with haematoxylin and eosin (H&E), creating a series with a cross-sectional interval of 50 mm. Additional sections were collected from each mega block and stained with Masson’s trichrome (MT) and Millers’ elastin (ME).25 In the other male and female specimen every 48th, 49th and 50th section were collected, creating three series with a cross-sectional interval of 250 mm, of which one series was stained with H&E and one series with MT. The remaining series was kept for additional stains with ME. The paraffin blocks containing the fetal pelvic specimens were serially cut in transverse sections of 8 and 10 mm and alternately stained with H&E and azan to reveal collagen. A series was stained using antibodies against alpha-smooth muscle actin (SMA; Sigma-Aldrich, A2547) and S-100 (S100; DAKO, Z-031101) to detect smooth muscle fibres and the peripheral neural network, respectively. The protocol used can be explored online at: www.caskanatomy.info/research.
740
A.C. Kraima et al. / EJSO 41 (2015) 738e745
Image acquisition All stained glass slides from the adult pelvic exenteration specimens were digitally scanned in Leeds with an Aperio XT slide scanner (Aperio, San Diego, California, USA) at 20 magnification, creating a resolution of 0.46 microns per pixel. The digital images were compressed with JPEG2000 quality 70 and viewed in Aperio ImageScope version 10.2.2.2319. Selected glass slides from the fetal pelvic specimens were photographed using an Olympus AX70-microscope with an Olympus D12 camera. Results Adult pelvic exenteration specimens In the male exenteration specimens, DVF was located between the seminal vesicles, prostate and mesorectum. It
extended from the top of the seminal vesicles and terminated in the perineal body. DVF covered the posterior surface of the seminal vesicles and consisted of a fascial condensation of collagen and smooth muscle fibres. Here, DVF was indistinguishable from the mesorectal fascia (Fig. 1a). At a more inferior level, DVF was situated posteriorly and adjacent to the prostate. Even at high microscopic magnification DVF could neither be distinguished from the prostatic fascia nor from the mesorectal fascia, indicating that they were fused. Laterally, the edges of DVF appeared fan-shaped and curved along the anterolateral angle of the mesorectum. The posterior parts of DVF were continuous with the mesorectal fascia. A small intervening space was detected between these fasciae filled with adipose tissue and several small nerve fibres (Fig. 1b). At successive inferior levels, multiple small nerves and ganglionated nerve plexuses were located in this intervening space. The prostatic fascia merged with the endopelvic
Figure 1. This figure shows the relationship of DVF to the seminal vesicles (SV) and prostate (P). The inferior hypogastric plexus (arrows in window a) is located laterally to the mesorectal fascia (arrowheads in window a). The arrows in detail window a.I show the lateral edges of DVF at superior level. The lowest arrow points out the part that fuses with the mesorectal fascia (arrowhead). The arrows in detail window a.II show that DVF is adherent to the mesorectal fascia (arrowheads). At inferior levels, nerves (N) are located in the intervening space (star in detail window b.I) which is located between the prostatic fascia (lower arrow) and mesorectal fascia (arrowheads). The prostatic fascia fuses with the endopelvic fascia (upper arrow). The rightmost arrowheads show that DVF and the mesorectal fascia are continuous. The arrow in window b.II shows that DVF cannot be distinguished separately from the mesorectal fascia (arrowheads). ME: Millers’ elastin, MT: Masson’s trichrome, O: obturator internus muscle, A: apex of prostate, R: rectum, M: mesorectum. Scale bars in windows a and b: 8 mm, windows a.I and a.II: 800 mm, window b.I: 2 mm and window b.II: 1 mm.
A.C. Kraima et al. / EJSO 41 (2015) 738e745
741
Figure 2. This figure illustrates the lateral edges of DVF at the prostatic base (P). Detail window a.I shows the multi-layered lateral edges. Between these layers, small nerve fibres can be detected (arrows). Note the close relation of DVF with the outer longitudinal layer of the rectal wall (OLL). The arrow in detail window a.II shows that at high magnification DVF is as closely related to the prostatic fascia as to the mesorectal fascia (arrowhead). ME: Millers’ elastin, LAM: levator ani muscle, R: rectum, M: mesorectum, ICL: inner circular layer of the rectal wall. Scale bars in window a: 8 mm, window a.I: 2 mm, window a.II: 300 mm.
fascia posterolaterally. At the apex of the prostate, the lateral continuation of DVF appeared less organized as multiple fascial condensations reached the pelvic side wall. Small nerve fibres, blood vessels and adipose tissue were still detectable between these fascial layers (Fig. 2). Due to the natural pelvic tapering, the mesorectum shrank in volume inferiorly. The anterior mesorectum was less voluminous relative to the posterolateral mesorectum, resulting in a closer relationship between the anterior rectal wall and DVF. The outer longitudinal layer of the rectal muscular wall approximated DVF at certain points. DVF narrowed at the level of the internal urethral orifice before
anchoring in the perineal body. At high magnification, the parts of DVF located posterior to the prostate consisted of a multi-layered condensation of collagen and smooth muscle fibres with a paucity of elastin. Small blood vessels and nerve fibres ran between these layers over the whole horizontal axis of DVF but were more abundant at inferior levels. In the female exenteration specimens, DVF was situated between the cervix uteri, vagina and rectum. It extended from the pouch of Douglas’ to its termination in the perineal body. The morphological anatomy was identical to that of the male specimens, consisting of a multi-layered
Figure 3. This figure shows the fan-shaped lateral edges at the level of the vagina (V). The arrows in detail window a.I show small nerve fibres lateral to DVF. The mesorectal fascia at the anterolateral mesorectum in this female specimen is separate from the lateral edge of DVF (arrowheads detail window a.I). Detail window a.II shows that DVF and the mesorectal fascia (arrowheads) are fused in the midline. In certain places, DVF is torn from the mesorectum (arrow) demonstrating the delicateness of the whole mount microscopic sections and the adherent DVF and anterior mesorectum. At a more inferior level, nerves (arrowheads in detail window b.I) are located in the intervening space (star in detail window b.I) which formed by the lateral extension of DVF, the mesorectal fascia and the endopelvic fascia. DVF is laterally continuous with the mesorectal fascia (arrowheads detail window b.I). DVF is adherent to the mesorectal fascia (arrowheads detail window b.II) and the posterior vaginal wall (VW). ME: Millers’ elastin, MT: Masson’s trichrome, B: bladder, R: rectum, M: mesorectum, U: urethra, O: obturator internus muscle. Scale bar in window a: 7 mm, window a.I: 3 mm, window b: 8 mm, window b.I:2 mm, windows a.II and b.II: 800 mm.
742
A.C. Kraima et al. / EJSO 41 (2015) 738e745
condensation of collagen and smooth muscle fibres. At the level of the cervix uteri and upper vagina, the lateral continuations of DVF appeared fan-shaped as well. In the midline, the posterior part of DVF was fused with the mesorectal fascia and in a lateral direction it curved along the mesorectal fascia. No nerve fibres were detected between these laterally curving edges of DVF and the mesorectal fascia (Fig. 3a). The inferior hypogastric plexus was identified laterally to the mesorectal fascia and ran along the curved mesorectum in an anterior direction. At a more inferior level, an identical intervening space described in the male specimens was formed in the female specimens by an angle between the lateral vagina, pelvic side wall and fused DVF and mesorectal fascia (Fig. 3b). The part of DVF located posteriorly to the vagina was comprised of more elastin, although most elastin was detected in the anterior vaginal wall. DVF was fused with the posterior vaginal wall over its whole vertical axis (Fig. 4). At high magnification, the inner and outer muscular layers of the vaginal wall were closely related and were invested by extensive vascular plexuses.
Fetal pelvic specimens In all fetal specimens, the mesorectum was recognized as a peripheral layer of densely packed mesenchymal cells surrounding the rectum (Fig. 5). There was a clear difference in density of the mesenchymal cells forming the mesorectum and the mesenchymal cells giving rise to the extraperitoneal tissues outside the mesorectum. This border represented the future mesorectal fascia. Collagen depositions were seen on visceral and parietal surfaces in the female foetuses aged 14 weeks and older, and the male fetus aged 20 weeks, but no evident fasciae were detected that would correspond to the adult morphological characteristics of fasciae. In both female and male specimens, the anterior mesorectum reached the vagina and prostate, respectively. At this site, an area rich in densely-packed collagen was detected that was continuous with the future anterolateral mesorectal fascia. We identified this area as the future DVF. At successive inferior levels, the attachment of the anterior mesorectum to the vagina and prostate narrowed until the levator ani muscle. The SMA stained sections revealed the smooth musculature surrounding the cervix uteri and vagina. From the pouch of Douglas until the perineal body, smooth muscle fibres invested to some extent the layer of densely packed mesenchymal cells. Autonomic nerves were located anterolaterally to the mesorectum at the level of the cervix uteri and upper vagina. We did not observe major nerve fibres belonging to inferior hypogastric plexus within the mesorectum. Only a minority of small nerve fibres ran between the rectum and vagina or prostate. In none of the fetal specimens, the relation between autonomic nerves and the lateral continuations of DVF were examinable.
Figure 4. This figure illustrates the lateral edges of DVF at the vagina. Detail window a.I shows the small nerve fibres (arrows) that run between the fan-shaped edges of DVF, which is adherent to the mesorectal fascia (arrowhead). Note the extensive vascular plexuses in the posterior vaginal wall (VW). The arrow in detail window a.II shows that at high magnification DVF is as closely related to vagina as to the mesorectal fascia (arrowheads). Small nerve fibres (N) run in the anterior mesorectum. ME: Millers’ elastin, LAM: levator ani muscle, R: rectum, ACL: anococcygeal ligament, OLL: outer longitudinal layer of the rectal wall, ICL: inner circular layer of the rectal wall. Scale bars in window a: 7 mm, window a.I: 3 mm, window a.II: 300 mm.
Discussion Until now, there has been no consensus on the anatomy of DVF and its relationship to the anterior dissection plane in TME. We have studied DVF in whole mount microscopic sections of adult pelvic exenteration specimens
A.C. Kraima et al. / EJSO 41 (2015) 738e745
743
Figure 5. Windows a, d and g are successive inferior levels from the cervix uteri (C), upper vagina (UV) until the lower vagina (LV). The inferior hypogastric plexus (IHP) and neurovascular bundles (NVB) are located laterally to the mesorectum (M) and rectum (R). Note that tiny nerve fibres traverse between the cervix uteri, vagina and the rectum. Windows b, e and h are consecutive slides relative to window a, d, and g and show the densely packed mesenchymal cells that are located between the anterior mesorectum and the cervix uteri and vagina (collagen is depicted in blue). The future mesorectal fascia is demonstrated by the arrowheads in all windows showing azan stained sections. The anterior mesorectum is adherent to the cervix uteri and vagina (arrows in detail window c.I and f.I). This relation disappears caudally as the mesorectum shrinks (arrow in detail window i.I). The smooth muscle fibres are revealed in detail windows c.II, f.II and i.II and the arrows indicate the fibres that slightly invest the anterior part of the mesorectum. SMA: smooth muscle actin, LAM: levator ani muscle. Scale bars in windows aeh: 500 mm, details windows: 200 mm.
and confirm that the optimal anterior plane in TME is located anterior to DVF. Macroscopically, it is a single anatomical entity extending from the pouch of Douglas to the perineal body. DVF is adherent to the anterior mesorectal fascia and cannot be distinguished from it even at high microscopic magnification. Any dissection carried out posterior to DVF puts the oncologic outcome at risk as this may easily lead to incomplete resection, tumour perforation and consequent tumour involvement of the circumferential resection margin. Rectal surgeons should be aware that DVF is anteriorly fused with the prostatic fascia in males and the posterior vaginal wall in females, meaning that
dissection anterior to DVF carries the risk of damaging these organs. Dissection behind DVF has been advocated by some researchers from whom the studies of Lindsey et al.13,26 are most notable. Based on operative analysis and histological analysis of surgical specimens, they reported that DVF is more closely related to the prostate and, in case of nonanteriorly and non-circumferentially located tumours, dissection should take place in the mesorectal plane between the fascia propria recti and DVF. In their series, the presence of DVF on the anterior surface of the mesorectum was histologically confirmed in 12 of the 30
744
A.C. Kraima et al. / EJSO 41 (2015) 738e745
specimens, which would indicate that dissection took place in the optimal anterior plane leaving DVF on the posterior prostate. As the latter was not histologically confirmed, the lack of DVF in the other 18 specimens could have been the result of suboptimal removal of the anterior mesorectum. Two of the three anterior planes as they suggested (the close rectal and mesorectal planes),7 would expose patients to suboptimal oncologic surgery, specifically at the level of the low rectum. We agree that DVF has macroscopically no discernible layers, but contradict the presence of two distinct microscopic layers as DVF consisted of a multilayered condensations of collagen, smooth muscle fibres and little elastin at high magnification in all whole mount microscopic sections. Moreover, several researchers studied the lateral edges of DVF in relation to the autonomic nerves. The Y-shaped lateral extensions as described by Peschaud et al.27 could not be confirmed. Kinugasa et al.14 reported on a lateral division into two or three laminae near the posterolateral corner of the prostate merging into or ending at the neurovascular bundles. One lamina extended dorsolaterally and separated the mesorectum from the autonomic nerves. Our findings confirm that the lateral edges of DVF consist of multiple laminae, though the most posterior part of DVF was continuous with the mesorectal fascia. We observed a tiny space between the lateral edges of DVF and the mesorectal fascia in one female specimen, but in the midline these were fused. As such, a border formed by a posterior lamina of DVF between the mesorectum and autonomic nerves could not be confirmed and consequently, dissection posterior to DVF as a wish to preserve the autonomic nerves cannot be supported. Is preservation of the autonomic nerves feasible when dissecting in front of DVF? Heald et al.28 stated that if the principles of TME are realized by following the innermost surgically identifiable areolar plane in anterolateral direction, gentle backward pressure on the specimen would allow surgeons to enter the plane anterior to DVF. If extra care is paid at the lateral edges of DVF, damage of the autonomic nerves could be averted. Our microscopic findings fully support these surgical descriptions. The neurovascular bundles are always at risk during dissection of the anterolateral mesorectum, but (sharp) dissection directly on the mesorectal fascia should allow preservation of the autonomic nerves. As DVF is adherent to the mesorectal fascia, there is no need to shift planes. It must be emphasized that not all small nerve fibres can be preserved. Some small nerve fibres ran between the prostatic fascia/posterior vaginal wall and DVF. We observed a similar distribution of these small nerve fibres in all fetal specimens. The inferior hypogastric plexus and major efferent fibres were located outside the mesorectum and very little nerve fibres were situated between the rectum and prostate/vagina. A small series of patients, in whom dissection was performed in the extramesorectal plane as proposed by Lindsey et al.,7 showed that this resulted in a justifiable rate of postoperative erectile dysfunction.29
Furthermore, the embryological origin of DVF cannot be elucidated based on microscopic analysis of the fetal specimens. Neither DVF nor peri-rectal fasciae could be recognized in our series that would correspond to the adult anatomy. A fascia morphologically comprises a single- or multi-layered condensation of collagen and smooth muscle fibres. However, we identified an area of densely-packed collagen as the future DVF, which was topologically similar to the adult anatomy of DVF. This would indicate that DVF might be formed in early fetal development by local condensation of mesenchymal cells, which is in agreement with the study of Aigner et al.30 Additionally, the absence of evident fasciae could suggest that mechanical stress on and stretch of the expanding volumes of the mesorectum, rectum and internal genital organs is also of relevance in the formation of DVF and peri-rectal fasciae, as has been proposed recently.31 In conclusion, DVF is adherent to and continuous with the mesorectal fascia. The anterior plane in TME is consequently anterior to DVF. Rectal surgeons must realize that there is no need to change planes when dissecting the anterolateral mesorectum. Autonomic nerves can be preserved by (sharp) dissection directly on the mesorectal fascia, which leads to the plane anterior to DVF. The usage of robot-assisted dissection might be valuable in this delicate anatomical region. Acknowledgements We would like to thank the GIFT donors who made this study possible and without whom no progress could be made in this area. Martin Waterhouse, Dave Turner and Mike Hale are gratefully acknowledged for their help in scanning the slides and processing the images. Aidan Hindley is gratefully acknowledged for his help in retrieving and administrating the GIFT specimens. Adam Kolesnik from the University of Warsaw is gratefully acknowledged for providing us with four fetal specimens. This study is funded by Technology Foundation STW (grant number 10903). PQ is funded by Yorkshire Cancer Research and the Experimental Cancer Medicines Centre. NW is funded by the Pathological Society of Great Britain and Ireland, The Academy of Medical Sciences and the National Institute for Health Research. The Aperio scanners are supported by the MRC Bioinformatics Centre (grant number MR/L01629X/1). Conflict of interest We declare no conflict of interest. References 1. Heald RJ. The ‘Holy Plane’ of rectal surgery. J R Soc Med 1988;81(9): 503–8. 2. Quirke P, Durdey P, Dixon MF, Williams NS. Local recurrence of rectal adenocarcinoma due to inadequate surgical resection. Histopathological study of lateral tumour spread and surgical excision. Lancet 1986;2:996–9.
A.C. Kraima et al. / EJSO 41 (2015) 738e745 3. Meredith KL, Hoffe SE, Shibata D. The multidisciplinary management of rectal cancer. Surg Clin North Am 2009;89:177–215. 4. Lange MM, van de Velde CJ. Urinary and sexual dysfunction after rectal cancer treatment. Nat Rev Urol 2011;8:51–7. 5. Wallner C, Lange MM, Bonsing BA, et al. Causes of fecal and urinary incontinence after total mesorectal excision for rectal cancer based on cadaveric surgery: a study from the Cooperative Clinical Investigators of the Dutch total mesorectal excision trial. J Clin Oncol 2008;26: 4466–72. 6. Church JM, Raudkivi PJ, Hill GL. The surgical anatomy of the rectum e a review with particular relevance to the hazards of rectal mobilisation. Int J Colorectal Dis 1987;2(3):158–66. 7. Lindsey I, Guy RJ, Warren BF, Mortensen NJ. Anatomy of Denonvilliers’ fascia and pelvic nerves, impotence, and implications for the colorectal surgeon. Br J Surg 2000;87:1288–99. 8. Moszkowicz D, Alsaid B, Bessede T, et al. Where does pelvic nerve injury occur during rectal surgery for cancer? Colorectal Dis 2011; 13:1326–34. 9. Bertrand MM, Alsaid B, Droupy S, Benoit G, Prudhomme M. Optimal plane for nerve sparing total mesorectal excision, immunohistological study and 3D reconstruction: an embryological study. Colorectal Dis 2013;15:1521–8. 10. Kinugasa Y, Murakami G, Suzuki D, Sugihara K. Histological identification of fascial structures posterolateral to the rectum. Br J Surg 2007;94:620–6. 11. Golligher JC. Anterior resection. In: Golligher JC, editor. Operative surgery of the colon, rectum and anus. 3rd ed. London: Butterworhts; 1980;, p. 143–56. 12. Heald RJ, Moran BJ. Embryology and anatomy of the rectum. Semin Surg Oncol 1998;15:66–71. 13. Lindsey I, Warren B, Mortensen N. Optimal total mesorectal excision for rectal cancer is by dissection in front of Denonvilliers’ fascia (Br J Surg 2004; 91: 121e123). Br J Surg 2004;91:897. 14. Kinugasa Y, Murakami G, Uchimoto K, Takenaka A, Yajima T, Sugihara K. Operating behind Denonvilliers’ fascia for reliable preservation of urogenital autonomic nerves in total mesorectal excision: a histologic study using cadaveric specimens, including a surgical experiment using fresh cadaveric models. Dis Colon Rectum 2006;49:1024–32. 15. Zhang C, Ding ZH, Li GX, Yu J, Wang YN, Hu YF. Perirectal fascia and spaces: annular distribution pattern around the mesorectum. Dis Colon Rectum 2010;53:1315–22. 16. Nano M, Levi AC, Borghi F, et al. Observations on surgical anatomy for rectal cancer surgery. Hepatogastroenterology 1998;45:717–26. 17. Moriya Y, Sugihara K, Akasu T, Fujita S. Nerve-sparing surgery with lateral node dissection for advanced lower rectal cancer. Eur J Cancer 1995;31:1229–32.
745
18. Bissett IP, Hill GL. Extrafascial excision of the rectum for cancer: a technique for the avoidance of the complications of rectal mobilization. Semin Surg Oncol 2000;18:207–15. 19. Clausen N, Wolloscheck T, Konerding MA. How to optimize autonomic nerve preservation in total mesorectal excision: clinical topography and morphology of pelvic nerves and fasciae. World J Surg 2008;32:1768–75. 20. Quirke P, Steele R, Monson J, et al. Effect of the plane of surgery achieved on local recurrence in patients with operable rectal cancer: a prospective study using data from the MRC CR07 and NCIC-CTG CO16 randomised clinical trial. Lancet 2009;373:821–8. 21. Leonard D, Penninckx F, Fieuws S, et al. Factors predicting the quality of total mesorectal excision for rectal cancer. Ann Surg 2010;252: 982–8. 22. den Dulk M, Putter H, Collette L, et al. The abdominoperineal resection itself is associated with an adverse outcome: the European experience based on a pooled analysis of five European randomised clinical trials on rectal cancer. Eur J Cancer 2009;45:1175–83. 23. van Leersum N, Martijnse I, den Dulk M, et al. Differences in circumferential resection margin involvement after abdominoperineal excision and low anterior resection no longer significant. Ann Surg 2014;259:1150–5. 24. Holm T, Ljung A, Haggmark T, Jurell G, Lagergren J. Extended abdominoperineal resection with gluteus maximus flap reconstruction of the pelvic floor for rectal cancer. Br J Surg 2007;94:232–8. 25. Everett MM, Miller WA. The role of phosphotungstic and phosphomolybdic acids in connective tissue staining. Histochemical studies. Histochem J 1974;6:25–34. 26. Lindsey I, Warren BF, Mortensen NJ. Denonvilliers’ fascia lies anterior to the fascia propria and rectal dissection plane in total mesorectal excision. Dis Colon Rectum 2005;48:37–42. 27. Peschaud F, Moszkowicz D, Alsaid B, Bessede T, Penna C, Benoit G. Preservation of genital innervation in women during total mesorectal excision: which anterior plane? World J Surg 2011;36:201–7. 28. Heald RJ, Moran BJ, Brown G, Daniels IR. Optimal total mesorectal excision for rectal cancer is by dissection in front of Denonvilliers’ fascia. Br J Surg 2004;91:121–3. 29. Kneist W, Junginger T. Male urogenital function after confirmed nerve-sparing total mesorectal excision with dissection in front of Denonvilliers’ fascia. World J Surg 2007;31:1321–8. 30. Aigner F, Zbar AP, Ludwikowski B, Kreczy A, Kovacs P, Fritsch H. The rectogenital septum: morphology, function, and clinical relevance. Dis Colon Rectum 2004;47:131–40. 31. Kim JH, Kinugasa Y, Hwang SE, Murakami G, Rodriguez-Vazquez JF, Cho BH. Denonvilliers’ fascia revisited. Surg Radiol Anat 2015;37: 187–97.
![[Transanal total mesorectal excision].](/assets/img/hold.png)
