Technique and Results of Transanal Endoscopic Microsurgery in Early Rectal Cancer Gerhard Buess,


Burkhard Mentges, MD, Klaus Manncke, MD, Michael Starlinger, Horst-Dieter Becker, MD, Tiibingen, Germany

The anatomy of the pelvis makes it difficult to perform local excisions in the rectum when the tumor is some distance from the anal verge. We have, therefore, developed a new minimally invasive technique for tumor resection. A rectoscope with a 4Qmm diameter permits tumor resection under stereoscopic control in the gas-dilated recta1 cavity. Excisions in full-thickness technique up to segmental resections with end-to-end anastomosis can be pf9fOrIkMXL



here is a growing consensus in the literature that, in selected cases of rectal cancer, local treatment is indicated [I]. However, surgeons disagree about the ease of exposure of the operative field with the different techniques. The arguments for local excision are that early rectal cancers with good or moderate differentiation have a low rate of metastatic spread and that conventional rectal resection by the abdominal approach carries a considerable risk regarding operative morbidity and mortality [2]. Local excision of small rectal cancers should be performed by the technique of full-thickness excision (disk excision), with a margin of clearance of at least 1 cm. The poor exposure and the limited distance from the anal verge that can be achieved using retractors in transanal surgery and the invasiveness and complication rates of the Mason and Kraske procedures [j-6] set the stage for the development of a new surgical technique combining the good exposure of endoscopy with advanced instrument technology. With the assistance of the Wolf Company (Knittlingen, Germany), we started to develop this technique in 1980, and it was ready for clinical application in 1983.

In selected cases, local excision of a small recta1 cancer can be regarded as appropriate treatment. However, most local resections of carcinomas are performed when removal of an adenoma is planned, and the postoperative histology shows a carcinoma. Since 1983, we have operated on 326 patients, 274 who have heen enrolled in a prospective clinical trial. Definitive histologic examination proved that 74 of these tumors were carcinomas. The rate of severe complications in patients with carcinomas was 9%, and the mortality rate was 0%. The advantages of this new technique are: The stereoscopic magnified view in the gas-dilated rectum allows precise surgery in an operative field that is otherwise difficult to reach. During the postoperative period, minimal discomfort and pain result in a short hospitalization.

METHODS The technique of transanal endoscopic microsurgery (TEM) [7-4: As with other techniques of minimally invasive sur-

From the Department of General Surgery, University of Tfibingen, Tiibingen, Germany. Requests for reprints should be addressed to Gerhard Buess, MD, Minimal Invasive Surgery, Klii Schnarrenberg, D-7400 Ttibingen, Germany. F’resentedat the 32nd Annual Meeting of the Society for Surgery of the Alimentary Tract, New Orieans, Louisiana, May 20-22.1991.

gery, the view to the operative field is gained by a rigid telescope (Figure 1). The operative field is extended by insufflation with carbon dioxide, and surgery is performed with elaborate surgical instruments. The recroscope. The diameter of the scope is 40 mm. Two different lengths, 12 and 20 cm, are available. The rectoscope is introduced into the rectum using an obturator. A glass window is then inserted at the rear end, and the tumor is visualized using manual air insufflation as in conventional rectoscopy. After the optimal position is defined, the rectoscope is fixed using the Martin arm, a double ball joint support instrument. The glass window is then removed, and the working insert is attached to the m&scope. Sealing elements prevent gas loss when the instruments are introduced (Figure 2). Surgical instruments.AU instruments are usually de signed specifically for endoscopic work. Scissors and forceps are angled to the right or to the left. The forceps, needle holder, and suction device are bent to allow easy access to the sacral cavity. A clip applicator is used to press a silver clip on the thread as a substitute for the surgical knot. The thread is monofilament and resorbable. Stereoscopictelescope.This telescope is angled at the tip and at the eyepiece so that the surgical working field




Flgwe 1. Operatingsystem for transanal endoscopic mlcrosurgery (TEM). Rectal dilation by gas insufflation.

lies in front and beneath the tip of the optic, The surgeon always has a direct stereoscopic view with up to a sixfold magnification. If a video camera is available, the operation can be followed on the video monitor by the whole team. Gtherwise, a single-view, flexible teaching optic can be used. Water can be injected automatically to clean the optic, which is important in case of spurting hemorrhage. Combined endosurgiculunit. Constant optimal exposure of the operative field is provided by automatic pressure-controlled gas insufflation. Carbon dioxide gas is used at a constant pressure of about 15 mm Hg. A water rinse for the optic and defined suction using a roller pump are integrated. Indications for local excision: Local excision Of Sessile adenomas is routinely performed in all tumors that can be easily reached with a rigid rectoscope. The extent of the adenomas can reach up to circular and to a length of 8 cm. For local excision of carcinomas, four subgroups of patients can be defined: (1) Patients with a good chance of curative treatment. Such patients have pT1 carcinomas with good or moderate differentiation (low-risk carcinomas according to Hermanek and Gall [ 101 and Morson [11]). Patients are selected after clinical staging, endoluminal ultrasound, and biopsy. Indication is accepted for all age groups. (2) Patients with a limited chance of curative treatment. Such patients have pT2 carcinomas with good or moderate differentiation (low-risk carcinomas according to Hermanek and Gall [IO]). Indication is accepted in elderly and high-risk patients. (3) Patients who present for palliative treatment. Such patients have carcinomas up to 4 cm in diameter if they are mobile (T3). Indication is only accepted in patients who are a high operative risk. (4) Patients with incidental carcinoma in adenoma. Most of our locally excised carcinomas belong to this subgroup. We perform the local excision of sessile villous adenomas in the extraperitoneal part of the rectum routinely by the technique of full-thickness excision. In 20% of cases, postoperative histology shows a 64


Flgere 2. Top. Assembly of ths complete operating system. Bottom. Front part of the rectoscope. !Stereoscopic optic at the top. Needle holder and forceps are also visible.

carcinoma in the center of the adenoma. The majority of these carcinomas are pT1. Preoperative examination: Apart from a standard clinical work-up (total coloscopy and clinical staging), the surgeon performs a rigid rectoscopy and the endoluminal ultrasound examination. Rigid rectoscopy is essential to localize the tumor. The distance of the tumor from the anal verge and the circumferential localization has to be determined. The positioning of the patient for the operation depends on the location of the tumor. Tumors



at the posterior wall are operated on with the patient in the lithotomy position; tumors at the anterior wall are operated on with the patient in a prone position; and tumors on the side walls are operated on with the patient in a side position. We perform the endoluminal ultrasound examination using an instrument that was developed in conjunction with the Kretz Company (Zipf, Austria) [12,13]. The ultrasound probe is attached to a rigid scope, and the entire rectum is filled with water. Examinations are carried out with an integrated 5-,8-, and lo-MHz scanner. In contrast to examinations using a water-filled balloon, which presses the tumor into the surrounding tissue THE AMERICAN

and obscures the different layers of the rectal wall, our technique preserves the anatomy, especially in soft tumors, so that the depth of infiltration can be better visualized. Since most of our patients with local excision of early cancers are classified in Group 4, we had to adapt the clinical staging [5] to also include the adenomas. We classify typical adenomas with their soft consistency as CS 0, irrespective of their mobility against the bowel wall. Preoperative preparation of patients and anesthesia: Bowel preparation with whole gut lavage and perio-

perative single-shot antibiotic prophylaxis is mandatory.








of the bowel wall by suture,.Theneedls Iholder mm

Tumor stage k74 carcinomas

F -----I





PT2 n=17



PT3 n=6



n=2/ \




& k22

radical secondary procedure FQure 6. Tumor stages of 74 carcinomas. White field: only local procedures. @ey flekt radical reopedlon because of tumor stage.

The procedure requires that the patient have either general or spinal anesthesia. During the last 2 years, we have increased the rate of regional anesthesia from 6% to 30%. Operative steps 111: TEM starts with rectoscopy using a glass window and manual insufflation. After lo66


calization of the tumor, the rectoscope is attached to the Martin arm, which provides stable positioning during the procedure. Now the entire equipment is assembled. In the case of adenomas, the margin of clearance is 5 mm; in carcinomas, it is 1 cm. The margin of clearance is defined by coagulation dots. Only adenomas at the anterior wall located higher than 12 cm are excised by the mucosectomy technique (Figure 3). Tumors located in the extraperitoneal part of the rectum, that is, up to 20 cm at the posterior wall, 15 cm at the side wall, and 12 cm at the anterior wall, are excised by using the full-thickness technique. TEM is not indicated for patients with carcinomas located on the anterior wall if they are higher than 12 cm from the anal verge. In all carcinomas, full-thickness excision is mandatory, and this cannot be performed safely at this location because of gas loss into the opened peritoneal cavity. The bowel wall is first transected aborally by means of a monopolar knife. During the preparation at the base of the tumor, the tissue is transected close to the rectal wall if the tumor is located anteriorly or laterally to avoid injury to the vagina, the urethra, or the opening of the peritoneum. In the case of posteriorly located tumors, a part of the retrorectal fat can be resected (Figure 4). Bleeding occurs frequently during full-thickness operations. Blood vessels are localized using the suction probe and coagulated with the tip. Spurting arteries are compressed with the forceps and coagulated (Figure 4). The resulting defect is closed by transverse continuous suture using monofil PDS thread 3.0 (Figure 5). Instead of a knot, a silver clip is pressed onto the thread. PATIENTS AND RESULTS Since 1983, 326 patients have undergone TEM [7]. Of these 326 patients, 214 have undergon TEM between 1986 and February 1991 and have been enrolled in a prospective clinical trial. The results of this subgroup



FQwm 7. Small cancer excised in full-


Carcinomas Excision technique full-thickness





intraperitoneal full-th.-ext.

33 5


full-thickness excision + perirectal fat 18 segment


ofihe local excision con&Wed by excision



partial wall excision





have been analyzed for this paper. Definitive histologic examination proved that 74 of these tumors were carcinomas. A large number of the carcinomas were diagnosed only postoperatively. The age distribution is typical for patients with rectal cancers. The predominant location was the middle third of the rectum (n = 26), 8 to 12 cm from the anal verge. More than two thirds of the tumors were pT1 tumors (F&ure 6). The operative time dependson the extent of operation. Median time for mucosectomy was 48 minutes and for segmental resection 120 minutes. In the group of patients who underwent cancer operations, no intraoperative complication occurred. The postoperative complication rate was 16% when minor compli-

cations were included; the rate of severe complications was 9%. Severe complications have occurred mainly after extended excisions with some tension to the suture line. In two patients, we found a dehiscense of the suture line, and, in another two patients, a rectovaginal fistula. In two of these patients, we had to perform a temporary colostomy. In one patient with T2 tumor, we had to perform an extirpation because of the tumor’s stage. The excision technique for carcinomas included full-thickness excision in 33 patients (Figures 7 and 8), full-thickness excision with resection of retrorectal (Figure 9) fat in 18 patients, and mucosectomy in 4 patients. The rate of complications after adenoma resection has been lower because of less extensive excisions.




Mortality in the adenoma group was 0.5%. One patient died on the sixth postoperative day after developing a pulmonary embolism. There were no deaths after the local excision of cancer. Follow-up of patients after local excision of carcinomas: After exclusion of all patients who then under-

went radical resection because of high-risk tumors as defined by Hermanek and Gall [IO], because of pT2 and pT3 stages, and because of reoperations of patients with pT1 low-risk tumors during the initial phase of this series (Figure 6), 29 patients with pT1 low-risk tumors remained for follow-up. The follow-up rate of the patients operated on till June 1989 is 93%. There was one local recurrence in the group of patients with pT1 low-risk tumors. The restaging of this tumor after diagnosis of recurrence showed that resection has not been complete. In the case of incomplete excision of a cancer, we routinely perform a radical moperation. COMMENTS

Following the pathologic studies on early rectal cancers presented by Morson [II] and Hermanek and Gall [IO], local excision of selected cancers is well accepted. Our contribution to local treatment for rectal cancer is the adaption of minimally invasive techniques to this important aspect of surgery. We developed the technique of TEM between 1980 and 1983 with its fust clinical application in 1983 [fl. Today, minimally invasive surgery is an important part of our clinical work with expanding indications (esophageal resection, cholecystectomy, appendectomy). The argument for starting our developmental work with TEM was that the rectum is the least accessible region in general surgery and that most locally treatable tumors are located here. Tumors in the lower rectum can be reached using equipment such as the Parks or the 68



F@ure 9. pT1 carcinoma with retm-


rectal fat and turn--free lymph node.

Mayo retractors, but mechanical dilation often obstructs visualization, which results in a less precise preparation. As a consequence of this, high rates of local recurrence after transanal removal of sessile adenomas have been observed [15]. The recurrence rate in patients with adenomas in our series was 4%. Tumors in the upper rectum cannot be reached using retractors, and disk excision is even more difficult than mucosectomy. Therefore, invasive surgical procedures such as Mason [3-51 or Kraske [q are required. Since the complication rates of these procedures are high [16-181, deep anterior resection is currently favored in all tumors of the upper part of the rectum. In contrast to the conventional techniques for local excision of rectal tumors, TEM gives optimal visual control when working in the gas-dilated rectum. The magnification and stereoscopic view add to the technical benefits of endoscopic microsurgery. This makes local rectal surgery both less invasive and more precise. Our complication rates are lower than after Kraske, Mason, or low anterior resections, especially after reset tion of adenomas. In more than 300 operations, there were no deaths after a local complication. The major complications are breakdown of the suture line, resulting in tension to the suture line, and vaginal fstula following extended excisions after dissection of the perirectal fat and coagulation of bleeders from the vaginal wall. After we became cognizant of this problem, we began to perform resections at the anterior wall in females more carefully, paying close attention to the rectovaginal fascia. Our clinical results support the arguments for local excision: surgery can be performed very precisely, especially with regards to maintaining the margin of clearance [19]. Our radical reoperations in the first years of our experience support the fact that the rate of metastases is low in favorable cases. We found no lymph node metastases in 10 patients who underwent reoperations of pT1



low-risk cancers. The low recurrence rate after resection of pT1 low-risk cancer suggests that clinical application should be broadened. The results of local resection of pT2 tumors with one recurrence in 15 patients are reasonable at the moment. The high rate of patients who refused radical reoperation made this experience possible. Compared with the more extended conventional surgical techniques, the advantages of TEM are as follow: painless postoperative course, unrestricted mobility, short hospital stay (patients can be discharged on the fourth or fifth postoperative day), and markedly reduced rehabilitation time. The problems of TEM are the same as with other minimally invasive procedures: The advanced technology is expensive. Special training in the technique is mandatory [20,21]. Knowledge in other microsurgical or endoscopic techniques and special talent are advantageous. Basic training courses are held at Washington University by Lee Smith. Video-supported intensive training courses using stage-orientated phantom models are performed in Tiibingen. The courses in Ttibingen are 5-day intensive training courses. A portion of these courses are in English. REFERENCES 1. Graham RA, Garnsey L, Milbum Jessup J. Local excision of rectal carcinoma. Am J Surg 1990; 160: 306-12. 2. Gall FP, Hermanek P. Die erweiterte Lymphknotendissektion beii Magen- und colon&t&n Karzinom. Chirurg 1988; 59: 202-10. 3. Mason AY. Surgical access to the rectum-a trans-sphincteric exposure. Proc R Sot Med 1970; 63: 91-4. 4. Mason AY. Transsphincteric surgery of the rectum. Prog Surg 1974; 13: 66-97. 5. Mason AY. Trans-sphincteric surgery for lower rectal cancer. In: Reifferscheid M, I-anger S, editors. Der Mastdarmkrebs. Stuttgart: Thieme Verlag, 1980. 6. Kraske P. Zur Exstirpation hochsitzender Mastdarmkrebse. Verh Dtsch Ges Chir 1885; 14: 464-74. 7. Buess G, Hutterer F, Theiss R, Boebel M, Isselhard W, Pichlmaier H. Das System fur die transanale endoskopische Rektumoperation. Chir 1984; 55: 677-80. 8. Buess G. Transanale endoskopische mikrochirurgie. In: Buess G, editor. Endoskopie. Von der Diagnostik bis zur neuen Chirurgie. K61n: Deutscher Arzte-Verlag, 1990; 288-3 11. 9. Buess G, Kipfmiiller K, Hack D, Gril/3ner R, Heintz A, Junginger T. Technique of transanal endoscopic microsurgery. Surg Endosc 1988; 2: 71-5. 10. Hermanek P, Gall FP. Early (microinvasive) colorectal carcinoma. Int J Colorectal Dis 1986; 1: 79-84. 1 I. Morson BC. Factors influencing the prognosis of early cancer of the rectum. Proc R Sot Med 1966; 59: 607-8. 12. Buess G, Heintz A, Frank K, Strunk H, Kuntz Ch. Endoluminale sonographie des rektums. In: Buess G, editor. Endoskopie. Von der Diagnostik bis zur neuen Chirurgie. Kiiln: Deutscher ArzteVerlag, 1990: 76-82. 13. Buess G. Endoskopie. Von der Diagnostik bis zur neuen Chirurgie. Kiiln: Deutscher Arzte-Verlag, 1990. 14. Cuschieri A, Buess G, Perissat J. Operative manual of endoscopic surgery. Berlin: Springer, 1992. 15. Schiessel R, Wunderlich M, Karner-Hanusch J. Transanale excision und anastomosentechnik. Chirurg 1986; 57: 773-8. 16. Schildbcrg FW, Wenk H. Der posteriore Zugang zum rectum.

Chirurg 1986; 57: 779-91. 17. Allgower M, Diirig M, Hochstetter A, von Huber A. The parasacral sphincter-splitting approach to the rectum. World J Surg 1982; 6: 539-48. 18. Hiiring R, Karavias Th, Konradt J. Die posteriore Proktorektotomie. Chirurg 1978; 49: 265-71. 19. Buess G, Kipfmuller K, Ibald R, et al. Clinical results of transanal endoscopic microsurgery. Surg Endosc 1988; 2: 245-50. 20. Buess G, Naruhn M, Motzung Th, Mentges B, Becker HD. Trainingsprogramm fur die Minimal-invasive Chirurgie. Chirurg 1991; 62: 276-83. 21. Kipfmiiller K, Buess G, Naruhn M, Junginger T. Training program for transanal endoscopic microsurgery. Surg Endosc 1988; 2: 24-7.


What is the diameter of your resectoscope? How far up from the anal verge in the rectum were the lesions in your series? In other words, how high do you go in the rectum with your scope? Your data give the appearance as if you had performed segmental resections through this scope. Greg V. Stiegmann (Denver, CO): Would you elaborate on the ultrasound staging that you use preoperatively, i.e., what factors do you consider in making a decision to do the minimal access operation? Is it purely ultrasound? Is it physical examination? Is it computed tomography? How do you evaluate your patients and what are your criteria? What do you do if you find an enlarged lymph node with the ultrasound? Thomas R. Gadacz (Baltimore, MD): Do you have any problem with postoperative pneumoperitoneum as related to the location of the resection? In a circumferential resection, do you do a running sutured, an interrupted sutured, or a stapled anastomosis? What’s your approach when you completely transect the rectum? Do you drain the presacral area? Dr. Cameron: Have you performed endoscopy in your patients at 5 days or 1 week to see whether these closures have stayed intact following your adenoma and carcinoma resections? Gerhard Buess (closing): Dr. Cameron, 40 mm. It depends on the indication. Technically, we can operate as far as the rigid endoscope can reach, i.e., up to 20 to 25 cm. When we find a pT1 carcinoma on the anterior wall of the rectum, which is higher than 11 or 12 cm, we don’t perform a local excision because we cannot do a fullthickness excision in this area. Full-thickness excision must be restricted to the extraperitoneal part of the rectum, which is up to 20 cm on the posterior wall but only 10 to 11 cm on the anterior wall. These were not proven carcinomas before operation. If we have an extensive carcinoma, we do not perform a local excision. Local excision in our series is only for pT1 carcinomas with good or moderate differentiation. Dr. Stiegmann, with computed tomography, we cannot assess the infiltration of small tumors. Therefore, we perform endoluminal ultrasound. When the endoluminal ultrasound shows that the muscle is infiltrated, then we usually do not perform a local excision. In those cases, we





perform local procedures only in very elderly or high-risk patients. In the small tumors, I have never seen markedly enlarged lymph nodes. Dr. Gadacz, we are using a running sutured anastomosis but interrupting with silver clips. Yes, we have had pneumo



peritoneum with the proximal lesions. There was no prob lem because the pneumoperitoneum rapidly resolved. Dr. Cameron, we do perform endoscopy after surgery. We find that approximately one third of the patients have a small dehiscence of the suture line, but this is without clinical relevance.


Technique and results of transanal endoscopic microsurgery in early rectal cancer.

The anatomy of the pelvis makes it difficult to perform local excisions in the rectum when the tumor is some distance from the anal verge. We have, th...
9MB Sizes 0 Downloads 0 Views