Mean ± SD anal sphincter resting pressure (A) and resting vector volume (B) in patients with midrectal cancer (group 2A) and low rectal cancer (group 2B) before and after neoadjuvant chemoradiation with total mesorectal excision. Asterisk indicates P<.05 by paired t test.
Ammann K, Kirchmayr W, Klaus A, Mühlmann G, Kafka R, Oberwalder M, De Vries A, Öfner D, Weiss H. Impact of Neoadjuvant Chemoradiation on Anal Sphincter Function in Patients With Carcinoma of the Midrectum and Low Rectum. Arch Surg. 2003;138(3):257-261. doi:10.1001/archsurg.138.3.257
Preoperative chemoradiation for patients with stage II and III midrectal and low rectal cancer may improve survival and decrease local recurrence rate. We evaluated the long-term impact of neoadjuvant chemoradiation on anal sphincter function.
Tertiary referral center.
From March 1, 1996, to January 31, 2002, 50 patients with midrectal and low rectal cancer who underwent total mesorectal excision were prospectively enrolled.
Patients received either surgical therapy alone (group 1, n = 22) or preoperative, combined chemoradiation (group 2, n = 28). Group 2 was divided into patients with midrectal (group 2A, n = 14) and low rectal (group 2B, n = 14) cancer. Anorectal manometry was performed preoperatively and a median of 384 days postoperatively.
Main Outcome Measures
Anal resting pressure, squeeze pressure, anal sphincter vector volumes, length of the high-pressure zone, sensory threshold of the pouch, and rectal capacity.
Preoperative manometric values were comparable between the groups. No statistically significant manometric differences occurred in group 1 postoperatively. Mean resting pressure (preoperative and postoperative, respectively: 89 ± 35 mm Hg, 53 ± 17 mm Hg), resting vector volume (605 ± 324 cm3, 142 ± 88 cm3), and maximal tolerable volume (144 ± 29 mL, 82 ± 44 mL) decreased significantly in chemoradiated patients postoperatively (P<.05). Manometric values of group 2B patients remained stable postoperatively, while mean resting pressure (73 ± 22 mm Hg vs 52 ± 14 mm Hg) and resting vector volume (631 ± 288 cm3 vs 145 ± 78 cm3) decreased significantly in group 2A patients (P<.001).
Total mesorectal excision does not influence anal sphincter function during long-term follow-up. Neoadjuvant chemoradiation results in disordered anal sphincter function in patients with midrectal cancer. Low and rectoanal anastomosis seems to obtain better anal sphincter function than higher anastomosis in chemoradiated patients.
RADIATION FOR the treatment of carcinoma of the pelvic region has been found to have a significant impact on anal sphincter function.1 Effects of irradiation on anorectal function are dose dependent. Although the total radiation dose is less in patients with rectal carcinoma than in patients treated for cervical or anal carcinoma, impairment of the anal sphincter has been described.2 In addition, operative trauma occurring with low anterior resection of the rectum may lead to further functional disorder. The introduction of nerve-sparing total mesorectal excision (TME) by MacFarlane et al3 resulted in a decreased incidence of complications due to nerval injury, such as urinary incontinence and impotence. However, quality of life after TME may be impaired by disturbed anal continence and fecal urgency. Neoadjuvant chemoradiation in patients with advanced rectal cancer has proven to show oncologic benefit4 but may have an impact on the functional entity of the anal sphincter. Recently, Gervaz et al5 published a short-term follow-up study of anal sphincter function after chemoradiation for rectal cancer. In their series, resting pressure of the anal sphincter was reduced after chemoradiation within the postoperative observation period of 8 to 12 weeks. As a conclusion, adequate shielding of the anal sphincter was therefore recommended. However, in another study published by Birnbaum et al,6 anorectal function was not altered after pelvic irradiation. In these patients, the mean distance of the lower edge of tumors from the dentate line was 7.4 cm. The surgical trauma itself was considered to account for any functional sphincter defects.
This study was performed to identify the long-term impact of chemotherapy, irradiation, and surgery on anal sphincter function in patients who underwent low anterior resection for low and midrectal cancer.
Patients with adenocarcinoma of the low and midrectum who underwent low anterior resection between March 1, 1996, and January 31, 2002, were enrolled in this study after giving informed consent. Collected data were entered into a prospective database. Preoperatively, rigid proctoscopy was performed to assess the distance of the distal edge of the cancer from the anal verge. In addition, abdominal computed tomography and endorectal ultrasound were used to determine tumor infiltration and lymph node status of the disease. Tumors with distal edges between 6 and 10 cm from the dentate line were defined as midrectal cancer, and tumors with distal edges within 6 cm from the dentate line were defined as low rectal cancer. Patients presenting with distal tumor edges closer than 3 cm from the dentate line were excluded from this study because they underwent rectal amputation.
The operative procedure was standardized according to the recommendations for nerve-sparing TME as described elsewhere.3 For reservoir reconstruction, a 5- to 7-cm colonic J-pouch was hand sewn and a staple-line anastomosis was performed. A 31-mm circular stapler (CEEA circular stapler; Tyco, Norwalk, Conn) was used in all of the patients. At the end of the procedure, the level of the staple-line anastomosis was documented routinely. None of the patients included in the study had hand-sewn coloanal anastomoses. Diverting loop ileostomy was established for a period of 4 to 6 weeks.
Patients with locally advanced rectal cancer, staged as uT3/uT4 (u indicates staging by means of endorectal ultrasound and/or computed tomography), underwent neoadjuvant chemoradiation according to a pilot study protocol of the Tyrolean Oncology Working Group. Each patient received preoperative, combined chemoradiation and a total radiation dose between 4400 and 4500 rad, at a single dose of either 110 rad administered twice a day or 180 rad once a day. Radiation fields in a 3-field technique (posteroanterior and 2 lateral fields)7 included the rectal canal and adjacent lymph nodes. In all of the study patients, the rectal sphincter was not included in the radiation fields. In conjunction with the radiotherapy, fluorouracil in a dose of 350 mg/m2 (5-Fluorouracil "Ebewe"; EBEWE Pharma GmbH Nfg KG, Unterach, Austria) was administered continuously through an implanted central venous catheter (Port-A-Cath Deltec CADD-1 system; Deltec, Inc, St Paul, Minn) on each treatment day.
Characteristics of patients who solely underwent low anterior resection (group 1) and patients who received neoadjuvant chemoradiation and low anterior resection (group 2) are given in Table 1.
Each combined chemoradiation treatment started on Monday and was interrupted on weekends. After 4 weeks of treatment, all patients were allowed a therapy-free recovery interval of 2 weeks and subsequently scheduled for surgery in the following week. Anorectal manometry was performed preoperatively and postoperatively as previously described.5,8 In brief, a stationary and continuous pull-through technique with a water-perfused catheter (Synectics Medical Inc, Irving, Tex) was used to assess anal resting pressure, squeeze pressure, and anal sphincter vector volumes. In addition, the length of the high-pressure zone and the rectoanal inhibitory reflex were determined. Sensory threshold of the neorectum was defined as the first appreciation of balloon filling. The filling volume that led to the urge for defecation gave the rectal capacity. Furthermore, the maximal tolerable volume was assessed.
Normal values of the anorectal manometry setting used in this study were obtained in healthy volunteers and yielded 50 to 100 mm Hg for resting pressure, 100 to 180 mm Hg for squeeze pressure, 2 to 3 cm for high-pressure zone length, and present rectoanal inhibitory reflex. Normal balloon filling volumes reached 20 to 60 cm3 for sensory threshold, 110 to 200 cm3 for rectal capacity, and 140 to 300 cm3 for maximal tolerable volume.
Quantitative data were expressed as mean ± SD. The range and interquartile range are reported. The Mann-Whitney test and the t test were used to compare preoperative and postoperative continuous data. Comparisons between different groups were calculated by means of the Fisher exact test. A P value of less than .05 was considered statistically significant (SPSS for Windows; SPSS Inc, Chicago, Ill).
Fifty patients who underwent TME gave informed consent to participate in this study. Twenty-two patients underwent surgery alone (group 1), while 28 patients additionally were treated with neoadjuvant chemoradiation (group 2). Table 1 presents the demographic data of patients in groups 1 and 2. There was no statistically significant difference in terms of age, tumor infiltration, or tumor stage between the 2 groups (Fisher exact test). The female-male ratio was 1.0 in group 1 and 0.3 in group 2 (P = .08), respectively. All patients in group 2 underwent chemoradiation preoperatively. Intermittent diarrhea with perianal skin irritation was found to bother patients in group 2, although symptomatic therapy such as topical corticosteroids could control these symptoms. None of the study patients had grade 4 toxic effects related to the chemoradiation treatment. All patients in this group completed the neoadjuvant therapy regimen.
Nine patients in group 1 (41%) and 14 patients in group 2 (50%) had low rectal cancers. Preoperative chemoradiation in combination with a diverting ileostomy did not result in anastomotic leakage. No surgical revisions were necessary, and the postoperative hospital stay (median, 10 days) was not influenced by the neoadjuvant therapy strategy. However, in contrast to the initial staging (uTNM) based on computed tomography and endorectal ultrasound, postoperative ypTNM staging of cancers showed a decrease in tumor infiltration in 14 patients in group 2. (The yp indicates the final staging after neoadjuvant chemoradiation [y] and postoperative histopathologic diagnosis [p].)
Manometry was performed 2 days before the beginning of the therapy, which was surgery in group 1 and chemoradiation in group 2. Repeated manometry was carried out a median of 384 days (interquartile range, 149-405 days) postoperatively. The manometric characteristics of group 1 and group 2 before and after treatment are given in Table 2. The differences in the postoperative interval did not statistically influence the results of manometric data. Length of the sphincter high-pressure zone did not vary between groups and between preoperative and postoperative measurements. After the operative procedure, anal sphincter measures as well as sensory and capacity measurements remained unaffected in group 1 compared with preoperative values. In contrast, mean resting pressure and resting vector volume decreased significantly in the chemoradiation group postoperatively (P = .03 and P = .003, respectively).
Resection of a low rectal cancer always included total excision of the rectum with coloanal anastomosis, whereas resection of a midrectal cancer resulted in a colorectal anastomosis preserving a small cuff of the rectum. To evaluate the influence of the tumor level, corresponding adjustments of the irradiation field, and level of resection on functional sphincter impairment, patients with neoadjuvant chemoradiation (group 2) were further grouped as having midrectal cancer (group 2A, n = 14) and low rectal cancer (group 2B, n = 14). Patients' demographics are given in Table 3. Female-male ratio was 0.56 and 0.17 in groups 2A and 2B, respectively. The mean age was comparable in both subgroups. Patients with cancer in the midrectum exhibited a tendency toward less tumor infiltration. Whereas all preoperative manometric values were comparable between groups 2A and 2B, postoperative manometry showed a significant impairment of resting pressure and resting vector volume in patients operated on for midrectal cancer (Figure 1). Preoperative mean ± SD values for resting pressure were 93 ± 20 mm Hg and 73 ± 22 mm Hg in groups 2A and 2B, respectively. After surgery, these values decreased significantly in group 2A (52 ± 14 mm Hg; P = .009) but remained statistically stable in group 2B (53 ± 21 mm Hg; P = .09). These results were paralleled by a significant decrease in resting vector volume in group 2A patients (631 ± 288 cm3 and 145 ± 78 cm3 preoperatively vs postoperatively; P = .002), whereas values of resting vector volume did not change in group 2B after surgery (393 ± 369 cm3 and 133 ± 106 cm3 preoperatively vs postoperatively; P = .1).
In addition, sensory threshold measurements showed significant differences between preoperative and postoperative values only in group 2A patients (57 ± 19 mL vs 41 ± 14 mL; P = .047). Maximal tolerable volume was significantly reduced in group 2A after surgery (129 ± 42 mL and 79 ± 36 mL preoperatively vs postoperatively; P = .01) but did not show a significant difference in group 2B (138 ± 29 mL and 100 ± 54 mL preoperatively vs postoperatively; P = .3). Squeeze pressure, squeeze vector volume, length of high-pressure zone, and capacity measurements did not show significant alterations in both subgroups postoperatively.
The anorectum is a functional entity that may be influenced by local inflammatory and malignant processes. Anal continence can be maintained by the anal sphincter even if the rectal reservoir has been removed, since it is a separate anatomic and physiological entity that is supported by the skeletal musculature of the pelvic floor.9 In past years, sphincter-preserving operative procedures have replaced abdominoperineal resection and setting of a permanent stoma in patients with midrectal and low rectal cancer as long as an adequate distal margin of resection of 2 cm can be achieved and the mesorectum is totally excised.10 In these patients, the descending colon may be anastomosed to the anal canal itself, thereby creating a new rectal reservoir. However, several studies have assumed that when the level of the anastomosis becomes less than 3 to 4 cm to the dentate line, the patient's quality of life decreases in terms of bowel function, urgency, and leakage.11- 14
MacFarlane and colleagues3 popularized the TME that is characterized by sharp excision of the entire mesorectum during rectum resection, leading to reduced complications such as impotence and urinary dysfunction as a result of identification and preservation of the autonomic nerves. This surgical technique enables reduction of the local recurrence rate and synchronously obtains good functional results. The current study demonstrates that TME without preoperative chemoradiation did not statistically influence anal sphincter function in a long-term follow-up.
It is currently recommended in patients with stage II and III rectal cancer to convert to a sphincter-preserving operation using preoperative combined chemoradiation therapy, since it has been shown to reduce the risk of local recurrence and to increase survival.15,16 In this study, only patients with locally advanced rectal cancer (uT3/uT4) were enrolled in the neoadjuvant chemoradiation protocol. It is notable that postoperative histologic assessment showed downstaging of rectal tumors after chemoradiation, resulting in ypT0 to ypT2 staging. This is in accordance with observations published by Valentini et al.4
It has been recognized that preoperative and postoperative chemoradiation may adversely influence the pouch function after low anterior resection. Kollmorgen et al17 reported in a study from the Mayo Clinic that postoperative radiation therapy led to significantly increased bowel movements and incontinence. The authors speculated that radiation-induced fibrosis of the neorectum accounts for poor functional outcome. This observation has gained increasing popularity in recent years; however, the anal sphincter may be affected directly when the anal canal is included in the irradiation field. A study by DaSilva et al18 confirmed this pathological mechanism, since chemoradiation was found to cause internal sphincter fibrosis. Damage to both the sphincter muscle and the neorectal reservoir may accelerate the negative influence of TME on postoperative incontinence in patients with rectal cancer.
Previous obstetric injury and vaginal delivery are important issues, since the anal sphincter may have been damaged and present pathologic findings on anomanometry. Routine preoperative and postoperative ultrasound of the sphincter system was not part of the study protocol. Chart review in female patients with a mean age of 65 years did not reliably refer to this topic. However, we assessed radial sphincter vector volume symmetry in all female patients. We did not find abnormal preoperative and postoperative manometric values or radial asymmetry in the 2 female patients in group 2B (low rectal cancer). On the other hand, manometry in women in group 2A (midrectal cancer) showed no radial sphincter vector volume asymmetry despite significant impairment of postoperative resting pressure values. This and the male preponderance in group 2 strongly suggest that obstetric injury with segmental sphincter damage did not account for the statistical significance.
The present study shows that TME without chemoradiation therapy does not lead to statistically significant alterations in manometric characteristics of the anal sphincter. However, when neoadjuvant chemoradiation is applied, the functional outcome of patients after TME is adversely affected. Postoperatively, the mean resting pressure and mean resting vector volume are significantly decreased in patients with adjuvant chemoradiation. This is in accordance with the findings of Gervaz et al,5 who reported a higher incidence of incontinence for liquid and solid stools determined by means of a standardized questionnaire in irradiated patients. It may still be speculated that the adaptation in maximal tolerable volume could favor impaired continence. However, Gervaz et al showed that time-dependent progressive improvement of continence did not occur in patients who received either neoadjuvant or adjuvant pelvic irradiation. In the present study, patients' anomanometry findings confirmed impaired anal sphincter function in long-term follow-up.
Short-term studies of anal sphincter function have demonstrated a 24% decrease in resting pressure after abdominoperineal resection and coloanal anastomosis. The application of adjuvant chemoradiation resulted in a further decrease of the mean resting pressure of 27% in these patients.19 Adequate shielding of the anal sphincter was therefore recommended in patients with rectal cancer whenever a sphincter-preserving procedure is considered. The particular experimental setting of our study was designed to prevent misleading results of functional measurements due to the acute concomitant inflammatory reaction of the rectoanal region soon after chemoradiation. As the operative procedure was performed in week 3 after the end of the chemoradiation therapy, it was not possible to evaluate the pure impact of chemoradiation on rectoanal function.
The results we obtained indicate a different pathophysiological process when the group of irradiated patients was subdivided into midrectal and low rectal cancer groups. Patients with low rectal cancer with coloanal anastomosis did not show significant impairment of the anal sphincter. In contrast, irradiated patients with midrectal cancer with colonic anastomosis to a rectal remnant had statistically significant dysfunction of the internal anal sphincter. Although preoperative anal sphincter resting measurements seemed to be impaired in patients with low rectal cancer, statistical comparison did not demonstrate a significant difference between the 2 groups. Our data indicate that chemoradiation in combination with coloanal anastomosis was superior to irradiation with colorectal anastomosis in regard to internal anal sphincter function.
Although shielding of the anal sphincter was considered imperative in neoadjuvant chemoradiation, it seems to be more important to carry out a deep anastomosis after rectal excision to prevent fibrosis of the rectal and perirectal stump, even in patients with midrectal cancer.
In conclusion, we have shown that TME provides good functional anal sphincter results in patients with midrectal and low rectal cancer during long-term follow-up. However, neodjuvant chemoradiation may lead to impaired sphincter function as long as a rectal stump remains for anastomosing the colonic J-pouch.
Accepted for publication November 3, 2002.
Corresponding author and reprints: Helmut Weiss, MD, PhD, Department of General Surgery, University Hospital, Anichstrasse 35, A-6020 Innsbruck, Austria (e-mail: email@example.com).