levator ani muscles caused by operative procedure in anal sphincter-preserving operation for rectal cancer

The American Journal of Surgery (2011) 201, 508 –513

Clinical Science

Damage to anal sphincter/levator ani muscles caused by operative procedure in anal sphincter-preserving operation for rectal cancer Atsushi Hirano, M.D., Keiji Koda, M.D., Ph.D.*, Chihiro Kosugi, M.D., Ph.D., Masato Yamazaki, M.D., Ph.D., Hideki Yasuda, M.D., Ph.D. Department of Surgery, Teikyo University, Chiba Medical Center, 3426-3 Anesaki, Ichihara City, Chiba 299-0111, Japan KEYWORDS: Three-dimensional vector manometry; Rectal cancer; Anal sphincterpreserving operation; Defecatory disorder; Anterior resection

Abstract BACKGROUND: Details of postoperative damage to anal sphincter tonus following sphincterpreserving operation for rectal cancer remain unclear. METHODS: Postoperative anal tonus was measured using 3-dimensional (3D) vector manometry in 56 patients. Anal length with pressure from any direction was defined as total length (TL). Length with circular pressure (LCP), which is only measurable using 3D manometry, was also evaluated. RESULTS: In operations associated with low anastomosis, both TL and LCP at rest were significantly shortened when compared with control (high interior resection [HAR]). In particular, degraded LCP at rest was obvious. Anal lengths in squeezing state were preserved except in cases with intersphincteric resection (ISR). Postoperative incontinence score inversely correlated with functional anal length at rest. CONCLUSIONS: Although the sphincter muscles are mechanically preserved, function of the internal sphincter and subsequent defecatory function can be degraded in cases with operative procedures including surgical maneuvers at the pelvic floor. © 2011 Elsevier Inc. All rights reserved.

Anal sphincter-preserving operation (SPO) for the surgical treatment of rectal cancer has been widely performed with fair local control and postoperative survival rates.1,2 Intersphincteric resection (ISR) has been proposed and widely executed for the last decade for patients with very low rectal cancer, which formerly was a definite indication for abdominoperineal resection (APR) with permanent colostomy.3,4 With SPO becoming a common procedure, postoperative defecatory disorders of varying degrees have become a major problem requiring attention.5 One potential * Corresponding author. Tel.: ⫹81 436 62 1211; fax: ⫹81 436 61 3961. E-mail address: [email protected] Manuscript received August 7, 2009; revised manuscript December 11, 2009

0002-9610/$ - see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2009.12.016

cause for such postoperative defecatory malfunctions is anal sphincter damage caused by surgery. Sphincter tonus,6 anal sensation,7 and recto-anal reflex8 have been reported to be degraded postoperatively following SPO. However, characteristic findings of sphincter damage according to types of SPO procedure have not yet been clearly delineated. Recently, 3-dimensional vector manometry (3D manometry) has been used for the evaluation of physiologic luminal pressure in the lower esophagus9 or anal sphincter.10,11 To date, postoperative damage to the anal sphincter following SPO for rectal cancer has not been evaluated extensively with 3D manometry. The current study examined postoperative sphincter status using 3D manometry in patients who underwent several types of SPO to clarify specific findings associated with surgical procedures.

A. Hirano et al.

Sphincter damage following rectal surgery

509

Materials and Methods Patients

Figure 1 Anal canal lengths defined by pressure profiling in each channel. (A) Chart of pressure distributions in channels 1– 8 and definitions of anal canal length. TL ⫽ total length; LCP ⫽ length with circular pressure. (B) Three-dimensional construction of 8-channel pressure profiling.

Table 1

Subjects comprised 56 patients (37 men, 19 women) who underwent SPO at either Teikyo University Chiba Medical Center or Chiba University Hospital between January 2005 and December 2007. High anterior resection (HAR), in which anastomotic height was above the peritoneal reflection, was performed in 14 patients. Low anterior resection (LAR), using tumor-specific mesorectal excision (TSME) with anastomotic height below the peritoneal reflection, was performed in 20 patients. Very low anterior resection (vLAR), using complete total mesorectal excision (TME) with the anastomotic line at or below the levator plane was applied in 12 patients. In these cases, a circular stapler was used for anastomosis. In the remaining 10 patients, either partial or total ISR with a trans-anal hand-sewn colo-anal anastomosis was performed. Median age was 63 years (range 52– 81 years). The interval between initial surgery or the time at which the covering stoma was closed and the time of manometric study was 1 year in 37 cases, 2 years in

Length of the anal canal and LCP/TL ratio at resting state defined by 3D manometry HAR (n ⫽ 14)

LAR (n ⫽ 20)

vLAR (n ⫽ 12)

ISR (n ⫽ 10)

TL

50 mm (41–58)

45 mm (23–58)

46 mm (33–54)

38 mm (29–44)

LCP

40 mm (31–49)

30 mm (6–42)

30 mm (18–44)

12 mm (0–26)

LCP/TL ratio

84.4% (63.3–96)

66.4% (26.1–81.3)

66.3% (42.9–91.7)

32.5% (0–89.7)

Data are expressed as median (range). *P ⬍.05, †P ⬍.01 (Mann-Whitney U test).

510

The American Journal of Surgery, Vol 201, No 4, April 2011 plication of 110 mL of glycerine enema. Both resting and squeeze pressures were recorded using the pull-through method with the patient in a left-lateral posture. A perfusion catheter, 5-mm thick, with 8 independent thin lumens, was specifically manufactured (Asahi Biomed, Yokohama, Japan). At 2 cm from the tip, 8 drainage holes were situated radially at 45° intervals. All lumens were independently connected to pressure transducers, then to a capillary water perfusion system with .5 mL/min per channel of water at a pressure of 150 mm Hg. The transducers were connected to an amplifier, then data were transferred to a personal computer for analysis using Polygram software (Japan Medtronic, Kawasaki, Japan). Before obtaining data, water pressure at the height of the anus was set to 0 cm, and 68 cm above this zero point was adjusted to 50 mm Hg. For the measurement of sphincter pressure, the catheter inserted inside the neorectum was manually pulled from 6 cm above the anal verge at 1 cm/s in harmonization with a signal sound from the computer. When squeeze pressure was measured, patients were asked to squeeze maximally for approximately 10 seconds. All measurements were repeated ⱖ5 times until stable data were obtained to minimize both inter- and intra-observer variability. Written informed consent was obtained from all patients before the examination. All study protocols were approved by the University Ethics Committee at Teikyo University Chiba Medical Center.

Evaluation of data

Figure 2 (A) Example of fair TL (46 mm), LCP (38 mm), and LCP/TL ratio (82.6%) in a patient who received HAR. No deformity was apparent in 3D reconstruction of tonus distribution, suggesting that the internal sphincter muscle was preserved undamaged. (B) Example of degraded LCP (18 mm) and LCP/TL ratio (36%) in a patient who underwent LAR. Although total length was intact, internal sphincter muscle in some direction at the proximal site was partly degraded.

7 cases, and ⱖ3 years in 12 cases. Among patients who received either ISR or vLAR, preoperative chemoradiotherapy was administered to 7 patients, and colo-J pouch reconstruction was performed in 4 patients.

3D manometry All 3D manometry was performed by 3 researchers (A.H., C.K., and K.K.) with sufficient experience in conventional single-channel manometry. Just before examination, patients were asked to defecate completely after ap-

Figure 1 summarizes the definitions of several anal canal lengths measured by 3D manometry. Under recording with the pull-through method from 6 cm above the anal verge, total length of the anal canal (TL) was defined as the length from the point at which any 1 channel exceeds the rectal baseline pressure to the point at which all channels fall to baseline. TL presumably represents the longest length of the anal canal measurable by manometry, and may well reflect the conventional functional anal canal length as measured by single-channel manometry. Length of circular pressure (LCP) was defined as the length from the point at which all 8 channels exceed the rectal baseline to the point at which any channel falls to baseline. LCP reflects the length receiving pressure from 360° of the anal canal, and is only measurable by 3D manometry since the length is determined based on pressures from all 8 channels. In addition, we also evaluated LCP/TL ratio in accordance with the surgical procedure for SPO. LCP/TL ratio may reflect some semicircular damage to the sphincter muscle, since LCP can only be defined from the length receiving circular pressure, whereas TL is measurable from areas receiving semicircular pressure.

Evaluation of defecatory status All patients were interviewed regarding recent defecatory status using a self-administered questionnaire within 1 month

A. Hirano et al. Table 2

Sphincter damage following rectal surgery

511

Length of the anal canal and LCP/TL ratio at squeezing state defined by 3D manometry HAR (n ⫽ 14)

LAR (n ⫽ 20)

vLAR (n ⫽ 12)

ISR (n ⫽ 10)

TL

52 mm (40–58)

49 mm (34–69)

48 mm (40–62)

43 mm (32–47)

LCP

43 mm (30–50)

41 mm (20–57)

38 mm (30–58)

25 mm (20–32)

LCP/TL ratio

81.8% (71.7–88)

78.8% (50–96.2)

80.7% (75–93.6)

56.8% (50–81.3)

Data are expressed as median (range). *P ⬍.05, †P ⬍.01 (Mann-Whitney U test).

before or after each examination. Fecal status was categorized based on criteria proposed by Wexner et al,12 taking into account the degree and frequency of incontinence. The Wexner incontinence scale score (Wexner score) measures 5 types of incontinence: solid, liquid, gas, wearing a pad, and lifestyle alteration. Each category was scored from 0 (never) to 4 (⬎1 episode/d). Perfect continence was thus scored as 0, whereas complete incontinence was scored as 20.

Statistical analysis TL, LCP, and LCP/TL ratio were compared among several SPOs, namely, HAR, LAR, vLAR, and ISR, to examine damage to the anal sphincter/levator ani muscles associated with various surgical procedures. Nonparametric testing (Mann–Whitney U test) was used to compare data between surgical procedures. Spearman’s rank correlation coefficient (␳) was calculated when correlations between 2 variables were analyzed. Values of 2-sided P ⬍ .05 were considered statistically significant.

Results

vLAR, and ISR, compared with HAR. In addition, TL at rest was significantly shorter in ISR than in all other SPOs. Median LCP at rest was 40 mm (range 31– 49 mm) in HAR. As with TL, LCP was significantly shorter in LAR, vLAR, and ISR, when compared with HAR. LCP was also significantly shorter in ISR than in all other SPOs. LCP/TL ratio at rest was significantly smaller in LAR, vLAR, and ISR than in HAR. LCP/TL ratio at rest fluctuated widely as the anastomotic line became more distal, differing significantly between individual patients. LCP/TL ratio was significantly smaller in ISR than in other SPOs. Figure 2 shows typical examples of fair and degraded anal sphincter tonus and length following SPO.

Squeezing state As shown in Table 2, both TL and LCP during squeezing were significantly shorter in ISR than in other SPOs for rectal cancer. However, no significant differences in TL or LCP during squeezing were apparent between HAR, LAR, and vLAR. LCP/TL ratio during squeezing was also significantly lower in ISR than in other SPOs, whereas no significant difference was seen among HAR, LAR, or vLAR.

Resting state Defecatory function and manometric data TL, LCP, and LCP/TL ratio at rest for the various SPOs examined are summarized in Table 1. Median TL at rest was 50 mm in HAR (range 41–58 mm). TL at rest showed wider ranges following LAR (23–58 mm), vLAR (33–54 mm), and ISR (29 – 44 mm), and was significantly shorter in LAR,

Both TL and LCP correlated significantly (P ⬍ .01) with Wexner score (Figure 3). Coefficient ␳ with Wexner score was ⫺.444 for TL and ⫺.607 for LCP, showing that continence worsened with reduced functional anal canal length.

512

Figure 3 Correlation between Wexner score with (A) total length of anal canal and (B) length with circular pressure according to the operation methods. Significant inverse correlations (P ⬍ .01) were noted (Spearman’s rank correlation coefficient (␳) was ⫺.444 for A and ⫺.607 for B).

No significant correlation was seen between Wexner score and preoperative chemoradiotherapy (n ⫽ 7) or colo-J pouch reconstruction (n ⫽ 4) among patients who received ISR or vLAR. Furthermore, anastomotic leakage (n ⫽ 2) was not associated with function in this series.

Comments Causes of defecatory malfunction seen postoperatively in patients who have undergone sphincter-saving surgery for rectal cancer has been studied for the last several decades. In particular, manometric studies have revealed that anal resting pressure and maximum squeeze pressure are both degraded following surgery,6 with reduced anal sensation7 or reduced physiologic recto-anal inhibitory reflex.8 These sphincter malfunctions may be caused by direct injury dur-

The American Journal of Surgery, Vol 201, No 4, April 2011 ing surgery13 or may be a result of denervation to the internal sphincter,14 which is controlled by the hypogastric nerve.15 Damage to the anal sphincter is one of the major backgrounds of postoperative defecatory disorder following SPO, and such damage may not recover with time,16 suggesting that further studies are needed regarding postoperative sphincter malfunction following SPO for rectal cancer. The technique of 3D manometry has recently been used for the evaluation of patients with anal incontinence11 or postoperative anal damage as seen in Hirschsprung’s disease.10 The present study examined anal sphincter tonus using 8-channel 3D manometry in patients who had undergone several types of SPO for rectal cancer. As for controls, we used the patterns of patients who had undergone HAR, since preliminary data obtained from normal individuals suggested that the pattern with HAR resembles that in normal individuals in many cases, with occasional exceptions (data not shown). For the evaluation of data, we focused on 2 different types of anal canal length, namely, TL and LCP. TL is the longest length of functional anal canal as defined by manometry, and can be measured by conventional single-channel manometry. LCP is the length of functional anal canal receiving pressure from 360°. LCP is a unique length that is only measurable by 3D manometry. In case either pressure disproportion or physical deformity should occur inside the anal canal postoperatively, TL may be preserved to the extent that sphincter/levator ani muscles were preserved uninjured. Meanwhile, LCP may be influenced more strongly by physical deformity of the anal canal or by pressure disproportion caused by scarring or denervation,8 since pressure disorder in even one direction reduces LCP. Under this hypothesis, we examined LCP/TL ratio for the detection of minor damage to sphincter tonus. Data indicated that both TL and LCP in the resting state were shortened after LAR, vLAR, and ISR when compared with HAR. This suggests that surgical maneuvers at the pelvic floor as executed in LAR, vLAR, and ISR, but not in HAR, damage the tonus of internal sphincter muscles. In ISR, both TL and LCP were significantly shortened compared with other SPOs, probably because internal sphincter muscle is removed to various degrees in ISR. LCP/TL ratio at rest was also significantly decreased in LAR, vLAR, and ISR when compared with HAR, indicating that the anal length with 360° pressure is more damaged than length with pressure from at least one direction (TL). The range of LCP/TL ratio at rest varied as the anastomotic line became lower, suggesting that the degree of damage to internal sphincter muscles was uneven across individual patients, even though similar operative methods had been executed. Both TL and LCP at rest showed significant inverse correlations with Wexner incontinence score, indicating preservation of functional anal canal length as a critical factor in SPO procedures for better postoperative defecatory function. However, patients with moderate length of the anal canal, such as a TL of approximately 40 mm, showed Wexner scores ranging from 0

A. Hirano et al.

Sphincter damage following rectal surgery

to 20 (Figure 3A). Patients with LCP of approximately 25 mm showed scores ranging from 0 to 12 (Figure 3B), suggesting that incontinence score is not determined solely by sphincter tonus. Other factors that may affect postoperative defecatory function include function of the neorectum17 and postoperative motility of the pelvic floor.18 Meanwhile, TL and LCP under squeezing state did not differ significantly among HAR, LAR, vLAR. Only ISR showed significant decreases in both TL and LCP when compared with other SPO procedures. These findings suggested that external sphincter/levator ani muscles both work properly even after LAR or vLAR. In addition, external sphincter/levator ani muscles may be damaged during ISR procedures depending on the individual patient. Given these findings, we concluded that measuring TL, LCP, and LCP/TL ratio both at rest and in a squeezing state using 3D manometry is valuable for the evaluation of postoperative anal sphincter tonus derived from internal or external sphincter muscles and associated damage following SPO for rectal cancer. Resting anal canal lengths are shortened to various degrees after surgical procedures with pelvic floor maneuvers, namely LAR, vLAR, and ISR. Since LCP/TL ratio is also decreased in those operations, pressure disproportion in the anal canal seems to be associated with these operations. Judging from LCP/TL fluctuations as the line of anastomosis becomes lower, damage in the internal sphincter was uneven across individual patients. Excessive range of dissection on the anal side of the main tumor, excessive degree of electric coagulation at the pelvic floor, mechanical dilatation at the insertion of the circular stapler, or other technical issues at the bottom of the pelvis (which shows a wide degree of interindividual variation in shape) may damage internal sphincter tonus either mechanically or through injury to the nerve supply. Careful surgical maneuvers that minimize sphincter damage are thus warranted in SPOs for rectal cancer. Further study is needed using 3D manometry to evaluate the recovery of postoperative anal tonus with time after surgery.

References 1. Enker WE, Cohen AM, Lauwers GY. Treatment of rectal cancer by low anterior resection with coloanal anastomosis. Ann Surg 1994;219:365–73.

513 2. Portier G, Ghouti L, Kirzin S, et al. Oncological outcome of ultra-low coloanal anastomosis with and without intersphincteric resection for low rectal adenocarcinoma. Br J Surg 2007;94:341–5. 3. Schiessel R, Karner-Hanusch J, Herbst F, et al. Intersphincteric resection for low rectal tumours. Br J Surg 1994;81:1376 – 8. 4. Saito N, Moriya Y, Shirouzu K, et al. Intersphincteric resection in patients with very low rectal cancer: a review of the Japanese experience. Dis Colon Rectum 2006;49:13–22. 5. Murphy J, Hammond TM, Knowles CH, et al. Does anastomotic technique influence anorectal function after sphincter-saving rectal cancer resection? A systematic review of evidence from randomized trials. J Am Coll Surg 2007;204:673– 80. 6. Miller AS, Lewis WG, Willamson MER, et al. Factors that influence functional outcome after coloanal anastomosis for carcinoma of the rectum. Br J Surg 1995;82:1327–30. 7. Nakahara S, Itoh H, Mibu R, et al. Clinical and manometric evaluation of anorectal function following low anterior resection with low anastomotic line using an EEA stapler for rectal cancer. Dis Colon Rectum 1988;31:762– 6. 8. Otto IC, Ito K, Ye C, et al. Causes of rectal incontinence after sphincter-preserving operations for rectal cancer. Dis Colon Rectum 1996;39:1423–7. 9. Kawahara H, Kubota A, Hasegawa T, et al. Lack of esophageal contraction is a key determinant of gastroesophageal reflux disease after repair of esophageal atrasia. J Pediatr Surg 2007;42:2017–21. 10. Till H, Heinrich M, Schuster T, et al. Is the anorectal sphincter damaged during a transanal endorectal pull-through (TERPT) for Hirshcprung’s disease? A 3-dimensional, vector manometric investigation. Eur J Pediatr Surg 2006;16:188 –91. 11. Damon H, Henry L, Roman S, et al. Influence of rectal prolapse on the asymmetry of the anal sphincter in patients with anal incontinence. BMC Gastroenterol 2003;19:3–23. 12. Jorge JMN, Wexner SD. Etiology and management of fecal incontinence. Dis Colon Rectum 1993;36:77–97. 13. Horgan PG, O’Connel PR, Shinkwin CA, et al. Effect of anterior resection on anal sphincter function. Br J Surg 1989;76:783– 86. 14. Jehle EC, Haehnel T, Starlinger MJ, et al. Level of the anastomosis does not influence functional outcome after anterior rectal resection for rectal cancer. Am J Surg 1995;169:147–53. 15. Carlstedt A, Nordgren S, Fasth S, et al. Sympathetic nervous influence on the internal anal sphincter and rectum in man. Int J Colorectal Dis 1988;3:90 –5. 16. Williamson MER, Lewis WG, Finan PJ, et al. Recovery of physiologic and clinical function after low anterior resection of the rectum for carcinoma: myth or reality? Dis Colon Rectum 1955;38:411– 8. 17. Koda K, Saito N, Seike K, et al. Denervation of the neorectum as a potential cause of defecatory disorder following low anterior resection for rectal cancer. Dis Colon Rectum 2005;48:210 –7. 18. Morihiro M, Koda K, Seike K, et al. Characteristic findings on defecography according to reconstruction method and defecatory disorder following sphincter-saving surgery for rectal cancer. Int J Colorectal Dis 2008;23:883–92.