Long-term Follow-up of Local Rectal Cancer Surgery by Transanal Endoscopic Microsurgery

Transcription

1 World J Surg (2008) 32: DOI /s Long-term Follow-up of Local Rectal Cancer Surgery by Transanal Endoscopic Microsurgery Xavier Serra-Aracil Æ Helena Vallverdú Æ Jordi Bombardó-Junca Æ Carles Pericay-Pijaume Æ Joan Urgellés-Bosch Æ Salvador Navarro-Soto Published online: 13 March 2008 Ó Société Internationale de Chirurgie 2008 Abstract Background In 1997 we launched a prospective program of transanal endoscopic microsurgery (TEM) for the treatment of rectal cancer. Methods Suitability for TEM was based on endorectal ultrasound results, classified as follows: (I) benign tumors; (II) adenocarcinomas ut0 and ut1 with un0; (III) adenocarcinomas ut2- un0, low histological grade with intention to cure; and (IV) advanced stage adenocarcinomas with palliative care Results Transanal endoscopic microsurgery was performed in 218 patients: 122 adenomas, and 96 adenocarcinomas: group II 72, group III 19, and group IV 5. Follow-up was [24 months (median 59 months) in 61 patients. Nine were lost to follow-up, and so 52 patients were studied: group II 38, group III 11, and group IV 3. The Kaplan-Meier probability of nonrecurrence of adenocarcinoma by group was 93% in tumors in situ (Tis) and T1; and 77.8% in T2. The Kaplan-Meier probability of survival by group was 100% in Tis and T1 and 82% in T2. Conclusions Rates of recurrence and long-term survival in Tis and T1 adenocarcinomas treated with TEM are X. Serra-Aracil (&) J. Bombardó-Junca S. Navarro-Soto Department of General and Digestive Surgery, Corporació Sanitaria Parc Taulí, Parc Taulí s/n, Sabadell, Barcelona, Spain jserraa@cauli.cat X. Serra-Aracil H. Vallverdú J. Urgellés-Bosch Department of General and Digestive Surgery, Hospital de Sant Boi, C/Bonaventura Calopa, Sant Boi de Llobregat, Barcelona, Spain C. Pericay-Pijaume Department of Oncology, Corporació Sanitaria Parc Taulí, Parc Taulí s/n, Sabadell, Barcelona, Spain similar to those in previously published reports using conventional surgery. Further studies are required in T2 adenocarcinomas to determine a definitive strategy. Introduction The standard treatment for rectal adenocarcinoma today is total mesorectal excision [1]. This technique involves anterior resection of the lower rectum or coloanus, frequently in combination with protective ostomy or APR (abdominoperineal resection) with definitive colostomy. The operation is not without complications, and the rates of associated morbidity and mortality by approach are 33 and 2%, respectively [2]. Moreover, silent genitourinary alterations and sexual dysfunction may occur in 20 30% of patients [3, 4]. The capacity of local surgery to treat rectal cancer depends on the extent of lymph node impairment, which is found in 0 12% of T1, 12 28% of T2, and 36 79% of T3 patients [5]. Local surgery in rectal cancer is controversial, however, because of the discordant results described by different authors, even in the initial stages of the disease [6, 7]. Furthermore, comparison is difficult because many of these reports are retrospective studies that used very different procedures for patient selection. Transanal endoscopic microsurgery (TEM), first described by Buess et al. [8], is performed with a rectoscope 4 cm in diameter and 15 or 20 cm in length with a window with four port sites and three-dimensional optical vision incorporating a camera for monitoring and specific forceps (scalpel, needle-holder, grasping forceps). Transanal endoscopic microsurgery allows easy exposure of the lesion and good control of the margins, with

2 World J Surg (2008) 32: total excision of the rectal wall. Publications of series using this technique in T1 cancers report recurrence rates of less than 5% and a disease-related survival of 100% at 5 years [9, 10]. We present the results of a prospective protocol of local surgery for rectal adenocarcinoma using TEM, with a lengthy follow-up period. Methods and patients The Corporació Sanitària Parc Taulí and the Hospital de Sant Boi are referral centers in Catalonia for the local treatment of rectal tumors by TEM. All patients prospectively diagnosed with rectal cancer undergo a strict preoperative study that includes complete colonoscopy with multifocal biopsy of the lesion, completion of a questionnaire on sphincter function [11], endorectal ultrasound, abdominal computed tomography (CT), and CEA and CA 19.9 tumor markers. All patients signed informed consent. Mechanical colon lavage was performed and the patients were administered prophylactic antibiotics and thromboembolic drugs. Surgery was carried out under general anesthesia. The patients were positioned so that the tumor site was in the inferior position. The TEM technique involved dissection of the tumor with margins of 5 10 mm from the lesion. Because of the high frequency of adenocarcinoma in the definitive specimen, wall resection was complete, even in cases in which the preoperative biopsy revealed an adenoma. Lavage was performed with saline solution containing iodine povidone prior to and after suturing the wound. Transanal endoscopic microsurgery presents technical limitations when the tumor is situated more than 20 cm posterior or more than 15 cm anterior or lateral to the anus. The technique presents partial limitations when treating large, almost circumferential lesions. Postoperative morbidity was recorded. Sphincter function was assessed by a questionnaire administered prior to surgery and 1 week and 3 months postoperatively [11]. All patients underwent a strict follow-up protocol including the following procedures during the first 2 years: endorectal ultrasound, rectosigmoidoscopy-biopsy, and CEA every 4 months. In the third, fourth, and fifth years endorectal ultrasound, rectosigmoidoscopy-biopsy and CEA were performed every 6 months. Total colonoscopy and abdominal CT were performed annually. The patients were assigned to four groups, based on endorectal ultrasound staging (Table 1). In the pathological examination, surgical specimens were staged according to the TNM classification of the AJCC (American Joint Committee on Cancer) from T0 T4 [12]. Poor cellular Table 1 Transanal endoscopic microsurgery (TEM) patient classification TEM groups Group I Group II Group III Group IV differentiation and lymphovascular or perineural invasion are likely indicators of regional lymph node metastasis [13]; group II patients presenting these factors were moved to group III and treatment was by consensus: that is, patients declined the recommendation of conventional surgery as salvage therapy and accepted the risks of recurrence. The criteria used in the pathological examination were always the same and were established by two teams of pathologists. We present the results over a minimum follow-up period of 24 months for tumor recurrence and survival in patients undergoing TEM for rectal adenocarcinoma. The description of the variables and statistical analyses were performed with SPSS program, version The quantitative variables are described by mean values and standard deviation when the distribution was considered normal, based on the Shapiro Wilks and Lilliefors tests. When the distribution was non-normal, median values, the interquartile interval and range are reported. The distributions are presented in a boxplot chart (Fig.1). Categorical variables are described in absolute numbers and percentages. The results of the statistical tests are given, whenever possible, with the confidence interval (CI) of 95%. Descriptive analyses of the follow-up data for both recurrence and survival were carried out using the Kaplan-Meier method. Results Preoperative endorectal ultrasound findings Benign tumors (adenomas): ut0 Adenocarcinomas: ut0 a and ut1 with un0 (intention to cure) Adenocarcinomas: low histological grade ut2 un0 ( early rectal cancer ) Advanced stage adenocarcinomas with palliative care a ut0, adenocarcinoma biopsy; no invasion of muscularis mucosae on ultrasound This prospective protocol for local TEM and follow-up was initiated in April Two hundred and eighteen patients underwent surgery in two centers; 122 had adenomas and 96 had adenocarcinomas. The 96 adenocarcinomas were classified by group (Table 1): group II comprised 72 adenocarcinomas ut0 and ut1 with un0; group III, 19 adenocarcinomas ut2 un0, with low histological grade and intention to cure; and group IV, 5 advanced stage adenocarcinomas treated with palliative care. The mean age of the patients was 66 years (CI 95%: years; range: years); 64 were men

3 1164 World J Surg (2008) 32: Table 2 Postoperative morbidity Postoperative morbidity (9%) Surgical complications Nonsurgical complications perforations (3.1%) 2 febrile syndromes 2 reinterventions 2 cardiac insufficiencies 1 direct suture with favorable evolution 2 self-limiting rectorrhagias 2 urinary tract infections N = Fig. 1 Lesion size: median: 40 mm; interquartile range (IQR): 20 mm The final study sample thus comprised 52 patients. The median length of follow-up was 59 months (IQR 16.5 months), ranging from 24 to 97 months. After pathological study patients were assigned to groups, as follows: group II (intention to cure, Tis (T in situ), T1 N0) comprised 38 patients; group III ( early rectal cancer with unfavorable histological features T2 N0) had 11; and group IV (advanced stage adenocarcinomas with palliative care) had 3. Two of the 38 patients in group II developed local recurrence (5.3%). According to the Kaplan-Meier method, the probability of non-recurrence at the end of the followup period was 93.3% (Fig. 3). Within group II, patients with Tis (n = 22) were differentiated from those with T1 (n = 16). There was one recurrence among Tis patients (4.5%). The Kaplan-Meier probability of non-recurrence was 94.4%. There was also one recurrence among the T1 patients (6.25%), with a probability of non-recurrence of 92.3%. 1,00 Survival Function 0 N = Fig. 2 Distance to anal margin: median: 9 cm; IQR: 4 cm (67%) and 32 were women (33%). The median distance to the anal margin was 9 cm, with an interquartile range (IQR) of 4 cm and a range from 1 to 20 cm (Fig. 1). The median lesion size was 40 mm, with an IQR of 20 mm and a range from 10 to 115 mm (Fig. 2). The median postoperative stay was 5 days (IQR 2 days). Postoperative morbidity was 9% (Table 2). Three perforations in the peritoneal cavity were recorded in the early stages of the series, two of which required reintervention within h; the accident was identified during surgery in the third case and suture was performed, with favorable evolution. Postoperative mortality in this group of patients with adenocarcinoma was zero. Sixty-one patients were studied for a minimum followup of 24 months. Nine patients who lived outside the referral areas of the two hospitals were lost to follow-up. 91 Cum Survival,90,80,70,60,50,40,30,20, Recurrence follow-up Survival Function Censored Fig. 3 Probability of non-recurrence according to the Kaplan Meier method. Recurrence per group: group II intention to cure: 38 patients

4 World J Surg (2008) 32: Table 3 Patients who were rescued with radical conventional abdominal surgery following relapse after local TEM surgery Group indication Pathology (pt) Distance to margin (cm) Lesion size (mm) Time to recurrence (months) Rescue intervention Pathology postrecurrence Stage postrecurrence Survival after recurrence (months) II ptis APR pt3, pn0 IIA 30 II pt APR pt2, N0 I 46 III pt Anterior resection pt3, pn0 IIA 54 III pt Anterior resection pt3, pn0 IIA 26 APR, abdominoperineal resection In group III (11 patients), two patients were rescued after TEM, as pathological criteria advised radical abdominal surgery. In addition, 2 recurrences were observed (22.2%). According to the Kaplan-Meier method, the probability of non-recurrence at the end of the followup period was 77.8%. One patient was upgraded from group II to III, due to the pathological findings after TEM. He is currently alive with no recurrence. Table 3 shows the characteristics of the 4 patients who were rescued with radical conventional abdominal surgery following relapse after local TEM. They are currently alive and disease-free after 30, 46, 54, and 26 months respectively (Table 3). Patients with locally advanced lesions (T2) did not receive radiation or chemotherapy. After consultation with the multidisciplinary committee, we decided to rescue patients with poor prognosis (two cases) with immediate radical surgery, and the others were strictly controlled. The two patients who were rescued after TEM are now diseasefree 36 and 40 months after the rescue operation. Pathological examination did not show any residual disease. All the recurrences occurred in the wall, close to the scar, though the margins in the pathology specimen were disease-free. Thanks to the follow-up protocol, recurrence was detected promptly. For this reason the radical surgery in these patients did not differ from conventional surgery in terms of technical difficulty, time, or postoperative morbidity. Four patients in group II died of other causes during follow-up (10.5%), two with Tis (9.1%) and two with T1 (12.5%). The remaining patients (n = 34) are currently alive and cancer-free. In group III, one patient developed hepatic metastasis at 12 months and died 20 months after surgery, with a Kaplan Meier probability of survival of 82%. Of the three group IV patients, two died of other causes during follow-up, and the third developed recurrence at 6 months. Discussion Conventional rectal cancer surgery is particularly aggressive, involving total mesorectal excision with APR or anterior resection of the rectum. In the search for an alternative strategy, the first results obtained with local surgery in patients with early stage rectal cancers were promising, with rates of recurrence of less than 8% [14 16]. However, later series with longer follow-up found that the recurrence rates in the initial stages of T2 and apparent N0 tumors identified by complementary tests were unacceptably high, above 25% [6, 9, 17]. Recent long-term results in T1 and apparent N0 tumors have also reported high recurrence rates, between 18% and 29% [6, 7, 9]. However, these series were from retrospective studies which, despite attempts to homogenize the cohorts, lacked uniform criteria for preoperative patient selection, surgical treatment, and follow-up. In contrast, prospective studies in T1 have shown rates of recurrence of 3.4% [18] and 9.5% [19]. For effective monitoring of recurrence and survival in patients with T1 tumors, it is essential to have a prospective protocol with a preoperative study including patient selection, standardized surgical technique (we recommend TEM), and strict follow-up control. The complementary explorations (endorectal ultrasound, nuclear magnetic resonance imaging, and last-generation abdominal CT) for patient selection should identify suitable candidates and rule out those with lymphatic or systemic dissemination. Compared with endoanal excision, TEM provides a better view of the lesion and continuous control of the margins, and it also allows resection of the complete wall. The recent application of the ultrasonic scalpel for tumor dissection in TEM facilitates performance, allows better hemorrhage control, and considerably shortens the length of surgery. According to Lezoche et al. [20] and Lee et al. [9], the results of recurrence reported with TEM in T1 were 0% and 4.1% at 34 and 60 months of follow-up, respectively. Our results are similar, with recurrence rates of 4.5% in Tis and 6.5% in T1 after a mean follow-up of 59 months. Comparison with the results of other series is difficult because of the differences in the local resection techniques used. For adequate selection of patients likely to evolve favorably in terms of recurrence and disease survival, it is

5 1166 World J Surg (2008) 32: important to specify the pathological characteristics of the surgical specimens in order to reduce the risk of lymph node metastasis. Kikuchi et al. [21] suggested that T1 be divided into three types, according to the degree of submucosal invasion: the higher the degree of submucosal invasion, the greater the probability of lymph node metastasis. Poor differentiation and lymphovascular or perineural invasion are also factors that raise the probability of lymph node metastasis [13]. We used these factors to define criteria of poor prognosis in our study. In a recent multivariate analysis in a long-term follow-up study, Wang et al. [22] demonstrated that the risk factors for lymph node metastasis are the degree of cell differentiation, lymphatic invasion, the presence of inflammation around the cancer, and budding at the invasive front of the tumor. Two patients in our series with some of these factors contributing to poor prognosis were reconverted to radical surgery. It has been reported that reconversion to radical surgery when the definitive study of the specimen shows higher staging than expected does not alter survival [23]. The main objective in T1 rectal tumors that are suitable for local surgery is to achieve 5-year survival. Despite a high incidence of recurrence in long-term retrospective studies, the 5-year survival is almost 90% [7]. In series using TEM, the survival for T1 is 100% [8, 20]; our series was close to reproducing this rate. Nonetheless, the final results should be evaluated because, if these patients are not treated with local surgery, they will undergo total mesorectal excision with a risk of operative mortality and secondary sequelae including genitourinary and sexual dysfunction, as well as the sequelae related to the temporary or permanent ostomy. As already stated, group III comprised patients who in our view required rescue after TEM surgery and those who, after being informed of this situation, declined radical surgery and accepted the risk of recurrence. For this reason, the pt1 with factors of poor prognosis and the pt2 were placed in the same group. What should be done with high grade T1 and T2 tumors? Could a combination of local surgery with adjuvant radiotherapy and chemotherapy provide similar results to those of radical abdominal surgery? It has been reported that local surgery alone achieves poor results in these tumors. Recent publications have described that the combination of local surgery with chemoradiotherapy achieves results similar to those of conventional surgery in terms of local recurrence and survival [24, 25]. Even in a retrospective review of 10 patients with T1 tumors treated with local excision in combination with chemoradiotherapy, Lamont et al. [26] did not observe recurrence; in 17 other patients in whom only surgery was performed, recurrence was found in four (24%). However, these were retrospective studies including a small number of patients. Studies of how to treat the initial stages of rectal cancer should not defend the merits of one type of surgery or another; the objective must be to establish a surgical procedure that is safe and best guarantees the quality of life of the patients. Patients should not be overtreated, but neither should the serious risks involved in surgery dissuade us from trying to eradicate the disease. A definitive response is most likely to be achieved with randomized, controlled, prospective comparisons of local and radical abdominal surgery, though, for ethical reasons, they may be difficult to perform. It is, therefore, important to know the results of different study groups using prospective protocols with uniform local surgical treatment that can be compared with other series. References 1. Heald RJ, Ryall RDH (1986) Recurrence and survival after total mesorectal excision for rectal cancer. Lancet 1: Law WL, Chu KW (2004) Anterior resection for rectal cancer with mesorectal excision: a prospective evaluation of 622 patients. Ann Surg 240: Kneist W, Junginger T (2004) Residual urine volume after total mesorectal excision: an indicator of pelvic autonomic nerve preservation? Results of a case-control study. Colorectal Dis 6: Shah EF, Huddy SPJ (2001) A prospective study of genito-urinary dysfunction after surgery for colorectal cancer. Colorectal Dis 3: Mellgren A, Sirivongs P, Rothenberger DA et al (2000) Is local excision adequate therapy for early rectal cancer? Dis Colon Rectum 43: Garcia-Aguilar J, Mellgren A, Sirivongs P et al (2000) Local excision of rectal cancer without adjuvant therapy: a word of caution. Ann Surg 231: Madbouly KM, Remzi FH, Erkek BA et al (2005) Recurrence after transanal excision of T1 rectal cancer: should we be concerned? Dis Colon Rectum 48: Buess G, Mentges B, Manncke K et al (1992) Techique and results of transanal endoscopic microsurgery in early rectal cancer. Am J Surg 163: Lee W, Lee D, Choi S et al (2003) Transanal endoscopic microsurgery and radical surgery for T1 and T2 rectal cancer. Surg Endosc 17: Taylor RH, Hay JH, Larsson SN (1998) Transanal local excision of selected low rectal cancer. Am J Surg 175: Jorge JMN, Wexner SD (1993) Etiology and management of fecal incontinence. Dis Colon Rectum 36: Compton C, Fenoglio-Preiser CM, Pettigrew N et al (2000) American Joint Committee on cancer prognostic factors consensus conference: colorectal working group. Cancer 88: Saclarides TJ, Bhattacharyya AK, Britton-Kuzel C et al (1994) Predicting lymph node metastasis in rectal cancer. Dis Colon Rectum 37: Morson BC, Bussey HJ, Samoorian S (1977) Policy of local excision for early rectal cancer of the colorectum. Gut 18: Hager T, Gall FP, Hermanek P (1983) Local excision of cancer of the rectum. Dis Colon Rectum 26:

6 World J Surg (2008) 32: Grigg M, McDermott FT, Pihl EA et al (1984) Curative local excision in the treatment of adenocarcinoma of the rectum. Dis Colon Rectum 27: Paty PB, Nash GM, Baron P et al (2003) Long-term results of local excision for rectal cancer. Ann Surg 236: Steele GD, Herdon JE, Bleday R et al (1999) Sphincter-sparing treatment for distal rectal adenocarcinoma. Ann Surg Oncol 6: Bleday R, Breen E, Jessup JM et al (1997) Prospective evaluation of local excision for small rectal cancers. Dis Colon Rectum 40: Lezoche E, Guerrieri M, Paganini AM et al (1998) Transanal endoscopic microsurgical excision of irradiated and non-irradiated rectal cancer. Surg Laparosc Endosc 8: Kikuchi R, Takano M, Takagu K et al (1995) Management of early invasive colorectal cancer: risk of recurrence and clinical guidelines. Dis Colon Rectum 38: Wang H-S, Liang W-Y, Lin T-Ch et al (2005) Curative resection of T1 colorectal carcinoma: risk of lymph node metastasis and long-term prognosis. Dis Colon Rectum 48: Hahnloser D, Wolff BG, Larson DW et al (2005) Immediate radical resection after local excision of rectal cancer: an oncologic compromise? Dis Colon Rectum 48: Lezoche EL, Guerrieri M, Paganini AM et al (2002) Long-term results of patients with T2 rectal cancer treated with radiotherapy and transanal endoscopic microsurgical excision. World J Surg 26: Chen Ch-Ch, Leu Sh-Y, Liu M-Ch et al (2005) Transanal local wide excision for rectal adenocarcinoma. Hepato-gastroenterology 52: Lamont JP, McCarty TM, Digan RD et al (2000) Should locally excised T1 rectal cancer receive adjuvant chemoradiation? Am J Surg 180: