Low colorectal anastomosis after radical pelvic surgery: A risk factor analysis

Low colorectal anastomosis after radical pelvic surgery: A risk factor analysis Ramin Mirhashemi, MD, Hervy E. Averette, MD, Ricardo Estape, MD, Roberto Angioli, MD, Reza Mahran, MD, Luis Mendez, MD, Guilherme Cantuaria, MD, and Manuel Penalver MD Miami, Florida OBJECTIVE: This study was conducted to analyze our experience with low (8-12 cm above the anal verge) and very low (5 cm above the anal verge. We believe that this resulted in improved vascularity of the anastomosis and therefore in fewer breakdowns. In conclusion, we were not able to document an improved healing rate of the anastomosis among irradiated patients with a protective colostomy, such as has been previously suggested by Berek et al15 and Harris and Wheeless.4 Patients who had a protective colostomy did, however, have a more benign postoperative course. Furthermore, an omental wrap around the anastomosis may be important in the prevention of breakdowns, presumably by improving vascularity. Finally, a protective colostomy appears to be unnecessary for patients with low colorectal anastomosis in the presence of significant ascites. In general, one can expect that patients who undergo low colorectal resection as part of a posterior exenteration will have lower complication rates of breakdowns and fistulas than will those who undergo very low colorectal resections as part of a total pelvic exenteration (2/37 vs 13/40). REFERENCES

1. Arenal JL, Benito C, Concejo MP, Ortega E. Colorectal resection and primary anastomosis in patients aged 70 and older: prospective study. Eur J Surg 1999;165:593-7. 2. Dehni N, Tiret E, Singald DJ, Cunningham C, Schlegel DR, Guiguet M, et al. Long-term functional outcome after low anterior resection. Dis Colon Rectum 1998;41:817-23. 3. Fegiz G, Angelini L, Bezzi M. Rectal cancer: restorative surgery with the EEA stapling device. Int Surg 1983;68:13-8. 4. Harris WJ, Wheeless CR Jr. Use of the end-to-end anastomosis stapling device in low colorectal anastomosis associated with radical gynecologic surgery. Gynecol Oncol 1986;23:350-7. 5. Burnett AF, Potkul RK, Barter JF, Barnes WA, Delgado G. Colonic surgery in gynecologic oncology: risk factor analysis. J Reprod Med 1993;38:137-41. 6. Hatch KD, Gelder MS, Soong S, Baker VV, Shingleton HM. Pelvic exenteration with low rectal anastomosis: survival, complications, and prognostic factors. Gynecol Oncol 1990;38:462-7. 7. Hatch KD, Shingleton HM, Potter ME, Baker VV. Low rectal resection and anastomosis at the time of pelvic exenteration. Gynecol Oncol 1988;32:262-7. 8. Hoffman MS, Lynch CM, Gleeson NC, Fiorica JM, Roberts WS, Cavanagh D. Colorectal anastomosis on a gynecologic oncology service. Gynecol Oncol 1994;55:60-5. 9. Wheeless CR Jr. Low colorectal anastomosis and reconstruction after gynecologic cancer. Cancer 1993;71:1664-6. 10. Clarke-Pearson DL, DeLong ER, Chin N, Rice R, Creasman WT. Intestinal obstruction in patients with ovarian cancer: variables associated with surgical complications and survival. Arch Surg 1988;123:42-5. 11. Penalver M, Averette H, Sevin BU, Lichtinger M, Girtanner R. Gastrointestinal surgery in gynecologic oncology: evaluation of surgical techniques. Gynecol Oncol 1987;28:74-7. 12. Terada K, Morley GW. Radical hysterectomy as surgical salvage therapy for gynecologic malignancy. Obstet Gynecol 1987;70: 913-6.

Mirhashemi et al 1379

13. Castaldo TW, Petrillin ES, Ballon SC, Lagasse LD. Intestinal operations in patients with ovarian carcinoma. Am J Obstet Gynecol 1981;139:80-4. 14. McGinn FP, Gartell PC, Clifford PC, Brunton FJ. Staples or sutures for low colorectal anastomoses: a prospective randomized trial. Br J Surg 1985;72:603-5. 15. Berek JS, Hacker NF, Lagasse LD. Rectosigmoid colectomy and reanastomosis to facilitate resection of primary and recurrent gynecologic cancer. Obstet Gynecol 1984;64:715-20.

Discussion DR C. ROBERT STANHOPE, Rochester, Minnesota. Mirhashemi et al described a retrospective analysis of 77 patients who underwent colorectal anastomosis. Forty of these patients underwent total exenteration, 36 after previous irradiation, and they had a 33% anastomotic breakdown rate. The other 37 patients who underwent anterior resection of the sigmoid colon associated with treatment for ovarian cancer (among whom only 1 patient had previous irradiation) had an anastomotic breakdown rate of 5.4%. Mirhashemi et al noted that neither protective colostomy, performance of an air test, use of reinforcing sutures for the anastomosis, nor an omental J flap offered protection. These results are not surprising, but they do lead us to further analysis of the factors related to anastomotic leakage. It is quite clear that adequate blood supply is the key and that ischemia as a result of anastomotic tension, rough handling, or the variable nature of hand sewing compared with stapled anastomosis may be important. Also of concern is circulatory status, whether from vessel ligation or vascular endarteritis after radiation therapy. Arterial circulation is of some major significance. The hemorrhoidal arteries are frequently sacrificed during either of the procedures described by Mirhashemi et al. The superior hemorrhoidal artery is ligated with the rectosigmoid resection, the middle hemorrhoidal artery is frequently ligated with transection of the rectum, and the inferior hemorrhoidal artery is ligated with anterior division of the hypogastric or internal iliac artery. A report from Milan, Italy, by Vignali et al1 provides some insight into what is occurring and perhaps where we must go to further analyze and develop a treatment plan for our patients. They prospectively evaluated 55 patients who underwent stapled straight anastomosis of the lower colon. They determined transmural colonic blood flow by means of laser Doppler flowmetry. Each of these patients underwent a baseline assessment before any manipulation of the bowel. Then the laser Doppler flowmetry was repeated after vascular ligation, and any decrease from the baseline in blood flow was calculated. Among their 55 patients, anastomotic leaks occurred in 8 (14.5%). The mean rectal stump blood flow reduction was 6.2% among patients who did not have anastomotic leaks but 16% among patients who did have anastomotic leaks, and this difference was statistically significant (P < .001). Vignali et al1 concluded that blood flow reduction at the rectal stump is associated with an increased risk of anastomotic leakage. This is objective documentation of what Mirhashemi et al have so vividly shown us.

1380 Mirhashemi et al

Perhaps further research efforts with laser Doppler flowmetry may be of value in identifying and quantifying alterations in microperfusion at the rectal stump. This could be of value in predicting which patients will be at little or great risk of anastomotic breakdown. My questions for Dr Mirhashemi are simple. (1) Were there any differences between hand-sewn and stapled anastomoses? (2) What is your plan for future patient management in response to these results? REFERENCE

1. Vignali A, Gianotti L, Braga M, Radaelli G, Malvezzi L, Di Carlo V. Altered microperfusion at the rectal stump is predictive for rectal anastomotic leak. Dis Colon Rectum 2000;43:76-82.

DR MIRHASHEMI (Closing). As gynecologic oncologists, we are frequently faced with the question of whether we should reanastomose the colon to the rectum in the face of irradiation during a supralevator pelvic exenteration. The data on this controversial topic have been somewhat vague. Our study of low and very low colorectal anastomoses is the largest to date in the gynecologic oncology literature. On the basis of our results, we do not recommend performing low or very low colorectal anastomosis in a field that has previously been treated with definitive radiotherapy. Our study indicates that a protective colostomy in the face of irradiation still resulted in a 50% breakdown rate of the anastomosis. There is no question that the radiation and the extent of pelvic devascularization during a total pelvic exenteration compromise the

December 2000 Am J Obstet Gynecol

anastomosis. I think that we would all agree about the importance of vascular integrity for the purposes of healing tissue. With respect to hand-sewn versus stapled anastomoses, there were too few hand-sewn anastomosis to realistically come to a conclusion as to whether one was better. The results from the general surgical literature seem to be more favorable for stapled anastomosis, and stapling is therefore the primary technique of choice for most surgeons. Your reference to the article by Vignali et al1 is quite interesting. The use of laser Doppler flowmetry to assess transmural colonic blood flow is an attractive concept to determine which patients are candidates for colorectal anastomosis. It would be of value to perform a pilot study to look at the use of this technique among gynecologic oncology patients, because they are unique with respect to perioperative risk factors, as discussed in our article. In conclusion, on the basis of this retrospective study we have decided to reserve the technique of stapled low and very low colorectal anastomosis for patients who have not undergone previous pelvic radiotherapy. In addition, we believe that the presence of ascites is not a contraindication for this procedure among patients who have undergone posterior exenteration for cytoreduction of ovarian cancer. REFERENCE

1. Vignali A, Gianotti L, Braga M, Radaelli G, Malvezzi L, Di Carlo V. Altered microperfusion at the rectal stump is predictive for rectal anastomotic leak. Dis Colon Rectum 2000;43:76-82.