Does Rectal Washout During Anterior
Resection Prevent Local
Tumor Recurrence?
Emmanuel A. Agaba, M.B.B.S., F.R.C.S. (Edinb.)
Academic Surgical Unit, Castlehill Hospital, Cottingham-Hull, United Kingdom
PURPOSE: Colorectal cancer is one of the leading causes of
cancer deaths in the industrialized nations. Left-sided tu-
mors, especially rectal, rectosigmoid, and sigmoid, account
for more than half of these tumors. Among many colorectal
surgeons, the practice of rectal washout with cytocidal
agents before anastomosis is common. It is widely believed
that this practice prevents implantation of free malignant
cells. It is unclear whether this translates into a reduction in
the incidence of local recurrence. This study was designed
to evaluate the effectiveness of cytocidal rectal washout in
reducing the incidence of local recurrence. METHODS:
Case notes and histology reports of all patients who under-
went curative anterior resection for adenocarcinoma of the
rectum and rectosigmoid between 1992 and 1994 were re-
viewed. A total of 141 patients were deemed suitable for the
study. Of these, 90 patients underwent rectal washout using
cetrimide before anastomosis. Fifty-one patients did not
have rectal washout before anastomosis. Local recurrences
between the two groups were compared. RESULTS: The
two study groups were identical in all respects. Overall, the
local recurrence rate for all comers was 5 percent (n = 7).
Among the washout group, the local recurrence rate was
4.4 percent (n = 4) compared with 5.9 percent (n = 3)
among the no washout group. CONCLUSIONS: Because of
the size of the study, we were unable to demonstrate the
benefit or lack thereof of cytocidal agents in reducing local
recurrence. [Key words: Rectal washout; Local recurrence]
C olorectal cancer is extremely common in the
Western world. Rectal cancer accounts for 30 to
35 percent of these cancers. After treatment, the
emergence of local recurrence (LR) is a major set-
Correspondence to: Emmanuel A. Agaba, M.B.B.S., F.R.C.S. (Ed-
inb.), 6 Mallen Drive, Tividale, Birmingham, B69 1LX, United King-
dom, e-mail: eagaba@hotmail.com
Dis Colon Rectum 2004; 47: 291–296
DOI: 10.1007/s10350-003-0046-1
© The American Society of Colon and Rectal Surgeons
291
back for the patients and their surgeons. Although
the reported incidence of LR is variable, the estimated
rate in the present era of total mesorectal excision
(TME) is approximately 14 percent of all curative re-
sections compared with 30 percent in the pre-TME
era.1,2 Often, these recurrences are associated with
severely disabling symptoms and are difficult to treat.
In general, most recurrences become overt within two
years of surgery3 and in most cases, the outcome is
fatal.
Although the exact mechanism that led to the emer-
gence of recurrence is unknown, many surgeons be-
lieve that implantation of viable tumor cells that oc-
curred at the time of primary resection is one of the
possible mechanisms by which local recurrences oc-
cur.4–6
Several studies have shown that free malignant cells
are collected on circular stapling devices during an-
terior resection,7,8 and it is possible that implanta-
tion of these viable malignant cells occur during anas-
tomosis. Several studies have demonstrated that
these free malignant cells can be destroyed by effec-
tive rectal irrigation using cytotoxic agents, which
have been shown to be effective in in vivo and in
vitro studies.9–11 Theoretically, the prevention of im-
plantation of viable cells during surgery or their de-
struction using cancericidal agents such as Cetrimide
and povidone-iodine should prevent such recur-
rences. It is unclear whether this translates into a re-
duction in the incidence of local recurrence. This
study was designed to evaluate the effectiveness of
cytocidal rectal washout in reducing the incidence of
local recurrence.
292 AGABA
PATIENTS AND METHODS
In December 1999, the medical records of all pa-
tients who underwent anterior resection for biopsy-
proven adenocarcinoma of the rectum and rectosig-
moid diagnosed between January 1992 and
December 1994 in a high-volume, tertiary center were
examined. A total of 141 patients (mean age, 66
(range, 32–86) years; 87 males) underwent curative
oncologic resection with flush/high ligation of the in-
ferior mesenteric vessels and total mesorectal exci-
sion. In all patients, the resected segments were ex-
amined for proximal, distal, and radical resection
margins and tumor clearance. The apical nodes were
identified and examined histologically to determine
their involvement. For this study, a resection was con-
sidered curative if the preoperative chest x-ray and
computed tomography of the abdomen and liver did
not reveal any evidence of metastasis and also if at
operation, there was no peritoneal seedling and the
lateral and distal margins were histologically free of
tumors. Local recurrence was defined as any growth
in the pelvis or at the anastomotic site after the pri-
mary curative resection. None of the patients had pre-
operative chemoradiotherapy.
Three senior colorectal surgeons with identical
training, case volume, survival rates, and local recur-
rence rates performed these operations. Two of the
three surgeons routinely administer on-table rectal
washout using 1 percent Cetrimide, but the third sur-
geon does not.
In preparation, all patients were routinely prepared
with low residue diet and Picolax in the 24 hours
preceding the operation. On the morning of the op-
eration, all patients were offered preoperative phos-
phate enemas. At operation, the left colon was mobi-
lized and the splenic flexure taken down.
Rectal washout was performed in 90 patients. In
preparation for the rectal washout, a cross clamp was
placed distal to the tumor after adequate mobilization
of the rectum, and a soft bowel clamp was placed
proximally to the tumor to ensure that the segment
bearing the tumor was isolated. A 24-F gauge Foley
catheter was introduced per rectum for instillation of
cytocidal agent. The closed rectal stump was washed
using 500 ml of 1 percent Cetrimide until the effluents
ran clean. Once clean, the segment bearing the tumor
was excised between cross clamp. Intestinal continu-
ity was restored using circular stapling device in a
manner previously described.12,13 Anastomosis integ-
rity was tested by air insufflation test. Donuts were
Dis Colon Rectum, March 2004
checked for completeness and in all cases sent for
histologic confirmation of clear resection margin.
Among the no washout group, the left colon and
rectum were mobilized in the standard fashioned.
Cross clamp was placed distal to the tumor and fired.
The tumor was excised and anastomosis fashioned in
a similar manner as those of the washout group.
In all patients, the Dukes stage (Astler-Coller modi-
fication) was determined, and later, the American
Joint Committee on Cancer system was used in retro-
spect for most of the patients. Each patient was fol-
lowed up initially three-monthly for the first year, six-
monthly for second, and yearly afterward for a further
three years after which they were discharged to their
family practitioner. During each visit, the patients
were offered rigid sigmoidoscopy and any suspicious
area was biopsied. Blood was taken for carcinoem-
bryonic antigen and the liver function test. At the end
of the first and second year, all patients had colonos-
copy, chest radiograph, and a CT scan of the liver.
Elevated carcinoembryonic antigen beyond the con-
trol (control 艋 5 ng/l) was investigated by endoscopy
and/or CT scan of the abdomen and pelvis. Because
the rate of local recurrence decreases with increasing
distance from the anal verge, rectal cancers were ana-
lyzed separately from those of rectosigmoid.
Postoperative chemotherapy (5-fluorouracil) was
offered to selected group of patients with poorly dif-
ferentiated to anaplastic Dukes C disease. Some pa-
tients who met this criterion were not offered chemo-
therapy because of other comorbidities or frailty.
Statistical Analyses
The effectiveness of rectal washout was determined
by the chi-squared test. The survival curves were de-
termined using the Kaplan-Meier method, and these
were compared using log-rank regression. Statistical
analysis was performed using GraphPad® Prism sta-
tistics software (GraphPad, San Diego, CA).
RESULTS
The two groups were identical for age, gender,
Dukes stage, American Joint Committee on Cancer
(TNM) classification scheme, and degree of tumor dif-
ferentiation (Tables 1–3). There were 80 rectal and 61
rectosigmoid adenocarcinomas. The donuts were
complete and free of any residual tumors in all pa-
tients.
The local recurrence rate for all comers was 5 per-
Vol. 47, No. 3 RECTAL WASHOUT DURING ANTERIOR RESECTION 293

Table 1. Table 3.
Comparison of the Groups (All Comers) Subset of Reporting Variables in Patients With
Rectosigmoid Adenocarcinomas
Washout No Washout
Characteristics (n = 90) (n = 51) Washout No Washout
Characteristics (n = 50) (n = 11)
Male/female ratio 56/34 31/20
Median age (yr) 63 61 Male/female ratio 30/20 6/5
Dukes stage Dukes stage
A 17 12 A 17 2
B 51 21 B 23 6
C 22 18 C 10 3
Histologic grade Histologic grade
Well differentiated 46 29 Well differentiated 13 3
Moderately differentiated 30 15 Moderately differentiated 28 7
Poor to anaplastic 14 7 Poor to anaplastic 9 1
Distal resection margin (mm)
Median 39 46
Range 30–75 32–77

Table 2.
Characteristics of Patients with Rectal Cancer
Washout No Washout
Characteristics (n = 30) (n = 50)
Male/female ratio 24/6 30/20
Median age (yr) 66 64
Mean tumor level (cm)a 6.5 6.2
Postoperative adjuvant
chemotherapy 10 11
Distal resection margin (mm) 40 45
Modified Dukes stage
A 6 19 Figure 1. Survival curve: washout and no washout
B 14 18 groups.
C 10 13
AJCC stage
T1-3 24 46
T4 6 4 Of these recurrences, only four deaths (washout
N0 20 34 group = 2, no washout group = 2) were directly at-
N1 9 15
N2 1 1 tributable to carcinomatosis; the remaining deaths
M0 30 48 were from causes unrelated to cancers. The charac-
M1 0 2 teristics of those with recurrent disease are shown in
AJCC = American Joint Committee on Cancer. Tables 4 and 5. All patients with recurrent disease
a
Level of tumor measured from anal verge. were offered postoperative chemoradiotherapy start-
ing six weeks after salvage surgery. The radiation
dose ranges between 50 and 60 Gy fractions of 1.8 to
cent (n = 7), five-year disease-free survival was 81 2 Gy.
percent (3-year disease-free survival = 98.5 percent),
and five-year mortality rate was 19 percent (n = 27)
during a median follow-up of 63 months. Specifically, DISCUSSION
the five-year mortality rate for the rectum-only group
was 20 percent (n = 16, washout group = 10, no wash- This study was designed to evaluate the effective-
out group = 6) and five-year disease-free survival was ness of cytocidal agents used for rectal washout in
91 percent (Fig. 1). preventing local recurrence (LR). Several reports have
The local recurrence rate for the rectal-washout shown the presence of viable tumor cells in the wash-
group was 4.4 percent (n = 4) compared with 5.9 ing samples, but it is unclear whether implantation of
percent (n = 3) for the no washout group. This dif- these cells alone can explain the presence of local
ference was not statistically significant (P = 0.0653; 95 recurrence.
percent confident interval, 0.1264–1.066). In the series by McCall et al.,14 the pooled LR rate
294 AGABA Dis Colon Rectum, March 2004

Table 4.
Characteristics of Patients With Recurrent Disease (Recurrent Rectal Adenocarcinoma)
Characteristics Washout Group No Washout Group
No. of patients 4 3
Dukes stage
B 1 (Moderately differentiated) 1 (Moderately differentiated)
C 3 (Poorly differentiated) 2 (Poorly differentiated)
AJCC (TNM) Stage
T3 2 1
T4 2 2
N1 3 2
Mean tumor size (cm) 6 6.2
Mean distal resection margin (mm) 45 52
Distance from anal verge (cm) 8 6.5
Intraoperative spillage None None
Postoperative chemotherapy 3 2
AJCC = American Joint Committee on Cancer.

Table 5.
Comparison of Patients Who Recurred With Those Who Did Not Recur (Rectal Cancer Only)
Characteristics Local Recurrence Group No Recurrence Group
No. of patients 7 73
No. of positive or close radial margin None None
Mean distal resection margin (mm) 48 56
Mean no. of positive lymph node 16 10
Mean size of tumor (cm) 8 10
Intraoperative tumor spillage No No
Postoperative chemotherapy 5 16

for those series in which it was categorically stated
that cytocidal washout was used routinely was 12.2
percent. When separated according to total mesorec-
tal excision (TME) and extended pelvic lymphadenec-
tomy (EPL), the pooled LR rates for patients who had
cytocidal washout was 1 to 2 percent less than the
pooled LR for the entire group. This is in keeping with
the fact that true anastomotic recurrences are rare and
make up a minority of all local recurrences. The
group concluded that the true benefit of cytocidal
washout was small. Although our study population
was small, the indications are in support of this con-
clusion, because local recurrences between the two
groups were identical.
Malignant cell implantation has been suggested to
explain the mechanism by which local recurrences
occur. Implantation of malignant cells tends to occur
when the mucosal surface is damaged and rarely oc-
curs in undamaged cells. In an experimental study
conducted by Hubens et al.,5 implantation of malig-
nant cells occurred readily in rats with damaged mu-
cosal surface. In a separate study by McGregor et al.,4
the authors showed that braided sutures were capable
of entrapping and transferring large amount of free
intraluminal tumor cells in vivo. Although free tumor
cells are incapable of implanting on a normal colonic
mucosa, surgical wounds with exposed muscle edges
and serosa represent a fertile medium for the growth
of free malignant cells.
The fact that anastomotic recurrences are uncom-
mon and account for 5 to 15 percent of all local re-
currences suggests that other factors may be respon-
sible for LR. These factors may not render themselves
amenable to cytocidal washout.
Although the cause of LR after curative resection for
rectal cancer may be multifactorial, it is believed that
intraoperative spillage of tumor cells may be a signifi-
cant risk factor. Intraoperative spillage may occur
from open rectal stump (double purse-string tech-
nique) or when the trocar punctures the sealed rectal
stump.11 In two separate studies conducted by Zirn-
gibl et al.15 and Slanetz,16 the authors have shown
convincingly that intraoperative spillage of tumors
significantly increases the risk of LR. When these cells
are implanted, local recurrence may occur despite us-
ing complete mesorectal excision. We did not record
any intraoperative spillage during the study period.
The report from the study conducted by Umpleby
and Williamson 17 suggests that chlorhexidine-
cetrimide and povidone-iodine are lethal to tumor
Vol. 47, No. 3 RECTAL WASHOUT DURING ANTERIOR RESECTION
cells at wide range of concentration. In theory, irri-
gating the rectal stump with these agents should pre-
vent implantation of these cells and therefore local
recurrence. This hypothesis has not yet been formally
tested in man.18
In general, most local recurrences are extraluminal
and out of the reach of rectal washout. As noted by
Dukes19 in 1943, rectal cancers preferentially spread
laterally by direct extension and involvement of me-
sorectum than distally in the intramural plane.
Although the current practice of proximal and distal
examination of resected specimen may not be able to
predict all local recurrences, a clear distal margin does
not exclude radial spread. In general, tumor cells are
rarely found in the bowel wall for more than a centi-
meter beyond the distal end of the tumor20–22; radial
resection margin on the other hand often exhibit dis-
continuous spread.23 Recurrences bear a direct rela-
tion to the radial spread. In a study conducted by
Quirke et al.,24 using an examination technique in
which radial examination was routinely conducted,
85 percent of all local recurrences can be predicted.
For this reason, radial margins are of prognostic sig-
nificance. In the present series, all seven patients who
recurred had clear distal and radial margins. It is pos-
sible that the degree of tumor differentiation and the
Dukes stage may have contributed to this adverse out-
come.
Although not specifically addressed in the present
study, the level of tumor in the rectum seems to be of
prognostic significance. Among recurrences, all
shared a commonality of low rectal tumor, moderate
to poorly differentiated disease, large tumor size, and
a clear, distal resection margin. Among those with
node-positive disease, in most cases, the nodes were
completely replaced by tumor cells.
Our observation supports the reports by Kapiteijn
et al.3 and McDermott et al.25 In the report by McDer-
mott et al. in which the group examined the LR after
a curative resection for rectal cancer in a series of
1,008 patients, the group recorded the highest inci-
dence of LR among patients whose tumor level was
< 6 cm from the anal verge.
By applying the technique of total mesorectal ex-
cision, Heald et al. achieved a ten-year actuarial LR
rate of 4 percent in 200 consecutive patients under-
going curative anterior resection for rectal cancer.26
By applying the same principle of complete mesorec-
tal excision, we achieved a local recurrence rate of 5
percent and a five-year, disease-free survival of 81
percent for all comers who underwent curative ante-
295
rior. This is similar to other series in which standard-
ized, complete, mesorectal excision was per-
formed.27,28 Because of the discontinuous nature of
radial spread, it is possible that mesorectal disease is
not apparent to the surgeon at the time of initial op-
eration24 and therefore may account for the late de-
velopment of LR. For this reason, adequacy of distal
resection margin is not the same for the bowel wall
and the mesorectum.14
In the present study, most of our patients were fol-
lowed up for five years. We believe that this is ad-
equate because most recurrences tend to occur within
the first two to three years. In fact all recurrences
during this study occurred within the first three years,
thus justifying our current practice.
Three, dedicated, colorectal surgeons who have re-
ceived subspecialty training in colorectal surgery and
specifically in the art of complete mesorectal excision
performed all curative resection in our series. Al-
though the role of surgeon-to-surgeon variables, sub-
specialty training, and effect of case volume on the
long-term outcome of patient after rectal cancer sur-
gery are well documented,1,29–31 we believe that
given the wealth of experience among the three, ded-
icated, colorectal surgeons (60 years of experience
between them), the effect of -surgeon variation may
not have contributed significantly in our series.
Postoperative adjuvant chemotherapy was used in
28 patients. Of these, five patients recurred. It is pos-
sible that the use of chemotherapeutic agent in high-
risk patients may have prevented some recurrences
among this group of patients.
Among the recurrences in our series, 71 percent (n
= 5) were Dukes C compared with 29 percent (n = 2)
Dukes B. The incidence of LR increases with increas-
ing depth of invasion and lymph node involvement.
Thus, LR is directly related to the Dukes stage. Other
authors reported similar findings and generally a poor
outlook between Dukes B and C compared with
Dukes A.1,31
CONCLUSIONS
Because of the size of the study, we were unable to
demonstrate the benefit or lack thereof of cytocidal
agents in reducing local recurrence. We believe that
the time has come for a large-scale multicenter trial to
address this important question. Local recurrence is
directly related to the Dukes stage, tumor differentia-
tion, and the closeness of the tumor to the anal verge.
296 AGABA
REFERENCES
1. Phillips RK, Hittinger R, Blesovosky L, Fry JS, Fielding
LP. Local recurrence following “curative” surgery for
large bowel cancer 1: the overall picture. Br J Surg 1984;
71:12–6.
2. Enker WE. Total mesorectal excision–the new golden
standard of surgery for rectal cancer. Ann Med 1997;29:
127–33.
3. Kapiteijn E, Marijnen CA, Colenbrander AC, et al. Local
recurrence in patients with rectal cancer, diagnosed be-
tween 1988 and 1992: a population-based study in the
west Netherlands. Eur J Surg Oncol 1998;24:528–35.
4. McGregor JR, Galloway DJ, McCulloch P, George WD.
Anastomotic suture materials and implantation metas-
tasis: an experimental study. Br J Surg 1989;76:331–4.
5. Hubens J, Lafullarde T, Van-Marck E, Vermeulen P,
Hubens A. Implantation of colon cancers cells on intact
and damaged colon mucosa and serosa: an experimen-
tal study in rat. Acta Chir Belg 1994;94:258–62.
6. Tsunoda A, Shibusawa M, Kawamura M, Murakami M,
Kusano M. Recurrent colonic cancer developing at the
site of a stapled stump: report of case. Surg Today 1997;
27:457–9.
7. Gertsch P, Baer HU, Kraft R, Maddern GJ, Altermatt HJ.
Malignant cells are collected on circular staplers. Dis
Colon Rectum 1992;35:1108–9.
8. Umpleby HC, Fermor B, Symes MO, Williamson RC.
Viability of exfoliated colorectal carcinoma cells. Br J
Surg 1984;71:659–63.
9. Tsunoda A, Shibusawa M, Tsunoda Y, Choh H, Takata
M, Kusano M. Implantation on the suture material and
efficacy of povidone-iodine solution. Eur Surg Res 1997;
29:473–80.
10. Sayfan J, Averbuch F, Koltun L, Benjamin N. Effect of
rectal stump washout on the presence of free malignant
cells in the rectum during anterior resection for rectal
cancer. Dis Colon Rectum 2000;43:1710–2.
11. Jenner DC, de Boer WB, Clarke G, Levitt MD. Rectal
washout eliminates exfoliated malignant cells. Dis Co-
lon Rectum 1998;41:1432–4.
12. Yule AG, Fiddian RV. ‘Two gun’ surgery-low anterior
resection of the rectum avoiding anorectal purse string
suture. Br J Surg 1983;70:100–15.
13. Knight CD, Griffen FD. An improved technique for low
anterior resection of the rectum using the EEA stapler.
Surgery 1980;88:710–4.
14. McCall JL, Cox MR, Wattchow DA. Analysis of local
recurrence rates after surgery alone for rectal cancer. Int
J Colorect Dis 1995;10:126–32.
Dis Colon Rectum, March 2004
15. Zirngibl H, Husemann B, Hermanek P. Intraoperative
spillage of tumor cells in surgery for rectal cancer. Dis
Colon Rectum 1990;33:610–4.
16. Slanetz CA Jr. The effect of inadvertent intraoperative
perforation on survival and recurrence in colorectal
cancer. Dis Colon Rectum 1984;27:792–7.
17. Umpleby HC, Williamson RC. The efficacy of agents
employed to prevent anastomotic recurrence in colo-
rectal carcinoma. Ann R Coll Surg Engl 1984;66:192–4.
18. Umpleby HC, Williamson RC. Anastomosis recurrence
in large bowel cancer. Br J Surg 1987;74:873–8.
19. Dukes C. The surgical pathology of rectal cancer. Proc
R Soc Med 1943;37:131.
20. Pollett WG, Nicholls RJ. The relationship between the
extent of distal clearance and survival and local recur-
rence rates after curative anterior resection for carci-
noma of the rectum. Ann Surg 1983;198:159–63.
21. Karanjia ND, Schache DJ, North WR, Heald RJ. ‘Close
Shave’ in anterior resection. Br J Surg 1990;77:510–2.
22. Phillips RK. Adequate distal margin of resection for ad-
enocarcinoma of the rectum. World J Surg 1992;16:
463–6.
23. Heald RJ, Husband EM, Ryall RD. The mesorectum in
rectal cancer surgery-the clue to pelvic recurrence? Br J
Surg 1982;69:613–6.
24. Quirke P, Durdey P, Dixon MF, Williams NS. Local re-
currence of rectal adenocarcinoma due to inadequate
surgical resection. Histopathological study of lateral tu-
mour spread and surgical excision. Lancet 1986;2:
996–9.
25. McDermott FT, Hughes ES, Pihl E, Johnson WR, Price
AB. Local recurrence after potentially curative resection
for rectal cancer in a series of 1008 patients. Br J Surg
1985;72:34–7.
26. MacFarlane JK, Ryall RD, Heald RJ. Mesorectal excision
for rectal cancer. Lancet 1993;341:457–60.
27. Heald RJ, Karanjia ND. Results of radical surgery for
rectal cancer. World J Surg 1992;16:848–57.
28. Editorial. Breaching the mesorectum. Lancet 1990;335:
1067–8.
29. Fielding LP, Stewart-Brown S, Dudley HA. Surgeon-
related variables and the clinical trial. Lancet 1987;2:
778–9.
30. McArdle CS, Hole D. Impact of variability among sur-
geons on postoperative morbidity and mortality and
ultimate survival. BMJ 1991;302:1501–5.
31. Pilipshen SJ, Heilweil M, Quan SH, Sternberg SS, Enker
WE. Patterns of pelvic recurrence following definitive
resections for rectal cancer. Cancer 1984;53:1354–62.