- Email: [email protected]
Results of 1,000 single-layer continuous polypropylene intestinal anastomoses
Results of 1,000 Single-Layer Continuous Polypropylene Intestinal Anastomoses Ernest Max, Donald R. Butts,
MD, MD,
W. Brian Sweeney, MD, H. Randolph Bailey, MD, Samuel C. Oommen, MD, Kenneth W. Smith, MD, Luis F. Zamora, MD, Gary B. Skakun, MD, Houston. TCUS
Between 1979 and 1988, we created intestinal anastomoses in 1,000 patients using a single-layer, continuous suturing technique and a polypropylene suture. The technique is easily learned, flexible in its application, and incurs less cost than most other techniques. The anastomoses involved all levels of the colon and the upper (intraperitoneal and extraperitoneal) rectum. All patients were followed for a minimum of 1 year. The clinically suspected anastomotic leak rate was 1%. Other morbidity included wound complications (2%), obstruction of the small intestine (2%), anastomotic stricture ( 1%) , and death ( 1%). No death was due to anastomotic complications. These rates of complications are comparable with, and in many instances lower than, those reported with other techniques of intestinal anastomosis.
From the Colon and Rectal Clinic, Houston, Texas. Rapwts for reprints should be addrcswl to Ernest Max, MD, Colon and Rectal Clinic, Smith Tower, Suite 2307,655O Fannin, Houston, Texas, 77030. Manuscript submitted August 13, 1990, and accepted in revised form December 3,199O.
any different techniques are currently used to join M segments of the intestinal tract following resection. These include anastomoses made with stapling devices or the sutureless biofragmentable anastomotic ring, single- and double-layer procedures, and interrupted and continuous suturing techniques using a variety of absorbable and nonabsorbable suture materials. No single method of anastomosis has proven to be clearly superior. This study retrospectively reviews 1,000 anastomoses performed utilizing a single-layer of a continuous polypropylene suture in order to evaluate the reliability and safety of this particular technique. PATIENTS
AND METHODS
A total of 1,000 consecutive patients who had a singlelayer continuous polypropylene anastomosis performed between 1979 and 1988 were reviewed. The anastomoses were performed by seven staff physicians and approximately 20 resident physicians under the supervision of the staff physicians. Of the 1,000 patients, 600 (60%) were women and 400 (40%) were men. The greater number of women was primarily due to a group of patients with advanced endometriosis and reflects our referral practice at The Women’s Hospital of Texas in Houston. The ages of the patients ranged from 16 to 96 years, with a mean of 56 years. Follow-up ranged from a minimum of 1 year to more than 10 years. The majority of procedures were elective and utilized a mechanical bowel preparation. Erythromycin base and neomycin were given orally beginning 18 hours before surgery. A perioperative course of one of the secondgeneration cephalosporins was also given. All anastomoses were constructed end to end; colostomies were made at the time of resection in less than 1% of patients. The indications for surgery are shown in Table I. Nearly half (45%) of the procedures were performed for colorectal cancer. Cancer, diverticulitis, and endometriosis together accounted for 82% of the resections. Resection for diverticulitis, endometriosis, inflammatory bowel disease, colostomy closure, or radiation stricture is generally associated with a higher incidence of anastomotic complications than is resection for colorectal cancer; these indications accounted for 450 (45%) of the resections. The levels of the anastomoses are depicted in Table II. A colocolic anastomosis, including anastomosis to the intraperitoneal rectum, was performed in 493 (49%) of the patients, and a colorectal anastomosis to the proximal extraperitoneal rectum was performed in 242 (24%) of the patients. Anastomoses to the distal extraperitoneal rectum were done using a stapling technique; these patients were excluded from this study. Ileocolic anasto-
THE AMERICAN JOURNAL OF SURGERY
VOLUME 162 NOVEMBER 1991
%I
TABLE I Indications for Resection
Indications Colorectal cancer Diverticuliiis Endometriosis lnflammatoty bowel disease
No. of Patients 448 210 162
52 29 23 17 IO 9 9 5 26
Benign villousadenoma Colostomyclosure Sigmoidvolvulus Rectal procidentia Multiplepolyposiscoli Trauma (endoscopicperforation) Metastaticovariancancer Other*
1,000
Total *Includesthree patients with radiation enteritis.
TABLE II Levels of the Anastomoses
Level
No. of Patients
Colocolic (includes intraperitoneal rectum) Colorectal (extraperitoneal rectum) lleocolic lleorectal (intraperitoneal) Total
493 242 234 31 1,000
moses accounted for 23% of the anastomoses and ileorectal 3%. Complications that were potentially related to the anastomosis were analyzed retrospectively; these included clinically suspected anastomotic leak, stricture, obstruction of the small intestine, wound complications, prolonged ileus, unexplained fever, suture-line bleeding, and death. Technique: The anastomosis is performed using an open-bowel technique. The portion of bowel to be resected is isolated by clamps, and the bowel is amputated close to the clamps, removing the crushed portions with the specimen (Figure 1). Both ends of the bowel lumen are irrigated with saline solution to remove any residual fecal matter (Fwe 2). The anastomosis is started at the mesenteric border of the bowel with a 4-O double-armed polypropylene suture, or two single-armed sutures placed close to each other, with the knot (or knots) placed outside of the lumen (Figure 3). Each suture bite incorporates approximately 5 mm of seromuscular and submucosal tissue with a minimal amount of mucosa (Figure 3) and is placed approximately 3 to 5 mm from the previous bite. The bowel edges are handled with atraumatic forceps. Both arms of the suture are continued around the anastomosis until they meet at the antimesenteric border, where the ends are tied together with a surgeon’s knot and two square knots to prevent untying (Figures 4 and 5). RESULTS Table III summarizes the postoperative complications found in this series. Twelve patients (1%) developed signs of an anastomotic leak. These included four patients who were noted to have a pelvic abscess without clinical or radiographic demonstration of anastomotic breakdown and a patient who developed a c&cutaneous fistula following colostomy. The leak was diagnosed radiographitally in two patients, endoscopically in one patient, and at
FlQW1.ThSpRXXldWl3iSpWformed using an open-bowel technique. The
pcrticnoftheboweltoberemovsdis amputated close to the clamps, removingtheau3hedporticnswiththe specimen. %2
THE AMERICAN
JOURNAL OF SURGERY
VOLUME 162
NOVEMBER 1991
the time of laparotomy and the creation of a stoma in one patient. The remaining three patients were diagnosed by clinical signs; they were treated with antibiotics and had healing documented by proctosigmoidoscopy in one patient and by follow-up contrast studies in two patients. Anastomotic stricture, defined as the inability to pass a flexible endoscope through the anastomosis, developed in seven patients (1%). Of these, three patients had undergone resection for advanced, complicated diverticulitis, two for extraperitoneal rectal cancer, one for solitary rectal ulcer syndrome, and one for sigmoid colostomy closure. Four of the seven patients were treated successfully with dilation. In one patient, the stricture was repaired by transrectal stricture-plasty. The remaining two patients required resection of the stricture. There were eight deaths among the 1,000 patients (1%). None was related to anastomotic complications. One patient, an 80-year-old man who had undergone a transverse colon resection for a villoglandular polyp, developed a wound dehiscence on the fifth postoperative day. He had closure of his dehiscence on postoperative day 5, at which time the anastomosis was inspected and found to be intact. He later died of pneumonia. The other seven deaths were due to cardiovascular complications. Five patients had an unexplained fever (temperature greater than 100.5’F with no obvious source) beyond the fifth postoperative day. These patients were examined for anastomotic leaks or abscesses with negative results. None of the live patients required reoperation, and all fevers resolved. Small bowel obstruction requiring readmission occurred in 22 patients (2%), 8 of whom required laparotomy. Prolonged ileus, defined as distention with nausea or vomiting that required the placement of a nasogastric
Fm 2. The proximaland distal ends of the bowel are irrigated with saline solutionto remove any residualfecal matter.
tube, was observed in 19 patients. None of these patients required reoperation. Wound complications occurred in 24 patients. Eighteen patients had an infection limited to the subcutaneous tissue, and 6 had fascial disruption. There were no sutureline hemorrhages. COMMENTS
The safety and effectiveness of a technique of intestinal anastomosis may be measured by the incidence of postoperative complications, especially the clinically suspected anastomotic leak rate. Table IV lists the rates of
Figure3.Wlththeuseofadouble armed4-o~suhre,theJsu ttrelineisstHedatthemesen&kbcc derofthebowel.Aknotinitsmk&xHor isplacedoutsidethelunen.lhebitt (Xffihs about 5 rI-&ofwz s&rnuxM * minimal amount of muoosa.
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
162
NOVEMBER
1991
463
F@re 4. The suhre line is continued around the anastomosisalternatingsutures and using atraumatic forceps to handle the bowel. The sutures are placed approximately3 to 5 mm epart.
Flgure5.Thetwosuturesmeetatthe antimesentericside of the bowel, whsre they are tied together.
TABLE III Compllcetlons In 1,000 Patlents Wlth a Single-Layer, Contlnuous Polypropylene Intestinal Anastomosls No. of Patients (%)*
Complications Anastomotic leak (clinically suspected) Colocutaneous fistula 1 (0) Pelvic abscess 4 (0) Identified leak 7 (I)
12 (1)
Wound complications Wound infection Fascial disruption
24 (2) 18 (2) 6 (1)
Obstruction of small intestine Stricture Prolonged ileus Unexplained fever beyond postoperative day 5 Suture line bleeding Death ‘Percentagesrounded
22 (2) 7 (1) 19 (2) 5 (1) 0 8 (I)+
oft to nearest whole percent.
*No death was due to anastomotic complications.
464
THE AMERICAN
JOURNAL OF SURGERY
complications reported in a number of series in which different intestinal reconstruction techniques were used [I-I I]. The incidence of anastomotic leaks ranges from 0% in Harder and Vogelbach’s [a series utilizing a single-layer absorbable continuous suture to 15% in Antonsen and Kronborg’s [I] series utilizing the EEA circular stapling device. It should be noted, however, that Harder and Vogelbach excluded patients with extraperitoneal colorectal anastomoses. Antonsen and Kronborg’s anastomoses, on the other hand, were exclusively to the extraperitoneal rectum, procedures that are generally associated with a higher incidence of complications [12,13]. Among series that included anastomoses at all levels, the incidence of clinical anastomotic leak ranged from 1% to 9% [8,11]. The average rate of clinically suspected anastomotic leaks in these 19 series, which included 3 series that were extraperitoneal colorectal only and 6 that excluded all extraperitoneal colorectal anastomoses, was 5%. The clinically suspected anastomotic leak rate in our series was l%, one of the lower rates reported. While patients with anastomoses to the most distal portion of
VOLUME 162
NOVEMBER 1991
CONTINUOUS POLYPROPYLENE
INTESTINAL
ANASTOMOSES
TABLE IV Complications Followlng Intestinal Anastomosls In Studies Uslng Dlfferent Reconstruction Techniques*
Study
No. of Patients
Follow-Up
Type of Anastomosis
Leak
Obstruction
(%)**
(%)
Wound (%)++
Death (%)
15+
_$
IO
3
Extraperitoneal colorectal only; strictures associated with leaks No death from anastomosis; one surgeon responsible for 50% of leaks
Comments
Ill
170
-
EEA circular stapling device
I21
101
-
Biofragmentable anastomotic ring (BAR)
2
4
10
2
85
_
8
1
12
2
16
-
Monofilament absorbable continuous suture Staples
6
6
0
6
162 54 222
6wk 6wk 6wk
Sutures Staples Biofragmentable anastomotic ring (BAR)
2* 2* 3*
2 4 4
6 6 5
2 4 2
Excluded high-risk patients and extraperitoneal colorectal; no death from anastomosis
1 mo
Interrupted sutures
2
16
5
Extraperitoneal colorectal excluded; 34% of deaths due to anastomotic leak
[3]
[41
[51
153
-
143
[7]
02
6*
1
-5
3
Extraperftoneal colorectal excluded
Single-layer monofilament resorbable synthetic continuous suture
ot
-§
-5
0
Extraperitoneal colorectal exeluded; 27% emergencies
EEA circular stapling device
4
4
-5
1
lntraoperatiie air leakage test
2
2
All leaks colorectal; no death from anastomosis
204
4-6wk
Single-layer interrupted braided polyamide
1
_$
1,000
f yr
Single-layer continuous polypropylene
1t
2
2
1
Distal ex-traperltonealcolorectal excluded; no death from anastomotic leak
327
2 m-4 yr
Single-layer interrupted Vicryl sutures
1
-5
-3
5
15% extraperitoneal colorectal
60
-
Single-layer interrupted braided polyamide Circular stapling device
3’
-5
5
0
All extraperitoneal colorectal; intraoperatiie air leakage test
i61 1990
-
Single-layer interrupted sutures
[lo]
[ll]
58
-
588 210
-
115
-
Monofilament steel wire sutures Two-layer cotton and chromic catgut sutures Single-layer braided multlllament
12*
_§
9
1
3 9
_§ -5
-§ _§
3 4
6
-0
-0
-5
Braided abandoned early in study
wire sutures lPerc8n+sges rounded
off lo nearest whole percent.
WI extraperitoneal cobrectal. *Exlrapeitoneal
cobrectal excluded.
wet stated. ‘*Leak Wound
is defined as dinically apparent leak. is dev7w.I as wound complications (infection and dehiscence).
the extraperitoneal rectum were excluded, approximately 25% of the resections were to the extraperitoneal rectum. In addition, and as noted earlier, our series had a large proportion (45%) of patients who had surgery for a disease process (diverticulitis, endometriosis, inflammatory bowel disease, colostomy closure, or radiation enteritis) that is generally associated with a higher rate of postoperative anastomotic complications than is resection for cancer [12-141. Thus, while the exclusion of very low cola rectal anastomoses may have favorably influenced the incidence of postoperative complications slightly, the relatively low rate of morbidity in our series is not in general THE AMERICAN
due to a pool of patients who would be expected to have fewer complications. The 1,000 resections were done by approximately 27 different surgeons, with remarkably similar results. This suggests that the low rate of complications was not the result of one or two surgeons’ perfected abiities with the technique, but that the technique itself is easily learned and can be used in different hands to produce consistent and satisfactory results. A singlelayer anastomosis has been advocated by a number of investigators because it causes less narrowing of the lumen and because a smaller amount of tissue is
JOURNAL
OF SURGERY
VOLUME
162 NOVEMBER
1991
&5
---
Figure 6. “Coiledspring”concept of the anastomosis.Pressure from withinexpands the loopedsuture line. strangulated [5,15]. Several studies have noted that devascularization, infection, and necrosis are more likely to occur after a double-layer anastomosis [15-l 71. A singlelayer continuous suturing technique may also be less likely to produce focal strangulation and tissue damage than a single-layer interrupted technique [6,14] because the continuous technique distributes tension more evenly around the lumen. One concern that has been frequently expressed with regard to the single-layer continuous technique using a nonabsorbable suture is that it might produce a nonyielding stricture due to “purse-stringing” of the anastomosis. This has not been proven to be the case. Instead, the continuous suture appears to act like a coiled spring, actually expanding in response to intraluminal pressure (Figure 6). In our series, seven patients (less than 1%) developed evidence of anastomotic stricture, and obstruction of the small intestine developed in 22 patients (2%). These rates appear to be low. However, as Luchtefeld and associates [28] have noted, the incidence of these complications is difficult to compare among various series because of differences in the definitions of stricture, stenosis, and obstruction, and because of different lengths of follow-up. Luchtefeld and co-workers surveyed the membership of the American Society of Colon and Rectal Surgeons for cases of intestinal stenosis following anastomosis and found the incidence in various studies to range from 0% at 2 weeks to 22% at 3 months. They noted that, of the 123 cases reported, 82 (67%) occurred after a stapled anastomosis and that a postoperative anastomotic leak or pelvic infection was often associated with the later development of stenosis. They also noted that the postop erative evaluation needs to extend to at least 6 months after surgery to detect the true incidence of stenosis. Studies have shown that monofilament steel and monofilament synthetic sutures produce significantly less tissue reaction and resist infection more effectively than catgut, silk, or braided synthetic sutures [21,19]. Absorbable suture materials have not been widely employed in the past because of the intraluminal stress placed on the suture line. However, several surgeons have used absorb able suture material with considerable success [6,9]. We have recently begun clinical trials with a monofilament absorbable suture material, Maxon, which absorbs after 60 to 90 days; initial results are encouraging. Monofilament suture material does have the disad466
THE AMERICAN JOURNAL OF SURGERY
vantage of being difficult to tie securely, thus requiring careful and time-consuming securing of knots to prevent untying. When used for interrupted sutures, much time must be spent securing the knots. A continuous suturing technique, however, obviates this problem and also reduces the risk of infection associated with the knots themselves [6]. A continuous technique is also performed considerably more rapidly than an interrupted technique and, in some hands, may be faster than using a stapling device. In a previous study, we found the average total time in the operating room when performing a singlelayer continuous anastomosis to be 127 minutes [14]. More recently, we have recorded in 100 patients the time necessary to perform the anastomosis itself and found it to range between 8 and 22 minutes, with an average of 12 minutes. Adloff and associates [20] found that the average total time in surgery for procedures involving a stapled anastomosis was 180 minutes, while, for a singlelayer interrupted sutured anastomosis, it was 170 minutes. McGinn and co-workers [IO] also found no significant difference in the time required to do a stapled anastomosis and that required for an anastomosis utilizing interrupted sutures. One final concern is that of expense. The continuous suturing technique not only incurs less expense than do procedures that utilize devices such as the circular stapler and biofragmentable anastomotic ring, but it is less expensive than the interrupted suturing technique that requires multiple packages of sutures. In their study, McGinn and associates [IO] found that a stapling technique was 10 times more expensive than an interrupted suturing technique; this difference was due in part to increased morbidity.
CONCLUSIONS This study has reinforced our earlier findings, documented in previously published studies [14,21,22], concerning the efficacy of the single-layer continuous polypropylene intestinal anastomosis. The technique reliably creates a wide-lumen anastomosis with minimal tissue reaction and a low incidence of postoperative complications, and it does so at a lower cost to the patient than most other techniques. Although the paper is not original, it does have a large series of patients and it demonstrates that a surgical group that pays attention to meticulous technique can do a large number of bowel anastomoses with safety. To what extent it is the suture and technique versus the overall skills of the surgeons remains to be seen.
REFERENCES 1. AntonsenHK, Kronborg0. Early complications after lowanterior resectionfor rectal cancer using the EEA staplingdevice:a prospectivetrial. Dis Colon Rectum 1987; 30: 579-83.
2. Cahill CJ, Betzler M, Gruwez JA, Jeekel J, Pate1 J-C, Zederfeldt B. Sutureless large bowel anastomosis: European experience with the biofragmentable ring. Br J Surg 1989; 76: 344-7. 3. Corman ML, Prager ED, Hardy TG Jr, Bubrick MF, Valtrac (BAR) Study Group. Comparison of the Valtrac biofragmentable
VOLUME 162 NOVEMBER 1991
CONTlNUOUSPOLYPROPYLENEINTESTINALANASTOMOSES
anastomosis in colon surgery: results of a prospective, randomized clinical trial. Dis Colon Rectum 1989; 32: 183-7. 4. Debas HT, Thomson FB. A critical review of colectomy with anastomosis. Surg Gynecol Obstet 1972; 135: 747-52. 5. Gambee LP, Gamjobst W, Hardwick CE. Ten years’ experience with single layer anastomosis in colon surgery. Am J Surg 1956; 92: 222-7. 6. Harder F, Vogelbach P. Singlelayer end-on continuous suture of colonic anastomoses. Am J Surg 1988; 155: 611-4. 7. Lazorthes F, Chiotassol P. Stapled colorectal anastomoses: per+ perative integrity of the anastomosis and risk of postoperative leakage. Int J Colorectal Dii 1986; 1: 96-8. 8. Ma&son NA, McIntosh CA, Krukowski ZH. Continuing expe rience with singlslayer appositional anastomosis in the large bowel. Br J Surg 1985; 72: S104-6. 9. McDonald CC, Baird RL. Vicrylo intestinal anastomosis: analysis of 327 cases. Dis Colon Rectum 1985; 28: 775-6. 10. McGinn FP, Gartell PC, Clifford PC, Brunton FJ. Staples or sutures for low cclorectal anastomoses: a prospective randomized trial. Br J Surg 1985; 72: 603-5. 11. Trimpi HD, Khubchandani IT, Sheets JA, Stasik JJ Jr. Advances in intestinal anastomosisz experimental study and an analysis of 984 patients. Dis Colon Rectum 1977; 20: 107-17. 12. Mattson PN, Corman ML, Caller JA, Veidenheimer MC. Anterior resection for adenocarcinoma: Lahey Clinic experience from 1963 through 1969. Am J Surg 1976; 131: 434-41. 13. Heberer G, Denecke H, Pratschke E, Teichmann R. Anterior and low anterior resection. World J Surg 1982; 6: 517-24.
14. Bailey HR, La Voo JW, Max E, Smith KW, Butts DR, Hamp ton JM. Single-layer polypropylene colorectal anastomosis: experience with 100 cases. Dis Colon Rectum 1984; 27: 19-23. 15. Templeton JL, McKelvey STD. Low colorectal anastomoses: an experimental assessment of two sutured and two stapled techniques. Dis Colon Rectum 1985; 28: 38-41. 16. Graffner H, Anderson L, Lijwenhielm P, Walther B. The healing process of anastomcsis of the colon: a comparative study using single, double-layer or stapled anastomosis. Dis Colon Rectum 1984; 27: 767-71. 17. Olsen BG, Letwin E, Williams HTS. Clinical experience with the use of a single layer intestinal anastomosis. Can J Surg 1968; 11: 97-100. 18. Luchtefeld MA, Milsom JW, Senagore A, Surrell JA, Mazier WP. Colorectal anastomotic stenosis: results of a survery of the ASCRS membership. Dis Colon Rectum 1989; 32: 733-6. 19. Hastings JC, Van Winkle W, Barker E, Hines D, Nichols W. Effect of suture materials on healing wounds of the stomach and colon. Surg Gynecol Obstet 1975; 140: 701-7. 20. Adloff M, Amoud JP, Beehary S. Stapled versus sutured colorectal anastomosis. Arch Surg 1980; 115: 1436-8. 21. Bailey HR, LaVoo JW, Max E, Smith KW, Hampton JM. Single layer continuous cclorectal anastomcsis. Aust N Z J Surg 1981; 51: 473-6. 22. Skakun GB, Rexnick RK, Bailey HR, Smith KW, Max E. The single-layer continuous polypropylene colon anastomosis: a prospective assessment using water-soluble contrast enemas. Dis Colon Rectum 1988; 31: 163-8.
THE AMERICAN JOURNAL OF SURGERY
VOLUME 162 NOVEMBER 1991
467
MD, MD,
W. Brian Sweeney, MD, H. Randolph Bailey, MD, Samuel C. Oommen, MD, Kenneth W. Smith, MD, Luis F. Zamora, MD, Gary B. Skakun, MD, Houston. TCUS
Between 1979 and 1988, we created intestinal anastomoses in 1,000 patients using a single-layer, continuous suturing technique and a polypropylene suture. The technique is easily learned, flexible in its application, and incurs less cost than most other techniques. The anastomoses involved all levels of the colon and the upper (intraperitoneal and extraperitoneal) rectum. All patients were followed for a minimum of 1 year. The clinically suspected anastomotic leak rate was 1%. Other morbidity included wound complications (2%), obstruction of the small intestine (2%), anastomotic stricture ( 1%) , and death ( 1%). No death was due to anastomotic complications. These rates of complications are comparable with, and in many instances lower than, those reported with other techniques of intestinal anastomosis.
From the Colon and Rectal Clinic, Houston, Texas. Rapwts for reprints should be addrcswl to Ernest Max, MD, Colon and Rectal Clinic, Smith Tower, Suite 2307,655O Fannin, Houston, Texas, 77030. Manuscript submitted August 13, 1990, and accepted in revised form December 3,199O.
any different techniques are currently used to join M segments of the intestinal tract following resection. These include anastomoses made with stapling devices or the sutureless biofragmentable anastomotic ring, single- and double-layer procedures, and interrupted and continuous suturing techniques using a variety of absorbable and nonabsorbable suture materials. No single method of anastomosis has proven to be clearly superior. This study retrospectively reviews 1,000 anastomoses performed utilizing a single-layer of a continuous polypropylene suture in order to evaluate the reliability and safety of this particular technique. PATIENTS
AND METHODS
A total of 1,000 consecutive patients who had a singlelayer continuous polypropylene anastomosis performed between 1979 and 1988 were reviewed. The anastomoses were performed by seven staff physicians and approximately 20 resident physicians under the supervision of the staff physicians. Of the 1,000 patients, 600 (60%) were women and 400 (40%) were men. The greater number of women was primarily due to a group of patients with advanced endometriosis and reflects our referral practice at The Women’s Hospital of Texas in Houston. The ages of the patients ranged from 16 to 96 years, with a mean of 56 years. Follow-up ranged from a minimum of 1 year to more than 10 years. The majority of procedures were elective and utilized a mechanical bowel preparation. Erythromycin base and neomycin were given orally beginning 18 hours before surgery. A perioperative course of one of the secondgeneration cephalosporins was also given. All anastomoses were constructed end to end; colostomies were made at the time of resection in less than 1% of patients. The indications for surgery are shown in Table I. Nearly half (45%) of the procedures were performed for colorectal cancer. Cancer, diverticulitis, and endometriosis together accounted for 82% of the resections. Resection for diverticulitis, endometriosis, inflammatory bowel disease, colostomy closure, or radiation stricture is generally associated with a higher incidence of anastomotic complications than is resection for colorectal cancer; these indications accounted for 450 (45%) of the resections. The levels of the anastomoses are depicted in Table II. A colocolic anastomosis, including anastomosis to the intraperitoneal rectum, was performed in 493 (49%) of the patients, and a colorectal anastomosis to the proximal extraperitoneal rectum was performed in 242 (24%) of the patients. Anastomoses to the distal extraperitoneal rectum were done using a stapling technique; these patients were excluded from this study. Ileocolic anasto-
THE AMERICAN JOURNAL OF SURGERY
VOLUME 162 NOVEMBER 1991
%I
TABLE I Indications for Resection
Indications Colorectal cancer Diverticuliiis Endometriosis lnflammatoty bowel disease
No. of Patients 448 210 162
52 29 23 17 IO 9 9 5 26
Benign villousadenoma Colostomyclosure Sigmoidvolvulus Rectal procidentia Multiplepolyposiscoli Trauma (endoscopicperforation) Metastaticovariancancer Other*
1,000
Total *Includesthree patients with radiation enteritis.
TABLE II Levels of the Anastomoses
Level
No. of Patients
Colocolic (includes intraperitoneal rectum) Colorectal (extraperitoneal rectum) lleocolic lleorectal (intraperitoneal) Total
493 242 234 31 1,000
moses accounted for 23% of the anastomoses and ileorectal 3%. Complications that were potentially related to the anastomosis were analyzed retrospectively; these included clinically suspected anastomotic leak, stricture, obstruction of the small intestine, wound complications, prolonged ileus, unexplained fever, suture-line bleeding, and death. Technique: The anastomosis is performed using an open-bowel technique. The portion of bowel to be resected is isolated by clamps, and the bowel is amputated close to the clamps, removing the crushed portions with the specimen (Figure 1). Both ends of the bowel lumen are irrigated with saline solution to remove any residual fecal matter (Fwe 2). The anastomosis is started at the mesenteric border of the bowel with a 4-O double-armed polypropylene suture, or two single-armed sutures placed close to each other, with the knot (or knots) placed outside of the lumen (Figure 3). Each suture bite incorporates approximately 5 mm of seromuscular and submucosal tissue with a minimal amount of mucosa (Figure 3) and is placed approximately 3 to 5 mm from the previous bite. The bowel edges are handled with atraumatic forceps. Both arms of the suture are continued around the anastomosis until they meet at the antimesenteric border, where the ends are tied together with a surgeon’s knot and two square knots to prevent untying (Figures 4 and 5). RESULTS Table III summarizes the postoperative complications found in this series. Twelve patients (1%) developed signs of an anastomotic leak. These included four patients who were noted to have a pelvic abscess without clinical or radiographic demonstration of anastomotic breakdown and a patient who developed a c&cutaneous fistula following colostomy. The leak was diagnosed radiographitally in two patients, endoscopically in one patient, and at
FlQW1.ThSpRXXldWl3iSpWformed using an open-bowel technique. The
pcrticnoftheboweltoberemovsdis amputated close to the clamps, removingtheau3hedporticnswiththe specimen. %2
THE AMERICAN
JOURNAL OF SURGERY
VOLUME 162
NOVEMBER 1991
the time of laparotomy and the creation of a stoma in one patient. The remaining three patients were diagnosed by clinical signs; they were treated with antibiotics and had healing documented by proctosigmoidoscopy in one patient and by follow-up contrast studies in two patients. Anastomotic stricture, defined as the inability to pass a flexible endoscope through the anastomosis, developed in seven patients (1%). Of these, three patients had undergone resection for advanced, complicated diverticulitis, two for extraperitoneal rectal cancer, one for solitary rectal ulcer syndrome, and one for sigmoid colostomy closure. Four of the seven patients were treated successfully with dilation. In one patient, the stricture was repaired by transrectal stricture-plasty. The remaining two patients required resection of the stricture. There were eight deaths among the 1,000 patients (1%). None was related to anastomotic complications. One patient, an 80-year-old man who had undergone a transverse colon resection for a villoglandular polyp, developed a wound dehiscence on the fifth postoperative day. He had closure of his dehiscence on postoperative day 5, at which time the anastomosis was inspected and found to be intact. He later died of pneumonia. The other seven deaths were due to cardiovascular complications. Five patients had an unexplained fever (temperature greater than 100.5’F with no obvious source) beyond the fifth postoperative day. These patients were examined for anastomotic leaks or abscesses with negative results. None of the live patients required reoperation, and all fevers resolved. Small bowel obstruction requiring readmission occurred in 22 patients (2%), 8 of whom required laparotomy. Prolonged ileus, defined as distention with nausea or vomiting that required the placement of a nasogastric
Fm 2. The proximaland distal ends of the bowel are irrigated with saline solutionto remove any residualfecal matter.
tube, was observed in 19 patients. None of these patients required reoperation. Wound complications occurred in 24 patients. Eighteen patients had an infection limited to the subcutaneous tissue, and 6 had fascial disruption. There were no sutureline hemorrhages. COMMENTS
The safety and effectiveness of a technique of intestinal anastomosis may be measured by the incidence of postoperative complications, especially the clinically suspected anastomotic leak rate. Table IV lists the rates of
Figure3.Wlththeuseofadouble armed4-o~suhre,theJsu ttrelineisstHedatthemesen&kbcc derofthebowel.Aknotinitsmk&xHor isplacedoutsidethelunen.lhebitt (Xffihs about 5 rI-&ofwz s&rnuxM * minimal amount of muoosa.
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
162
NOVEMBER
1991
463
F@re 4. The suhre line is continued around the anastomosisalternatingsutures and using atraumatic forceps to handle the bowel. The sutures are placed approximately3 to 5 mm epart.
Flgure5.Thetwosuturesmeetatthe antimesentericside of the bowel, whsre they are tied together.
TABLE III Compllcetlons In 1,000 Patlents Wlth a Single-Layer, Contlnuous Polypropylene Intestinal Anastomosls No. of Patients (%)*
Complications Anastomotic leak (clinically suspected) Colocutaneous fistula 1 (0) Pelvic abscess 4 (0) Identified leak 7 (I)
12 (1)
Wound complications Wound infection Fascial disruption
24 (2) 18 (2) 6 (1)
Obstruction of small intestine Stricture Prolonged ileus Unexplained fever beyond postoperative day 5 Suture line bleeding Death ‘Percentagesrounded
22 (2) 7 (1) 19 (2) 5 (1) 0 8 (I)+
oft to nearest whole percent.
*No death was due to anastomotic complications.
464
THE AMERICAN
JOURNAL OF SURGERY
complications reported in a number of series in which different intestinal reconstruction techniques were used [I-I I]. The incidence of anastomotic leaks ranges from 0% in Harder and Vogelbach’s [a series utilizing a single-layer absorbable continuous suture to 15% in Antonsen and Kronborg’s [I] series utilizing the EEA circular stapling device. It should be noted, however, that Harder and Vogelbach excluded patients with extraperitoneal colorectal anastomoses. Antonsen and Kronborg’s anastomoses, on the other hand, were exclusively to the extraperitoneal rectum, procedures that are generally associated with a higher incidence of complications [12,13]. Among series that included anastomoses at all levels, the incidence of clinical anastomotic leak ranged from 1% to 9% [8,11]. The average rate of clinically suspected anastomotic leaks in these 19 series, which included 3 series that were extraperitoneal colorectal only and 6 that excluded all extraperitoneal colorectal anastomoses, was 5%. The clinically suspected anastomotic leak rate in our series was l%, one of the lower rates reported. While patients with anastomoses to the most distal portion of
VOLUME 162
NOVEMBER 1991
CONTINUOUS POLYPROPYLENE
INTESTINAL
ANASTOMOSES
TABLE IV Complications Followlng Intestinal Anastomosls In Studies Uslng Dlfferent Reconstruction Techniques*
Study
No. of Patients
Follow-Up
Type of Anastomosis
Leak
Obstruction
(%)**
(%)
Wound (%)++
Death (%)
15+
_$
IO
3
Extraperitoneal colorectal only; strictures associated with leaks No death from anastomosis; one surgeon responsible for 50% of leaks
Comments
Ill
170
-
EEA circular stapling device
I21
101
-
Biofragmentable anastomotic ring (BAR)
2
4
10
2
85
_
8
1
12
2
16
-
Monofilament absorbable continuous suture Staples
6
6
0
6
162 54 222
6wk 6wk 6wk
Sutures Staples Biofragmentable anastomotic ring (BAR)
2* 2* 3*
2 4 4
6 6 5
2 4 2
Excluded high-risk patients and extraperitoneal colorectal; no death from anastomosis
1 mo
Interrupted sutures
2
16
5
Extraperitoneal colorectal excluded; 34% of deaths due to anastomotic leak
[3]
[41
[51
153
-
143
[7]
02
6*
1
-5
3
Extraperftoneal colorectal excluded
Single-layer monofilament resorbable synthetic continuous suture
ot
-§
-5
0
Extraperitoneal colorectal exeluded; 27% emergencies
EEA circular stapling device
4
4
-5
1
lntraoperatiie air leakage test
2
2
All leaks colorectal; no death from anastomosis
204
4-6wk
Single-layer interrupted braided polyamide
1
_$
1,000
f yr
Single-layer continuous polypropylene
1t
2
2
1
Distal ex-traperltonealcolorectal excluded; no death from anastomotic leak
327
2 m-4 yr
Single-layer interrupted Vicryl sutures
1
-5
-3
5
15% extraperitoneal colorectal
60
-
Single-layer interrupted braided polyamide Circular stapling device
3’
-5
5
0
All extraperitoneal colorectal; intraoperatiie air leakage test
i61 1990
-
Single-layer interrupted sutures
[lo]
[ll]
58
-
588 210
-
115
-
Monofilament steel wire sutures Two-layer cotton and chromic catgut sutures Single-layer braided multlllament
12*
_§
9
1
3 9
_§ -5
-§ _§
3 4
6
-0
-0
-5
Braided abandoned early in study
wire sutures lPerc8n+sges rounded
off lo nearest whole percent.
WI extraperitoneal cobrectal. *Exlrapeitoneal
cobrectal excluded.
wet stated. ‘*Leak Wound
is defined as dinically apparent leak. is dev7w.I as wound complications (infection and dehiscence).
the extraperitoneal rectum were excluded, approximately 25% of the resections were to the extraperitoneal rectum. In addition, and as noted earlier, our series had a large proportion (45%) of patients who had surgery for a disease process (diverticulitis, endometriosis, inflammatory bowel disease, colostomy closure, or radiation enteritis) that is generally associated with a higher rate of postoperative anastomotic complications than is resection for cancer [12-141. Thus, while the exclusion of very low cola rectal anastomoses may have favorably influenced the incidence of postoperative complications slightly, the relatively low rate of morbidity in our series is not in general THE AMERICAN
due to a pool of patients who would be expected to have fewer complications. The 1,000 resections were done by approximately 27 different surgeons, with remarkably similar results. This suggests that the low rate of complications was not the result of one or two surgeons’ perfected abiities with the technique, but that the technique itself is easily learned and can be used in different hands to produce consistent and satisfactory results. A singlelayer anastomosis has been advocated by a number of investigators because it causes less narrowing of the lumen and because a smaller amount of tissue is
JOURNAL
OF SURGERY
VOLUME
162 NOVEMBER
1991
&5
---
Figure 6. “Coiledspring”concept of the anastomosis.Pressure from withinexpands the loopedsuture line. strangulated [5,15]. Several studies have noted that devascularization, infection, and necrosis are more likely to occur after a double-layer anastomosis [15-l 71. A singlelayer continuous suturing technique may also be less likely to produce focal strangulation and tissue damage than a single-layer interrupted technique [6,14] because the continuous technique distributes tension more evenly around the lumen. One concern that has been frequently expressed with regard to the single-layer continuous technique using a nonabsorbable suture is that it might produce a nonyielding stricture due to “purse-stringing” of the anastomosis. This has not been proven to be the case. Instead, the continuous suture appears to act like a coiled spring, actually expanding in response to intraluminal pressure (Figure 6). In our series, seven patients (less than 1%) developed evidence of anastomotic stricture, and obstruction of the small intestine developed in 22 patients (2%). These rates appear to be low. However, as Luchtefeld and associates [28] have noted, the incidence of these complications is difficult to compare among various series because of differences in the definitions of stricture, stenosis, and obstruction, and because of different lengths of follow-up. Luchtefeld and co-workers surveyed the membership of the American Society of Colon and Rectal Surgeons for cases of intestinal stenosis following anastomosis and found the incidence in various studies to range from 0% at 2 weeks to 22% at 3 months. They noted that, of the 123 cases reported, 82 (67%) occurred after a stapled anastomosis and that a postoperative anastomotic leak or pelvic infection was often associated with the later development of stenosis. They also noted that the postop erative evaluation needs to extend to at least 6 months after surgery to detect the true incidence of stenosis. Studies have shown that monofilament steel and monofilament synthetic sutures produce significantly less tissue reaction and resist infection more effectively than catgut, silk, or braided synthetic sutures [21,19]. Absorbable suture materials have not been widely employed in the past because of the intraluminal stress placed on the suture line. However, several surgeons have used absorb able suture material with considerable success [6,9]. We have recently begun clinical trials with a monofilament absorbable suture material, Maxon, which absorbs after 60 to 90 days; initial results are encouraging. Monofilament suture material does have the disad466
THE AMERICAN JOURNAL OF SURGERY
vantage of being difficult to tie securely, thus requiring careful and time-consuming securing of knots to prevent untying. When used for interrupted sutures, much time must be spent securing the knots. A continuous suturing technique, however, obviates this problem and also reduces the risk of infection associated with the knots themselves [6]. A continuous technique is also performed considerably more rapidly than an interrupted technique and, in some hands, may be faster than using a stapling device. In a previous study, we found the average total time in the operating room when performing a singlelayer continuous anastomosis to be 127 minutes [14]. More recently, we have recorded in 100 patients the time necessary to perform the anastomosis itself and found it to range between 8 and 22 minutes, with an average of 12 minutes. Adloff and associates [20] found that the average total time in surgery for procedures involving a stapled anastomosis was 180 minutes, while, for a singlelayer interrupted sutured anastomosis, it was 170 minutes. McGinn and co-workers [IO] also found no significant difference in the time required to do a stapled anastomosis and that required for an anastomosis utilizing interrupted sutures. One final concern is that of expense. The continuous suturing technique not only incurs less expense than do procedures that utilize devices such as the circular stapler and biofragmentable anastomotic ring, but it is less expensive than the interrupted suturing technique that requires multiple packages of sutures. In their study, McGinn and associates [IO] found that a stapling technique was 10 times more expensive than an interrupted suturing technique; this difference was due in part to increased morbidity.
CONCLUSIONS This study has reinforced our earlier findings, documented in previously published studies [14,21,22], concerning the efficacy of the single-layer continuous polypropylene intestinal anastomosis. The technique reliably creates a wide-lumen anastomosis with minimal tissue reaction and a low incidence of postoperative complications, and it does so at a lower cost to the patient than most other techniques. Although the paper is not original, it does have a large series of patients and it demonstrates that a surgical group that pays attention to meticulous technique can do a large number of bowel anastomoses with safety. To what extent it is the suture and technique versus the overall skills of the surgeons remains to be seen.
REFERENCES 1. AntonsenHK, Kronborg0. Early complications after lowanterior resectionfor rectal cancer using the EEA staplingdevice:a prospectivetrial. Dis Colon Rectum 1987; 30: 579-83.
2. Cahill CJ, Betzler M, Gruwez JA, Jeekel J, Pate1 J-C, Zederfeldt B. Sutureless large bowel anastomosis: European experience with the biofragmentable ring. Br J Surg 1989; 76: 344-7. 3. Corman ML, Prager ED, Hardy TG Jr, Bubrick MF, Valtrac (BAR) Study Group. Comparison of the Valtrac biofragmentable
VOLUME 162 NOVEMBER 1991
CONTlNUOUSPOLYPROPYLENEINTESTINALANASTOMOSES
anastomosis in colon surgery: results of a prospective, randomized clinical trial. Dis Colon Rectum 1989; 32: 183-7. 4. Debas HT, Thomson FB. A critical review of colectomy with anastomosis. Surg Gynecol Obstet 1972; 135: 747-52. 5. Gambee LP, Gamjobst W, Hardwick CE. Ten years’ experience with single layer anastomosis in colon surgery. Am J Surg 1956; 92: 222-7. 6. Harder F, Vogelbach P. Singlelayer end-on continuous suture of colonic anastomoses. Am J Surg 1988; 155: 611-4. 7. Lazorthes F, Chiotassol P. Stapled colorectal anastomoses: per+ perative integrity of the anastomosis and risk of postoperative leakage. Int J Colorectal Dii 1986; 1: 96-8. 8. Ma&son NA, McIntosh CA, Krukowski ZH. Continuing expe rience with singlslayer appositional anastomosis in the large bowel. Br J Surg 1985; 72: S104-6. 9. McDonald CC, Baird RL. Vicrylo intestinal anastomosis: analysis of 327 cases. Dis Colon Rectum 1985; 28: 775-6. 10. McGinn FP, Gartell PC, Clifford PC, Brunton FJ. Staples or sutures for low cclorectal anastomoses: a prospective randomized trial. Br J Surg 1985; 72: 603-5. 11. Trimpi HD, Khubchandani IT, Sheets JA, Stasik JJ Jr. Advances in intestinal anastomosisz experimental study and an analysis of 984 patients. Dis Colon Rectum 1977; 20: 107-17. 12. Mattson PN, Corman ML, Caller JA, Veidenheimer MC. Anterior resection for adenocarcinoma: Lahey Clinic experience from 1963 through 1969. Am J Surg 1976; 131: 434-41. 13. Heberer G, Denecke H, Pratschke E, Teichmann R. Anterior and low anterior resection. World J Surg 1982; 6: 517-24.
14. Bailey HR, La Voo JW, Max E, Smith KW, Butts DR, Hamp ton JM. Single-layer polypropylene colorectal anastomosis: experience with 100 cases. Dis Colon Rectum 1984; 27: 19-23. 15. Templeton JL, McKelvey STD. Low colorectal anastomoses: an experimental assessment of two sutured and two stapled techniques. Dis Colon Rectum 1985; 28: 38-41. 16. Graffner H, Anderson L, Lijwenhielm P, Walther B. The healing process of anastomcsis of the colon: a comparative study using single, double-layer or stapled anastomosis. Dis Colon Rectum 1984; 27: 767-71. 17. Olsen BG, Letwin E, Williams HTS. Clinical experience with the use of a single layer intestinal anastomosis. Can J Surg 1968; 11: 97-100. 18. Luchtefeld MA, Milsom JW, Senagore A, Surrell JA, Mazier WP. Colorectal anastomotic stenosis: results of a survery of the ASCRS membership. Dis Colon Rectum 1989; 32: 733-6. 19. Hastings JC, Van Winkle W, Barker E, Hines D, Nichols W. Effect of suture materials on healing wounds of the stomach and colon. Surg Gynecol Obstet 1975; 140: 701-7. 20. Adloff M, Amoud JP, Beehary S. Stapled versus sutured colorectal anastomosis. Arch Surg 1980; 115: 1436-8. 21. Bailey HR, LaVoo JW, Max E, Smith KW, Hampton JM. Single layer continuous cclorectal anastomcsis. Aust N Z J Surg 1981; 51: 473-6. 22. Skakun GB, Rexnick RK, Bailey HR, Smith KW, Max E. The single-layer continuous polypropylene colon anastomosis: a prospective assessment using water-soluble contrast enemas. Dis Colon Rectum 1988; 31: 163-8.
THE AMERICAN JOURNAL OF SURGERY
VOLUME 162 NOVEMBER 1991
467