Complications after colorectal resections during peritonectomy and HIPEC in advanced peritoneal carcinomatosis from ovarian cancer.

INTRODUCTION The current options for treating patients with primary or recurrent diffuse ovarian carcinomatosis include peritonectomy and hyperthermic intraperitoneal chemotherapy (HIPEC). The main role of peritonectomy in this integrated procedure is to achieve maximal cytoreduction by multiple parietal and visceral resections, whereas HIPEC serves to sterilize microscopic or millimetric residual sites of tumor. Among the various visceral resections needed for maximal cytoreduction, colorectal resections account for nearly 50%. Despite consensus on the oncologic appropriateness of colorectal resections to achieve optimal cytoreduction, technical controversies persist and information is lacking on how these procedures influence outcome and survival, AIM: to identify a reasonable surgical strategy for colorectal resections to reach optimal cytoreduction and minimize operative risks. METHODS From a series of 70 patients prospectively enrolled from November 2000 to April 2009 in a single-center phase-II study on the use of peritonectomy and HIPEC (closed technique at the end of surgery) in the treatment of diffuse primary or recurrent peritoneal carcinomatosis from ovarian cancer we selected for this study all 52 consecutive patients who also underwent colorectal resection. • Surgical Technique for Peritonectomy The extent of peritoneal carcinomatosis was classified according to the peritoneal cancer index (PCI). Aggressive surgical cytoreduction to leave the patient with no visible disease then proceeded in three stages: treatment of the parietal peritoneum, visceral resections and lymphadenectomy •Surgical Technique for Colorectal Resection Involvement of the pelvis the and cul-de-sac along as well as the uterus and adnexa or recurrent disease of the pelvis ! en bloc resection of the internal genitalia or pelvic recurrence along with the rectum and sigmoid colon (TME). Right iliac fossa carcinomatosis involving the cecum, appendix, terminal ileum or ascending colon ! standard right hemicolectomy. Involvement of the pelvis and all colonic segments, with nodules penetrating deeply into the colonic wall ! total colectomy, rectal resection and terminal ileostomy. We generally preferred to construct an ostomy and postpone restoring intestinal continuity for a second look. In most patients, before restoring intestinal continuity we waited for at least 6 months after post-peritonectomy systemic chemotherapy. The completeness of cytoreduction (CC) was scored as proposed by Sugarbaker. Statistical Analysis A multiple regression test was used to analyze the influence of morbidity and mortality risk factors on patient’s outcome. The Kaplan-Meier method was used to construct survival curves and the log-rank test was used to assess the significance of differences between curves. The Cox regression model was used to determine the prognostic value of independent variables. P values <0.05 were considered to indicate statistical significance. The NCSS package was used to analyze the data base and perform statistical tests. DISCUSSION At a mean follow-up of 30.2 months (range 4-79), the estimated mean survival was 33.2 months and the mean disease- free survival was 27 months. In 74.3% of the cases (52/70 patients) a colorectal resection was needed to achieve satisfactory cytoreduction levels. Our experience in these patients therefore suggests that rectal resection alone or associated with other colonic resections is the crucial surgical step on which patients’ outcome depends. The major anatomic and pathological prognostic factors reflecting clinical outcome were colorectal wall involvement and CC score. In contrast to most investigators, a technical point we underline is the need for a low rectal resection leaving a rectal stump no longer than 5 cm completely removing the mesorectum. We also used inferior mesenteric artery ligation at its origin from the aorta (high tie) and the inferior mesenteric vein at the inferior pancreatic border including a large amount of the mesocolon and adequate lymphadenectomy. In most patients in our series (75%) the tumor infiltrated the muscular layers up to the mucosa; in 25% the tumor involved only the intraperitoneal rectal or colonic wall serosa, the peritoneal pouch or mesorectum without infiltrating the muscular layers. As many as 22 of the 52 patients (42.3%) undergoing colorectal resection in our series had mesenteric lymph-node metastases alone or in association with typical ovarian node metastases and 20 (41.6%) of patients who underwent rectal resection had mesorectal lymph-node metastases. This pattern of malignant spread shows a direct relationship between infiltration of the colorectal wall and mesenteric lymph-node metastases and suggests that optimal surgical management of these patients must include the resection procedures commonly used for primary large bowel carcinoma. Even though some investigators underline this concept, the appropriate surgical management of large bowel involvement in primary and recurrent diffuse peritoneal ovarian carcinomatosis in practice remains unapplied. In this scenario, for example, some resect a limited rectosigmoid segment (15 cm) constructing the colorectal anastomosis high (9-10 cm from the anal verge) without excising the entire mesorectum. Many also provide poor or no information on mesenteric lymphadenectomy and when they supply information resect a mean 5 lymph nodes, inadequate for oncologic exeresis. Besides, hardly surprisingly, this sleeve fashion resection leads to a high percentage of microscopic residual disease on the colorectal stump (20%). Hence in our opinion, mistakenly, more emphasis is placed on restoring intestinal continuity with a colorectal anastomosis without a colostomy than on observing the necessary oncologic rules. After rectal resection we generally avoided restoring intestinal continuity immediately and according to the entity of colorectal resection construct a colostomy or ileostomy. We postpone restoring intestinal continuity until after systemic postperitonectomy chemotherapy ends and after a further 6 months follow-up for patients who remain disease free. This strategy minimizes the numerous operative risks in critically ill patients, many of whom have intestinal obstruction (30%), all of whom have diffuse carcinomatosis (mean PCI 18), more than 50% of whom also require intestinal anastomoses or local excision of tumor implants from the large and small bowel wall, and who have generally suffered a mean blood loss of 1700 mL, and finally, all of whom have to undergo HIPEC. This strategy also has the distinct therapeutic advantage of allowing second-look surgery, especially in apparently disease-free patients with low tumor markers. Our unpublished experience shows that 7 of the 12 patients who underwent surgery to restore intestinal continuity had minimal recurrent disease that was resected during reconstruction surgery. These 7 patients also underwent a second HIPEC procedure. Of the 12 patients who underwent bowel reconstruction, 2 with a coloanal anastomosis, 7 with a colorectal stapler anastomosis, and 3 patients who underwent total colectomy with ileostomy had an ileorectal anastomosis with a J pouch. Three other patients available for reconstruction refused a new operation. Late reconstruction is particularly safe because the first operation improves the patient’s general conditions and leaves a stiff, healthy rectal stump. All patients who underwent reconstruction except one in whom a rectovaginal fistula developed, had an uncomplicated postoperative course. Comparing patients undergoing cytoreduction with and without HIPEC, Ryu et al. observed higher rates of intraabdominal complications such as intestinal perforation, intestinal obstruction, and sepsis in patients who underwent HIPEC. In our series only two patients, both of whom had only mild pelvic carcinomatosis, had colorectal anastomoses, all the others had colostomy or ileostomy, depending on the extent of colonic resection. In our series the mean PCI was relatively high (mean 18.8) and colorectal resections were invariably needed to achieve optimal cytoreduction. When we investigated the prognostic role of colorectal resection in the outcome of our patients our findings again underline growing evidence on the role of maximal cytoreduction and clearly show that colorectal resections are merely a step in this process similar to the other visceral or parenchymal resections necessary and unavoidable to reach radical cytoreduction. Not with standing the small number of patients studied, our results seem to suggest as a major negative pathologic factor colorectal wall involvement reaching the mucosal layer given that none of these patients in our series survived. This new finding implies that whenever diagnostic procedures identify these high risk patients surgery might usefully be preceded by neoadjuvant chemotherapy. Conclusion Colorectal resection are an unavoidable step in achieving maximal cytoreduction and hence in improving outcome and survival. The high rate of deep infiltration into the colorectal wall along with the pericolic, mesenteric and mesorectal lymphonode metastases suggest that colorectal resections should follow the strict oncologic rules applied for primary cancer. To minimize the operative risks, after rectal resection anastomosis should be postponed for at least 6 months in patients who remain disease-free after systemic post-peritonectomy CHT ends, that allows second-look surgery in appatently disease-free patients.