MULTIVISCERAL RESECTIONS IN COLORECTAL CANCER. Cleveland Clinic Symposium 2014

MULTIVISCERAL RESECTIONS IN COLORECTAL CANCER

Authors:

Sérgio Carlos Nahas 1

 Leonardo A. Bustamante L.2

Caio Sérgio R. Nahas 3

1- FACS, MASCRS, Professor of Colorectal Surgery. Chief, Section of Colorectal Surgery. Chief of Residency Program, Hospital das Clínicas - University of São Paulo. Chief, Section of Colorectal Surgery , Instituto do Câncer do Estado de São Paulo. University of São Paulo School of Medicine.

 2- Clinical Fellow of Colorectal Surgery, Hospital das Clínicas - University of São Paulo.

 3- FACS, Assisting Professor of Colorectal Surgery, Cancer Institute - University of São Paulo.

Complete surgical resection is critical to achieve long term survival in colorectal cancer.1 About 12% of all colorectal cancers (CRC) are adhesive to neighboring organs or are even invading them. When staging predicts invasion of adjacent pelvic organs (T4a stage disease) en bloc resection of involved organs is required to avoid tumor dissemination and positive resection margins. The presence of an involved margin after rectal cancer surgery is associated with local recurrence and poor survival.2 Accordingly, guidelines have recommended en-bloc multivisceral resections for treatment of these tumors, since most studies have shown that this procedure improves the possibility of R0 margins associated with a better local control and improved overall survival 3. Multivisceral resection (MVR) is often indicated during the operation due to the lack of preoperative exact data. Moreover, multivisceral resections are associated with high perioperative mortality and morbidity. During operative procedure it is often impossible to distinguish true invasion from inflammatory adhesions. Therefore, these resections are challenging even for an experienced colorectal surgeon.4 The en bloc resection of the tumor is of pivotal importance and was shown to be associated with a highly significant improvement in 5-year survival. These advanced tumors are challenging to treat, as patients may present with obstruction, fistula and pain.5 Multimodal therapy with complete resection of disease, including en bloc resection of affected adjacent organs, is the standard management. Previous studies reported 5-year survival rates of 36–53 % and local recurrence rates of less than 20 %. In the majority of these studies, achieving an R0 resection was the most important prognostic factor.6,7 A recent meta-analysis demonstrated no improvement in survival and local recurrence when an extended lymphadenectomy was performed compared with standard total mesorectal excision. However, when lymphatic spread is suspected either clinically or radiologically, extended lymphadenectomy is warranted to obtain an R0 resection. 8 The range for MVR goes from maximally invasive exenterative surgery to more limited operations, for example en bloc partial cystectomy.

In 2008 in the United States it was found that the majority of patients with locally advanced CRC did not receive MVR, despite improved survival with MVR. 9 There is a lack of randomized controlled trials about MVR , and much of the evidence for radical resections is by necessity based on small case series and retrospective studies. In our experience from the Cancer Institute of Hospital das Clínicas of University of São Paulo Medical School (ICESP/HCFMUSP), between January 2009 and September 2013, 112 procedures were done. A thorough preoperative staging workout was done: CEA levels, CT scans, specific MRI, sometimes PET, CT angiography, CT urography. As well, the nutritional and psychological status of the patient was assessed. It was followed by multidisciplinary discussions involving radiology, oncology, radiotherapy, urology, gynecology, vascular surgery, orthopedics and plastic surgery. This was done to confirm the resection possibilities, to foresee difficulties and impossibilities, types of reconstruction and oncologic type of surgery - radical or palliative. Median age was 63 years old (23-86), and 77 (68.7%) patients were females. In a systematic review, 87.0% of MVR were performed for primary Colorectal Cancer,10 in our study primary tumors were from rectum in 80%, 9% from colon, and 11% from anus. Patients with anal and rectal cancer received preoperative chemoradiation therapy in 100% and 45%, respectively. Most (92%) of our procedures were laparotomies. The correlations between cT4 and pT4 was 53.5%, as shown in previous reports.10,11 In four cases, no malignancy could be found in the specimens, despite histological confirmation before neoadjuvant treatment. The organs most frequently resected were ovaries and annexes (37%), followed by uterus (30%), vagina (26%), bladder (21%), and prostate (13%). Sacrum, small bowel, ureter and others were resected in few cases. In others papers the bladder was the most commonly involved organ (53.2 %). 7,10 Besides the primary, two or more organs were excised in 43% of the procedures. Clear margins (R0 resection) were obtained in 81 patients (72%). The complete resection rate of 72 per cent is within the range (70–82 per cent) for primary rectal malignancy in contemporary series. R0 resection was a positive prognostic factor in several studies. A meta-analysis and single-center study demonstrated a significantly reduced overall survival for patients in whom a transection of the tumor from adhesive structures was attempted. 12,13 Complete pathologic response (pT0N0) was observed in 4.5% of rectal tumors. 24% of patients with rectal tumors needed sphincter resection with terminal colostomy. 41% of patients submitted to abdominoperineal resection had the perineal defect closed by gluteal-fold flaps. Median blood transfusion was 1.7 units (0-6), length of stay was 11 (6-25) days. MVR for CRC morbidity rates are between 11% and 49%.10,14 The most common complications reported were wound infection (16.0 %), bowel obstruction or ileus (6.7 %), urinary complications (6.2 %), intra-abdominal abscess (5.4 %), anastomotic leak (3.7 %), eventration or dehiscence (2.1 %), intestinal fistula (1.9 %), bleeding (1.6 %), and urinary fistula (.8 %). Other complications recorded included leg weakness, medical complications including cardiac and pulmonary morbidity, and venous thromboembolism.10,14 Our overall morbidity rate was 45 per cent. Wound infection was the most frequent: 18%. It was followed by ileus 9%, urinary complications 5%, intraabdominal abscess 3%, anastomotic leak 2%, abdominal wall dehiscence (partial or complete) 1.5%, intestinal fistula 1% and bleeding 1%. 22% of patients suffered moderate complications and 10 per cent developed significant complications requiring intervention. While MVR is associated with a significant morbidity rate, perioperative mortality of 4.5 % (5 patients). It was comparable with previously published data on mortality following major colorectal cancer resection, with rates varying from 0 to 13% . 5,10 Fifteen percent developed local recurrence. Mortality was 23% in 27 (5-57) months follow up period. In a systematic review, the weighted mean overall 5-year survival rate was 50.3 %. 10 Direct comparison with other studies is difficult because of the inclusion of emergency patients once we are a public institution. Moreover we had a high proportion of rectal cancer. It is also important to highlight the inclusion of palliative and emergency resections in this analysis and the immense variability of presentation of these very advanced cases. Those are some of the reasons why prospective randomized trials are very difficult to be conducted in MRV. They are mostly treated as individual cases and depend on the expertise and skills of the whole multidisciplinary and well organized surgical team. Our results show clearly that they should not be neglected as inoperative and left hopelessly. Conclusion Multivisceral resection employed for colorectal tumors with invasion of adjacent organs is characterized by high complexity, requiring multidisciplinary approach for preoperative evaluation, surgical procedure and postoperative care. It should be performed in high volume specialized centers and in selected patients. MVR for locally advanced CRC provides the possibility of long-term cure. The best chance of cure is at the initial operation if R0 resection can be achieved. Despite advances in multimodal imaging techniques, the need for MVR sometimes is identified preoperatively. In fact, more than 20 % of MVRs were ‘‘negative’’ on pathological analysis. MVR is not an ideal option purely for palliation, given the morbidity associated with the procedure.