Transanal endoscopic microsurgery

Transanal Endoscopic Microsurgery Experience with 75 Rectal Neoplasms Dina Lev-Chelouche, M.D., David Margel, M.D., Gideon Goldman, M.D., Micha J. Rabau, M.D. From the Department of Surge~ B&C, Tel Aviv Sourasky Medical Center, and Sackler t~culty of Medicine, Tel Aviv University, Tel Aviv, Israel PURPOSE: The aim of this study was to describe a single institution's experience with transanal endoscopic microsurgery in patients with benign and malignant rectal tumors. PATIENTS: Between January 1992 and April 1998, 75 patients with a mean follow up of 38 months, underwent transanal endoscopic microsurgery excision of benign (46) or malignant (29) rectal tumors, located 3 to 18 cm from the dentate line. RESULTS:A total of 3 of 46 (6.5 percent) patients with benign tumors underwent conversion to radical surgery owing to tumor size. During the follow-up period, benign tumor recm'rence was observed in four (9 percent) patients, three of whom were managed by repeat transanal endoscopic microsurgery, whereas one required radical surgery. Histologlc staghag of malignant tumors was T1 (10), T2 (10), and T3 (9). Seven patients with either inadequate resection margins or T3 tumors were complimented with radical surgery. Of the remaining 22 patients, 11 received adjuvant radiation therapy whereas 11 had no Mrther treatment. Four (18 percent) had recurrent disease, which was managed by repeat transanal endoscopic microslugery in two, radical surgery in one, and laser ablation in one. No cancer-related deaths were observed during the foRow-up period. There was one operative mortality in a cardiac-crippled patient. Postoperative complications were mainly of a minor character and included fever, urinary retention, mad bleeding; none of which required reintervention. Rectourethral fisttfla developed in one patient who underwent repeat transanal endoscopic microsurgery excision for a T3 malignancy. Fecal soiling was transient in three patients and persisted in two. CONCLUSION: Transanat endoscopic microsurgery excision is a safe and precise technique that is well tolerated even in high operative risk patients. Transanal endoscopic microsurgery may become a procedure of choice for benign rectal tumors and selected early malignant neoplasms. [Key words: Rectal tumors; Early rectal cancer; Local excision; Transanal endoscopic microsurgery (qTM)]

procedures, including the transanal, transsphincteric, or posterior approach methods, have b e e n r e c o m m e n d e d as a definite procedure for benign adenomas and selected early rectal cancers. ~-7 This concept is appealing, because sphincter function is preserved and a permanent colostomy is avoided. It is especially attractive in regard to the operative m a n a g e m e n t of elderly and frail patients, with lower morbidity and mortality rates. Moreover, local surgical treatment is curative for benign lesions and offers a more than 60 percent local cure rate for malignant tumors w h e n no lymphatic or distant spread has taken place. 8, 9 Neoplasms located in the lower rectum can be adequately excised by the direct transanal approach. How'ever, middle and u p p e r rectal tumors are difficult to expose and, therefore, an attempt to remove them b y this approach can be complicated and unsatisfactory7 in terms of completeness of excision, which is the keystone for avoiding local recurrence. 1° A posterior approach provides somewhat improved exposure in these more cephalad tumors but m a y Iead to fecal fistulae or sphincter impairment in a significant percentage of patients. Transanal endoscopic microsurgery (TEM), as described b y Buess, .1, ~2 has been suggested as a novel alternative approach for the local excision of rectal neoplasms, especially those located in the middle and upper rectum. The few available reports on TEM excision of rectal tumors have suggested that the described procedure is a safe and efficient tectmique for the removal of rectal tumors. 12-14 In this report, w e have studied the outcome of a large series of patients with benign and malignant rectal tumors undergoing TEM excision in a single institution.

Lev-Chelouche D, Margel D, Goldman G, Rabau MJ. Transanal endoscopic microsurgery: experience with 75 rectal neoplasms. Dis Colon Rectum 2000;43:662-668. arcinoma of the rectum and some large villous a d e n o m a s are conventionally treated by anterior or combined abdominoperineal resection (APR). These traditional curative resections have often b e e n considered "radical" and sometimes disabling. Despite this aggressive approach, the reported five-year survival for rectal carcinoma is low, with an overall incidence of recurrent disease of up to 32 percent. ~, 2 Local resection

C

PATIENTS AND METHODS Patients Seventy-five patients with rectal tumors u n d e r w e n t TEM excision at the Tel Aviv Sourasky Medical Center b e t w e e n J a n u a r y 1992 and April 1998 (Table 1.). Cri-

Address reprint requests to Dr. Rabau: Department of Surgery B&C, Tel Av/v Sourasky Medical Center, 6 Weizmann Street, Tel Aviv 64239, Israel. 662

TEM FOR RECTAL NEOPLASMS

Vol. 43, No. 5

663

Table 1. TEM~Tumor Characteristics Parameter

Number Size (cm) Distance from dentate line (cm) teria for patient selection were mobile tumors, occupying less than one-third of rectal circumference or smaller than 3 cm in diameter; benign or well to moderately differentiated malignant histology with T1-T2 staging on transanal ultrasound (TRUS) w h e n performed. T3 cancers were included in the study for reasons of nonuse of ultrasound in early experience (n = 3) and poor-risk patients (n = 6). Also included were patients with p o o r surgical risk or those w h o refused radical surgery. Forty-six patients, 28 males, m e a n age, 69.3 years, had histologically proven villous a d e n o m a situated 3 t o 18 (mean, 7) cm from the dentate line. The 20 a d e n o m a s resected before the end of 1995 were at a m e a n distance of 6.3 cm from the dentate line because this was considered to be the learning curve, and even distal lesions were referred for TEM. Later on, only those tumors at a distance greater than 5 cm, at the level of the mid-rectum and higher, were resected in such a manner. The m e a n diameter was 2.5 cm. Twenty-nine patients, 16 males, m e a n age, 76.2 years, had adenocarcinoma of the rectum, situated 3 to 15 (mean, 7.75) cm from the dentate line. In the learning curve, the m e a n distance was initially 7,1 cm and later on 8.4 cm. The m e a n diameter was 3.25 cm.

Technique Before the procedure, the entire colon was evaluated with either colonoscopy or barium enema. Most of the cancers were evaluated with transrectal ultrasonography for preoperative staging. Preoperatively, a formal bowel cleansing was performed with whole gut tavage and perioperative antibiotics. The same team of surgeons performed all operations under general anesthesia. The tumors were resected by means of TEM as described by Buess. TM 12 In brief, the position of the patient in the operating r o o m was d e p e n d e n t on tumor location. Because the bevel of the rectoscope must face downward, patients were positioned accordingly. The surgical instruments, as well as the stereoscopic telescope, were m o u n t e d in such a w a y that the rectoscope was sealed and gas could not leak out. A video camera was connected so

Adenoma

46 1-7 (mean, 2.5) 3-18 (mean, 7)

Adenocarcinoma

29 2-5 (mean, 3.25) 3-15 (mean, 7.75)

that the operation could also be followed on a video monitor b y the entire team. Constant optimal exposure of the operative field was provided by automatic pressure controlled CO 2 insufflation. The margin of clearance was defined by coagulation clots; a d e n o m a s were removed with a 5-mm margin of normal mucosa, and dissection was undertaken in the submucosal plane. For carcinomas or large a d e n o m a s with firm areas or histologic evidence of atypia or dysplasia, w h e n the risk of harboring an occult cancer is increased, a 1-cm margin of normal tissue surrounding the lesion was obtained, and a full-thickness excision was conducted. When voluminous tumors were resected, the u p p e r part of the lesion was sometimes removed piecemeal; however, the base was always resected in one piece. The resuiting defect was closed b y transverse polydioxanone 3-0 sutures. After complete histopathologic examination of the tumor, patients were allocated to either a c o m p l e m e n tary radical surgery (anterior resection or APR), adjuvant radiotherapy, or no further treatment. Indications for radical resection after TEM were T3 tumors, inadequate resection margins, lymphatic or venous invasion, or p o o r histologic differentiation. The indication for adjuvant radiotherapy for cancer patients was T2 tumor staging; it was also suggested for patients at p o o r surgical risk and requiring complementary surgery and for those patients refusing radical surgery.

Pathologic Evaluation Immediately after surgery, specimens were spread on a cork specimen board b y means of fine needles to avoid anatomic distortion by formalin fixation, Before hematoxylin and eosin staining, margins were demarcated by ink coloration.

Follow-Up Patients were examined at three-month intervals for the first two postoperative years and six monthly thereafter. Complete clinical examination, digital, rectal, or rectovaginal palpation, as well as a rigid sig-

664

LEV-CHELOUCHE E T A L

moidoscopy, were performed. In addition, patients operated on for cancer underwent periodic metastatic workup. All local recurrences were verified by biopsy. RESULTS Complications Postoperative complications are summarized in Table 2. Minor complications such as fever and urinary retention occurred in a few patients and were all reversible. Anastomotic dehiscence manifested by early postoperative fever and pus-like discharge p e r a n u m occurred in three patients, all of w h o m responded to conservative measures of antibiotic therapy. Four patients suffered minor bleeding, and one required blood transfusion. Transient anal leakage occurred in three patients and persistent mild anal leakage in two. Repeat TEM resection for a recurrent T3 tumor caused an urethral injury in an elderly patient, which later led to the development of a rectourethral fistula. One patient had an intraperitoneal perforation, necessitating immediate conversion to anterior resection. This cardiac-crippled patient succumbed to fatal cardiac failure 18 hours later and was the only mortality in our study (1/75; 1.3 percent)). Immediate

Results and Follow-Up

The mean length of postoperative stay was 5.5 (range, 2-13) days and has been shortened to three days since 1996. Mean follow-up for patients with benign lesions was 36 (range, 5-76) months. Of the TEM procedures performed for 46 villous adenomas, three (6.5 percent) were converted immediately to low" anterior resection or APR because of poor operative exposure and inability to see the complete lesion. In four of the remaining 43 patients (9 percent), local recurrence of a villous adenoma was observed after 6, 20, 20, and 28 months. Of the latter, three Table 2. TEM--Postoperative Complications Fever Urinary retention

Anastomic dehiscence Intraperitoneal perforation Bleeding Urethra] injury

Transient anal leakage Persistent anal leakage Mortality

6 4 3 1 4 1 3 2 1

Dis Colon Rectum, May 2000

patients had reexcision by means of TEM and one required APR. No further recurrences have been revealed to date. The postoperative histologic staging of patients with malignant tumors is shown in Table 3. Eight patients with T1 tumors underwent no further treatment, as did three patients with T3 tumors whose general condition was either too debilitating or they refused additional treatment. Seven patients (24 percen0 underwent early complementary radical surgery; four immediately during the TEM procedure, including two T1 and two T2 tumors because of the inability to achieve adequate exposure and, thus, complete resection. One of the latter was a cardiaccrippled patient with a T2 tumor located in the anterior wall 12 cm from the dentate line. An intraperitoneal perforation was found during the procedure, and the operation was immediately converted to formal anterior resection. The remaining three patients underwent interval radical resection after pathological results T3 tumors. Eight T2 and three T3 patients underwent adjuvant radiation therapy. Mean follow-up time for patients with cancerous lesions was 34 (4-72) months. No recurrence was noted in T1 tumors. Local recurrence developed in 4 patients (18 percent) 12 to 36 months after TEM; two patients with T2 after adjuvant radiotherapy and two with T3 tumors who did not receive adjuvant therapy. Patients with local recurrence after T2 resection underwent repeat TEM or APR. One patient with local recurrence after T3 resection underwent repeat TEM excision after refusing radical salvage surgery. However, he further required a salvage APR after suffering urethral injury during the repeat TEM, which evolved into a rectourethral fistula. The other patient with T3 recurrence underwent local laser control because of poor general condition. The two patients with T3 tumors died less than one year postrecurrence of causes unrelated to cancer, whereas the two patients with T2 tumors are alive with no evidence of disease. During the overall fotlow up period, there have been four mortalities, none cancer-related. DISCUSSION The key factor in avoiding local recurrence after removal of a rectal adenoma is complete excision with sufficient tumor free margins. To achieve this goal, adequate exposure of the adenoma is required. Adenomas of the lower rectum are easily accessible

TEM FOR RECTALNEOPLASMS

Vol. 43, No. 5

665

Table 3.

TEM--FolIow-Up and Outcome Malignant Tumors Parameter No. No further treatment Early complementary surgery Adjuvant radiation therapy Local recurrence Local repeat resection Salvage radical surgery

Adenomas 46 43 3 4 3 1

by the transanat route. However, adenomas of the middle and upper rectum are often unsatisfactorily exposed transanally and require transabdominal or posterior approach routes for their removal. Since the introduction of TEM, a new route is available that permits excellent exposure and safe access to middle and upper rectal adenomas. The magnified stereoscopic view allows extremely accurate excision with safe margins as reflected by the 9 percent recurrence rate described in our series compared with published recurrence rates of 12 to 25 percent for locally excised adenomas.i5, 16 This finding could be a result of the improved exposure afforded by TEM. Because alternative techniques for complete removal of an adenoma involve major surgical procedures, the attending surgeon may be inclined to perform transanal excision for mid-rectal adenomas even in cases where the exposure is unsatisfactory. Some of the lesions in the present study, mainly those resected during the early years, were located in the distal rectum. These lesions were addressed by TEM to gain sufficient experience with the technique. However, there are disadvantages to TEM instrumentation w h e n operating in the distal rectum. First, the lowering of the rectoscope may slip externally, allowing carbon dioxide to escape. If this occurs, this operative field collapses and visibility is lost. Second, bleeding may occur from the hemorrhoid veins encountered near the anal canal. Brisk hemorrhage is difficult, but not impossible to control with TEM instrumentation. Furthermore, this is an expensive procedure. We suggest that limiting the surgical field to the transanal approach while using TEM for the middle and u p p e r rectum may reduce the overall recurrence rate after local excision of rectal adenomas. Local excision of rectal cancer with curative intent still remains a debatable issue. Several studies have shown favorable results after local excision of rectal cancer 4, 6. 7 The most important factor for satisfactory

T1

T2

T3

Total

10 8 2

10 2 8 2

9 3 3 3 2

29 11 7 11 4

1

1

2

1

1

2

outcome is appropriate patient selection. Criteria for patient eligibility for local control include mobile, well or moderately differentiated T1 and T2 tumors on TRUS, with a diameter of less than 3 cm, occupying less than 50 percent of the bowel wall circumference. These favorable features are associated with a low propensity for lymph node metastases, resulting in improved cure r a t e s F Preoperative tumor staging by means of clinical examination and imaging modalities achieve 80 to 90 percent accuracy. In addition, some of the patients are candidates for local excision because of significant comorbidity, whereas others refuse to undergo extensive surgery that might leave them with permanent disabilities such as incontinence, colostomy, or sexual derangements. This is w h y most of the reported series on local excision of rectal cancers also include patients with more advanced tumors. In our report, nine patients had T3 tumors, whereas only three had complementary radical surgery. An important factor in achieving satisfactory results is obtaining full-thickness tumor-free margins of the excised specimen. Because exposure of rectal neoplasms using ordinary anorectal retractors may be insufficient, not infrequently tumor excision is incomplete, leading to local failure. In our report, most of the resected tumors were located in mid-rectum (at an average distance of 7.75 cm from the dentate line). When adequate exposure was obtained, incomplete resection with involved margins was observed only in patients with T3 tumors (3/9, 33 percent). As in previous studies, our guidelines for postoperative treatment were T1 completely excised tumors, no further treatment; T2 completely excised tumors, adjuvant irradiation, positive surgical margins; or T3 tumors (excluding patients with major co-morbidity or refusal to undergo surgery), early complementary radical surgery. In those patients in w h o m these guidelines were strictly adhered to, a low incidence of local recurrence

666

LEV-CHELOUCHE E T A L

was observed; in the present series, this was the case in 2 of 16 of patients (12.5 percent) with T1 or T2 tumors. However, w h e n these guidelines were not strictly followed, as was the case in the six patients with T3 tumors, the local recurrence rate was 33 percent. It is important to emphasize, however, that the present study is still lacking a significant five-year follow-up period. No local recurrence was observed with T1 tumors. Only two of the 11 (18 percent) patients with tumors w h o received adjuvant radiotherapy suffered recurrences, and it might be of interest to note that none of the three patients with completely excised T3 tumors w h o underwent radiotherapy suffered recurrence. Thus, irradiation after TEM m a y prove effective in preventing local recurrence. Because rectal vasculature remains almost unaltered after local excision, the relative radioresistance of cancer cells in hypoxic tissue is avoided. There were 24 postoperative TEM related complications, most of which were minor and reversible. Postoperative bleeding or suture line dehiscence were of a self-limiting nature, responding to conservative measures only. After local excision during TEM, suturing of the remaining gap in the rectum is performed without releasing the rectal wall flaps. The larger the lesion, the bigger the remaining gap, the greater the tension that will be exerted on the suture line, thus increasing the risk of dehiscence. However, it would seem that extraperitoneal suture line dehiscence after TEM is of a significantly milder course than dehiscence after anterior resection and pelvic anastomosis. This may b e because after TEM, pelvic anatomy is unaltered and the perirectal spaces, as well as the peritoneal reflection, remain intact, walling off the inflammatory process and facilitating drainage through the suture line into the rectal lumen. Mild anaI incontinence, probably caused by the lengthy anal dilation resulting from the use of the 40-mm diameter operating rectoscope, occurred in five patients (6.6 percent); in three of them, the anal sphincter regained its function within three months, whereas two patients continued to suffer from persistent mild anal leakage. In two patients, major complications occurred; in one an attempt at full-thickness excision of a tumor located in the anterior wall of the rectum resulted in an intraperitoneal perforation. We suggest that resection of malignant neoplasms located on the anterior wall 10 cm or more above the dentate line should be avoided. Care should be exercised w h e n operating on anterior lesions, especially in fe-

Dis Colon Rectum, May 2000

males where there m a y be inadvertent entry into the vagina or the peritoneal cavity through the cul-de-sac, which is quite low in some females and unpredictable in its location. In the second patient, urethral injury occurred during a second attempt at TEM excision of a recurrent anterolateral invasive carcinoma. Because the anatomy is distorted after the first local excision and perirectal tumor infiltration hinders margin free tumor resection, we suggest that such tumors would be better m a n a g e d by salvage radical surgery rather than repeat TEM.

CONCLUSION TEM excision of rectal neoplasms is a rewarding technique for the removal and cure of a d e n o m a and selected carcinoma of middle and lower u p p e r rectum. Respect for patient selection guidelines, and mastery of the surgical technique and its pitfalls, produces results that are satisfactory with relatively short hospital stay and acceptable recurrence and complication rates.

REFERENCES 1. Danzi M, Ferulano GP, Abate S, Califano G. Survival and location of recurrence following abdominoperineaI resection for rectal cancer. J Surg Oncol 1986;311:235-9. 2. Sagar PM, Pemberton JH. Surgical management of locally recurrent rectal cancer. Br J Surg 1996;83:293-304. 3. Nivatvongs S, Balcos EG, Schottler JL, Goldberg SM. Surgical management of large, villous tumors of the rectum. Dis Colon Rectum 1973;16:508--16. 4. Rouanet P, Saint Aubert B, FabreJM, et al. Conservative treatment for low .rectal carcinoma by local excision with or without radiotherapy. BrJ Surg 1993;80:1452-6. 5. Roberts PL. Mucosectomy for treatment of rectal neoptasia. Semin Colon Rectum Surg 1996;7:210~1. 6. Nagle D. Full thickness locat excision of favorable rectal cancer. Semin Colon Rectum Surg 1996;7:215-20. 7. Bleday R, Breen E, Jessup JM, Burgess A, Sentovich SM, Steele G Jr. Prospective evaluation of local excision for small rectal cancers. Dis Colon Rectum 1997;40:388-92. 8. Roche B, Marti MC, Egeli R. Long term survival after transanal excision of rectal tumours. Tech Coloproctol 1996;1:6-9. 9. Kim DG, Madoff RD. Transanal treatment of rectal cancer: ablative methods and open resection. Semin Surg Oncol 1998;15:101-13. 10. Parks AG. A technique for excising extensive villous papillomatous changes in the lower rectum. Proc R Soc Med Lond 1968;61:441-2.

Vol. 43, No. 5

TEM FOR RECTALNEOPLASMS

11. Buess G, Kipfmuller K, Ibatd R, et al. Clinical results of transanal endoscopic microsurgery. Surg Endosc 1988; 2:235-50. 12. Buess G. Mentges B, Manncke K, Stalinger M, Becker HD. Technique and results of transanal endoscopic microsurgery in early rectal cancer, AmJ Surg 1992;163: 63-70. 13. Steele RJ, Hershman MJ, Mortensen NJ, Armitage NC, Scholefield JH. Transanal endoscopic microsurgery: initial experience from three centers in the United Kingdom. Br J Surg 1996;83:207-10. 14. Saclarides TJ. Transanal and endoscopic microsurgery: a single surgeon's experience, Arch Surg 1998;133: 595-9. 15. Chiu YS, Spencer RJ. Villous lesions of the colon. Dis Colon Rectum 1978;21:493-5. 16. Jahadi MR, Bailey W. Papillary adenomas of the colon and rectum: twelve-year review. Dis Colon Rectum 1975;18:249-53. 17. Saclarides TJ, Bhattacharyya AK, Britton-Kuzel C, Szeluga D, Economou SG. Predicting lymph node metastases in rectal cancer. Dis Colon Rectum 1994;37: 52-7. Invited Editorial To the Editor--Dr. Lev-Chelouche and colleagues provide a useful addition to the literature in support of transanal endoscopic microsurgery (TEM). They present a review of 75 patients w h o underwent tl~s procedure, a procedure first performed by Buess and Mentges in 1983. Tumors they removed were both benign and malignant, varied in size, and with a distance from the dentate line of 3 to 18 cm. The follow-up is complete, and although somewhat limited for malignancy, a mean of 34 months, it is longer follow-up than provided by most reviews on TEM. What the article demonstrates is that some surgeons can learn and properly apply the TEM technique to the benefit of select patients. What remains an issue regards the overall risks and benefits of TEM relative to established procedures. A more global issue regards whether it is time for this procedure to b e c o m e widely available, that is, whether it is time for all surgeons to embrace the procedure, purchase the equipment, and develop the skills. To put this into perspective, it seems most reasonable to examine this type of surgery for two separate indications: benign and malignant tumors. For benign tumors of the distal sigmoid and rectum, colonoscopic polypectomy is considered the first line of treatment. Failing this, the procedure of choice is transanal excision for lesions within reach (usually to

667

a proximal distance of 10 cm), and low anterior resection (LAR) for those more proximal. First considering distal benign tumors, the practice of transanal excision requires little in the way of special equipment; most patients are dismissed the same day as surgery; and further, it carries little morbidity'. In contrast, for TEM the cost of the equipment alone is over $40,000,: and this cost does not include the costs of the surgical procedure. The mean postoperative stay in this study was 5.5 days, reduced to 3 days late in the mature phase of the experience. The morbidity was low, and although two patients suffered permanent, mild incontinence, there is little in the literature to compare this with transanal excision, although it is probably similar. The operative mortality rate reported in this series reminds us that it is a real surgical procedure and, further, that any patient not suitable for anesthesia and abdominal surgery should not be considered for TEM. TEM has few if any advantages over local excision. However, w h e n comparing TEM with LAR for benign disease, TEM becomes more attractive. It can avoid an abdominal exposure and all the complications associated with abdominopelvic surgery, both short and long term. It can avoid the necessity for an intraperitoneal anastomosis in most patients. Although Dr. Lev-Chelouche and colleagues decided to abdominally explore the one patient w h o s e peritoneal cavity was entered, others have usually been able to repair this via TEM with little consequence.:. 2 Although the authors report a 9 percent recurrence rate, this is of much less concern with benign disease, and was usually remedied with repeat TEM. Although the benefits have not been quantitated or directly compared, this subset of patients would at least theoretically stand to gain the most from TEM. Next, reviewing this procedure for malignant lesions, it must be scrutinized not just as a technical procedure with risks, benefits, and costs, but most importantly it must be accountable to standard oncologic principles. In this regard, a minimal five-year follow-up should be mandatory. Unfortunately, this review and others previously reporting on TEM are off the mark. A mean follow-up of five years is not acceptable, because a patient with a local recurrence followed up for ten years will not adjust for the patient followed for one month without one. Even though eight patients were treated with radiation the> apy and the mean follow-up was only 34 months, Lev-Chelouche and colleagues reported a local recurrence rate of 18 percent. This is not acceptable w h e n