| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Departments of Surgery (HGM, PP, DW, JGG, AMC), Epidemiology and Biostatistics (ER), Radiation Oncology (BDM), and Medicine (LS), Memorial Sloan-Kettering Cancer Center, New York, New York.
Correspondence: Address correspondence and reprint requests to: Jose G. Guillem, MD, MPH, 1275 York Ave., Room C-1077, New York, NY 10021; Fax: 646-422-2318; E-mail: guillemj{at}mskcc.org
| ABSTRACT |
|---|
|
|
|---|
1 cm in patients with locally advanced rectal cancer requiring preoperative CMT. Methods: Ninety-four consecutive patients, status post curative low anterior resection for rectal cancer after preoperative CMT, were identified from the prospective Colorectal Service Database. Distal margin length, tumor grade, tumor-node-metastasis stage, presence of lymphovascular and perineural invasion, and tumor distance from the anal verge were examined for their effect on recurrence and survival. Median follow-up was 44 months.
Results: Distal margin length ranged from .1 to 9.5 cm (median, 2.0 cm) and did not correlate with local recurrence (hazard ratio, 1.1; P = .34) or recurrence-free survival (hazard ratio, 1.1; P = .29) by univariate analysis. Kaplan-Meier estimates of recurrence-free survival and local recurrence at 3 years for the
1 cm versus >1 cm and the
2 cm versus >2 cm groups were not significantly different. Groups were well matched for other clinicopathologic variables.
Conclusions: Our data suggest that for patients with locally advanced rectal cancer undergoing resection and preoperative CMT, distal margins
1 cm do not seem to compromise oncological outcome.
Key Words: Rectal cancer Distal margin Combined-modality therapy Total mesorectal excision
| INTRODUCTION |
|---|
|
|
|---|
Because distal intramural spread (DIS) of tumor rarely exceeds 1 to 2 cm in most rectal cancers37 and because local control and survival do not seem to be compromised by shorter distal resection margins,714 the generally accepted practice is to aim for obtaining a 2-cm distal margin. However, because the likelihood of distal spread beyond 1 cm increases with tumor stage,5 this policy may not be justified in locally advanced rectal cancers requiring preoperative CMT.
A recent report of a small series of patients receiving preoperative CMT followed by sphincter-sparing (SS) resection demonstrated that patients with distal resection margins of
1 cm did not have compromised disease-free survival versus those with margins >1 cm and those undergoing abdominoperineal resection (APR). However, no definitive conclusions were possible with regard to differences in local recurrence (LR) because only 1 of 28 patients undergoing restorative resection had local failure.15 The purpose of this study was, therefore, to assess the effect of a distal margin
1 cm on local control and survival in a large series of rectal cancer patients who were treated with preoperative CMT and underwent restorative resection by using the technique of sharp mesorectal excision (SME).
| MATERIALS AND METHODS |
|---|
|
|
|---|
Preoperative CMT
All patients in this analysis underwent preoperative CMT at Memorial Sloan-Kettering Cancer Center. The standard off-protocol regimen used during the study period consisted of two cycles of 5-FU and leucovorin (LV) (bolus x 5 days) repeated every 28 days, concurrently with 4680 cGy of pelvic radiation in 180-cGy fractions with a 360-cGy boost to a total dose of 5040 cGy. The details of this regimen have been previously reported.16 Eighty-nine patients received 5040 cGy of radiation, two patients received 4500 and 5000 cGy, and in four patients the final radiation dose was not available, but records indicate that the patients completed a full course of therapy. Eighty-eight patients received 5-FU/LV in combination with radiation. Seven patients received radiation in combination with irinotecan, according to a previously published report.17 Patients underwent surgery between 4 and 7 weeks after completing CMT. After surgery, 81 patients received additional 5-FU/LV (most commonly 4 cycles), 6 patients did not receive chemotherapy, and records regarding postoperative chemotherapy were unavailable in 7 patients.
Surgical Technique
All patients underwent LAR by using the technique of SME. This technique involves dissection in the areolar plane between the fascia propria (visceral fascia) of the rectum and the parietal pelvic fascia. For upper rectal tumors, the rectum and mesorectum were divided 5 cm distal to the caudal tumor edge. For tumors of the mid and distal rectum, a total mesorectal excision was performed.18
Pathologic Assessment
The prospective Colorectal Service Database was queried for distal margin length, tumor differentiation, tumor-node-metastasis stage, presence or absence of lymphovascular or perineural invasion, mucinous histology, and distance from the anal verge. Tumor-node-metastasis staging represents the final pathologic stage of the resected specimen after CMT. For the purposes of analysis, patients who were T0N0M0 on final pathologic assessment were designated as stage 0. Stages I to IV were defined according to the International Union Against Cancer classification.19 There were 8 stage 0, 38 stage I, 30 stage II, 17 stage III patients, and 1 stage IV patient. The one stage IV patient had liver metastases that were curatively resected within 3 months of LAR.
Distal Margin Assessment
In general, distal margins were measured and recorded by the pathologist before tissue fixation in the unpinned specimen. The "donuts" created by circular intraluminal staplers were not included in the measurement of distal margin length, but they were examined and found to be negative for tumor in all cases. When a complete pathologic response of the primary tumor to CMT occurred (no evidence of gross or microscopic tumor remaining), the distal margin was reported as the distance between the caudal edge of the residual scar and the cut margin of the rectum.
Patient Follow-Up
Median follow-up was 44 months (range, 24118 months). Disease status was available for all surviving patients between January 2000 and January 2001. When an assessment of disease status was not available by either the attending colorectal surgeon or the medical oncologist during this time period, patients or their primary physicians were contacted by telephone. Medical records of all patients, including imaging studies, endoscopy reports, operative reports, carcinoembryonic antigen determinations, and clinic notes, were examined for evidence of local and distal recurrence.
Statistical Analysis
Time to recurrence was measured from the date of surgery. Survival and recurrence distributions were estimated with the Kaplan-Meier method.20 For the analysis of recurrence-free survival (RFS), events were considered to be recurrence (local or distant) and/or death. Distal margin length was treated as both a continuous measure and a categorical variable (
1 vs. >1 cm, and so on). The relationship between time to recurrence and distal margin was tested by using the Cox proportional hazards model21 when the distal margin was considered continuous and by using a log-rank statistic22 when the distal margin was treated categorically. Comparison between groups for other clinicopathologic variables was performed with Fishers exact test for categorical variables and with a Students t-test for continuous variables. All analyses were performed with SASTM, version 8 (SAS Institute, Cary, NC).
| RESULTS |
|---|
|
|
|---|
|
2 cm margins (n = 53) and those with >2 cm margins (n = 41). Kaplan-Meier distributions of LR and RFS were compared for the
2 cm versus >2 cm groups and were noted to be not significantly different (P = .99 and P = .80, respectively). Estimates of LR and RFS at 3 years between the
2 cm and >2 cm groups were 8% versus 11% and 88% versus 82%, respectively (Table 1).
|
1 cm (n = 17) and >1 cm (n = 77) groups. Kaplan-Meier estimates of LR (Fig. 2) and RFS (Fig. 3) were compared and were not significantly different (P = .93 and P = .88, respectively). Estimates of LR and RFS at 3 years between the
1 cm and >1 cm groups were 12% versus 9% and 82% versus 85%, respectively.
|
|
A summary of the other clinicopathologic variables between the
1 cm and >1 cm groups can be found in Table 2. Comparison with Fishers exact test did not reveal any significant differences between groups for stage distribution, mucinous histology, presence of lymphovascular or perineural invasion, or differentiation. The mean distance from the anal verge was, however, significantly different between the
1 cm and >1 cm groups (P = .0002). Similarly, comparison between the
2 cm and >2 cm groups did not reveal any significant differences, except that the mean tumor distance from the anal verge was also significantly shorter for the
2 cm patients (P = .0002). Univariate analysis did not reveal any associations with either LR or RFS for any of the clinicopathologic variables in Table 2, although there was a significant difference in RFS by stage (P = .03). RFS was clearly improved between stage I and stage III/IV patients (Fig. 1).
|
| DISCUSSION |
|---|
|
|
|---|
1 cm do not seem to adversely affect LR or survival. One study, however, did demonstrate an increased incidence of anastomotic recurrence when the distal resection margin was <.8 cm.11
The importance of distal margin length in patients with advanced rectal cancer undergoing preoperative CMT, however, has not been fully studied. Preoperative CMT may theoretically eradicate any foci of distal intramural tumor spread before surgery, such that distal margins of
1 cm would be even less likely to compromise outcome in this patient population. However, these patients have advanced and potentially more aggressive rectal cancer, and the conclusions presented previously may not necessarily hold for this group of patients.
One study has addressed the effect of distal margin length in patients receiving preoperative CMT. Thirty-six patients underwent either SS procedures (n = 28) or APR (n = 8) after preoperative radiotherapy and 5-FUbased chemotherapy.15 There was no significant difference in disease-free survival between patients undergoing SS procedures with
1 cm margins and those with >1 cm margins (P = .06) or those undergoing APR (P = .27). Because LR occurred in only 1 of the 28 SS patients, no meaningful comparison between SS groups could be made, although LR occurred in only 1 of 9 patients with distal margins <5 mm. This study, however, does not provide information regarding other clinicopathologic variables that might have potentially masked differences in outcome attributable to distal margin length. Also included were different SS procedures (LAR and transsacral resection) and recurrent colorectal cancer patients (n = 3).
Our results, in a larger and relatively homogeneous group of rectal cancer patients undergoing preoperative CMT followed by potentially curative LAR, demonstrate that distal margins
1 cm do not seem to compromise LR or RFS. Although a number of different clinicopathologic variables are associated with outcome in curatively resected rectal cancer, comparison between groups in this study indicates that they are well matched. As expected, the mean distance of tumor from the anal verge was significantly shorter in the
1 and
2 cm groups. Despite the fact that more-distal rectal cancers are associated with a relatively poor prognosis, outcome in the shorter-margin groups was equivalent to that in those with longer distal margins.
All patients in this study underwent rectal cancer resection with the technique of SME, as described by Enker et al.23 and Heald et al.24 Use of this technique results in negative circumferential margins in 93% of cases25 and LR rates as low as 3% to 5% in large clinical series of patients with all stages of disease, with most not receiving adjuvant therapy.23,26 In previous series of patients undergoing LAR for rectal cancer with total mesorectal excision, distal resection margins of
1 cm did not negatively influence LR12 or survival.12,15
The negative prognostic effect of positive circumferential margins is well established.25 In rectal cancer patients undergoing preoperative CMT followed by SS resection, circumferential margins of
3 mm were associated with increased overall recurrence and disease-free survival.15 However, the retrospective nature of our study did not allow a rigorous measurement of the radial margin. Patients with positive radial margins (considered noncurative resections) were excluded to avoid the potentially confounding influence of this variable.
Although our data clearly support our conclusions, our study is limited by virtue of its retrospective design. Furthermore, although it is commonly believed that approximately 80% of rectal cancer recurrences occur within 2 years, the development of recurrence may be delayed after preoperative CMT. In a recent study of patients receiving preoperative radiotherapy for locally advanced rectal cancer, almost 5 years of follow-up were required to detect 80% of all recurrences.27 Our median follow-up of 44 months compares favorably with other reports, but a median of at least 60 months would seem to be optimal. Finally, the favorable outcome of our patients supports the efficacy of preoperative CMT and SME but limits the number of treatment failures available for analysis.
It is important to emphasize that obtaining a negative distal margin of at least 1 to 2 cm remains an important goal in restorative rectal cancer resection. However, in the setting of a curatively resected patient (negative circumferential and distal margins), status post preoperative CMT with an otherwise favorable tumor, a distal resection margin <1 cm does not necessarily portend a poor prognosis. Our data suggest that these patients do not seem to require further resection in an effort to procure a greater distal margin, but that they may be followed nonoperatively without compromising oncological outcome.
Ongoing, prospective studies at our institution will aim to address the relative contributions of distal and circumferential margin status to local control and survival in patients undergoing restorative rectal cancer resection with rigorous pathologic assessment of circumferential margin status and DIS. Long-term follow-up of these patients will elucidate the relative importance of distal and circumferential margin status in patients with advanced rectal cancer undergoing preoperative CMT.
| Acknowledgments |
|---|
Supported by National Cancer Institute grant R01 CA 82534-01 (JGG).
| Footnotes |
|---|
Although preoperative combined-modality therapy (CMT) can facilitate sphincter preservation, the optimal margin of distal clearance remains undefined. Our results indicate that distal margins
1 cm do not compromise outcome in locally advanced rectal cancer patients receiving CMT.
Received for publication April 10, 2002. Accepted for publication August 26, 2002.
| REFERENCES |
|---|
|
|
|---|
1 cm distal margins sufficient? Ann Surg Oncol 2001; 8: 1639.This article has been cited by other articles:
![]() |
H. B. Neuman, D. Schrag, C. Cabral, M. R. Weiser, P. B. Paty, J. G. Guillem, B. D. Minsky, W. D. Wong, and L. K. Temple Can Differences in Bowel Function After Surgery for Rectal Cancer Be Identified by the European Organization for Research and Treatment of Cancer Quality of Life Instrument? Ann. Surg. Oncol., May 1, 2007; 14(5): 1727 - 1734. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Han and S. Galandiuk Induction Chemoradiation for Rectal Cancer Arch Surg, December 1, 2006; 141(12): 1246 - 1252. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Rengan, P. Paty, W. D. Wong, J. Guillem, M. Weiser, L. Temple, L. Saltz, and B. D. Minsky Distal cT2N0 Rectal Cancer: Is There an Alternative to Abdominoperineal Resection? J. Clin. Oncol., August 1, 2005; 23(22): 4905 - 4912. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. A. Paraskeva, S. Purkayastha, and A. Darzi Laparoscopy for Malignancy: Current Status Surgical Innovation, March 1, 2004; 11(1): 27 - 36. [Abstract] [PDF] |
||||
![]() |
M. L. Smidt, J. A. Wegdam, A. J. A. Bremers, and R. P. Bleichrodt Missing Evidence for the Adequacy of a 1-cm Distal Margin in Resected Rectal Cancer Ann. Surg. Oncol., August 1, 2003; 10(7): 823 - 824. [Full Text] [PDF] |
||||
![]() |
J. G. Guillem, H. G. Moore, P. B. Paty, A. M. Cohen, and W. D. Wong Adequacy of Distal Resection Margin Following Preoperative Combined Modality Therapy for Rectal Cancer Ann. Surg. Oncol., August 1, 2003; 10(7): 824 - 824. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |