Key points

Introduction

Colorectal cancer is the third most common cancer worldwide, accounting for approximately 600,000 deaths per annum (CRUK Cancer Stats: http://www.cancerresearchuk.org/cancer-info/cancerstats/types/bowel/mortality/ ). The potentially curative primary treatment for patients with both colon and rectal cancer remains surgery. Adjuvant treatment with chemotherapy provides additional benefit for those patients with more advanced colonic tumors.

Over the past few decades, there have been significant improvements in the treatment of patients with colonic and rectal cancers. In rectal cancer, the role of earlier diagnosis, improved preoperative staging, neoadjuvant therapy, total mesorectal excision, and laparoscopic surgery have improved outcomes from oncological and patient recovery perspectives. Of these, arguably the most important for the surgeon was the advent of total mesorectal excision (TME) by Heald and Ryall.

Current thinking is that colonic tumors spread via hematogenous, lymphatic, and possibly perineural routes, with the lymphatics anatomically following the arterial supply. Current practice is to excise a proportion of the draining lymphatic bed to accurately stage the cancer and also clear possible lymphatic metastases. Recently, significant debate has centered on the degree of lymphatic clearance required; several reports have demonstrated improved oncologic outcomes with wider lymphovascular resections compared with current standard practice. Whether these improved outcomes are secondary to improved lymph node yield or an alternative technical effect has not yet been ascertained.

Introduction

Colorectal cancer is the third most common cancer worldwide, accounting for approximately 600,000 deaths per annum (CRUK Cancer Stats: http://www.cancerresearchuk.org/cancer-info/cancerstats/types/bowel/mortality/ ). The potentially curative primary treatment for patients with both colon and rectal cancer remains surgery. Adjuvant treatment with chemotherapy provides additional benefit for those patients with more advanced colonic tumors.

Over the past few decades, there have been significant improvements in the treatment of patients with colonic and rectal cancers. In rectal cancer, the role of earlier diagnosis, improved preoperative staging, neoadjuvant therapy, total mesorectal excision, and laparoscopic surgery have improved outcomes from oncological and patient recovery perspectives. Of these, arguably the most important for the surgeon was the advent of total mesorectal excision (TME) by Heald and Ryall.

Current thinking is that colonic tumors spread via hematogenous, lymphatic, and possibly perineural routes, with the lymphatics anatomically following the arterial supply. Current practice is to excise a proportion of the draining lymphatic bed to accurately stage the cancer and also clear possible lymphatic metastases. Recently, significant debate has centered on the degree of lymphatic clearance required; several reports have demonstrated improved oncologic outcomes with wider lymphovascular resections compared with current standard practice. Whether these improved outcomes are secondary to improved lymph node yield or an alternative technical effect has not yet been ascertained.

Current surgical practice for colonic resection

The traditional approach to surgical colon cancer resection involves removal of the primary tumor with adequate proximal and distal resection margins, and a clear circumferential resection margin (which may require en bloc resection of the abdominal wall or other viscera) together with an anatomically defined mesenteric lymphovascular pedicle. These operations may be performed via either a traditional open approach or laparoscopically. It has now been shown by a wide number of studies, including large randomized controlled trials (RCT), such as COST, CLASICC, and COLOR, that oncological outcomes from laparoscopic colonic surgery are equivalent to open surgery. The necessity to include resection of the lymphovascular mesentery is based on the tenet that in addition to hematogeneous spread, colonic tumors most commonly spread initially via the lymphatic system, which anatomically follows the colonic arterial supply. Historically it is held that en bloc lymph node resection is necessary not only for staging (Cady-Fisher) but also to reduce tumor burden (Halsted). The Cady-Fisher model of cancer progression proposes that systemic spread occurs as an early event in cancer development, whereas the Halsted theory suggests a more stepwise progression in the development of metastases. More contemporary views on the role of lymphadenectomy have seen it purely for staging purposes, the results of which effect management by defining adjuvant treatment options. Generally speaking, adjuvant chemotherapy is reserved for those patients with lymph node involvement (stage III/IV) or poor prognosis stage II cancers. Inadequate assessment of lymph nodes for malignancy will theoretically lead to understaging, resulting in increased mortality through undertreatment. Understaging may be consequent on either the surgeon not removing enough lymph nodes or the pathologist not identifying and examining all lymph nodes present in the specimen. In relation to the latter, various options have been evaluated to increase the identification and assessment of lymph nodes, including fat dissolution chemicals and ex vivo sentinel lymph node (SLN) identification. With respect to the operative harvesting of nodes, it is recommended to perform a “high tie” of the vascular pedicle to maximize the number of lymph nodes within the colonic mesentery.

What is the evidence to support current colonic cancer surgical practice?

It is important to identify the evidence for current practice in colonic cancer surgery. As with most other solid organ malignancies, primary treatment is surgical. In the past, there has been a reluctance to use neoadjuvant therapy in colonic cancer surgery because of concerns over accurate radiological staging and the risk of bowel obstruction during treatment. However, recently published results from the FOxTROT trial show that this option is feasible and safe, and may potentially induce downstaging of disease. In this randomized controlled trial, patients with radiologically staged T3 and T4 colonic tumors were randomized to either preoperative chemotherapy and then surgery or standard postoperative chemotherapy. Early recently reported data indicate low levels of chemotherapy side effects, equivalent levels of postoperative morbidity, and significant downstaging, including 2 reports of pathologic complete response. The trial is now being extended to detect meaningful oncological outcomes at 2 years. Currently there is insufficient evidence to support the routine use of preoperative neoadjuvant chemotherapy in colonic cancer outside of a trial environment.

The tumor must be completely resected; other than isolated case reports in the palliative setting, there is no evidence to support the role of debulking surgery in colon cancer. Because of the high mobility of the colon, resection margins are determined by the degree of lymphovascular dissection required rather than distance from the tumor per se. In general, a distal margin of 5 cm is deemed the minimum necessary, because of the possibility of intramural spread. There is no evidence to support local resection of colon cancer via colotomies. There is only one randomized trial comparing oncological outcomes in “segmental colectomies” versus hemicolectomies. Segmental colectomy was defined as a localized, not pedicle-based resection of the intestine and this was compared with a traditional pedicle-defined hemicolectomy. In this prospective trial, survival rates were equivalent between patients with left-sided cancers treated with segmental colectomy compared with hemicolectomy. Interestingly, other contemporary retrospective studies have replicated these findings, demonstrating equivalent outcomes comparing segmental colectomies and hemicolectomies for left-sided cancers. These isolated data should be interpreted with caution, and on a background of understanding the importance of lymphadenectomy and patterns of lymph node spread. Recent data reinforcing current practice for wide lymphadenectomy and therefore hemicolectomy demonstrates that early lymphatic spread occurs far wider than the juxtatumoral lymph nodes.

The original data suggesting the importance of removing potentially involved lymph nodes was published by Gilchrist and David in 1938. It has since been shown that in stage II and stage III disease, high lymph node yield is positively correlated with survival. Interestingly, this study also showed that irrespective of involvement or not, the number of lymph nodes analyzed was an independent predictor of outcome even in node-negative cases. The investigators of this study concluded that the surgeon alone may be an important controlling variable. These data and similar studies were reviewed in 2007 by Chang and colleagues, who made broadly similar conclusions. The association of outcome with lymph nodes has been further dissected to reveal a deeper association with that of the ratio of metastatic to total number of identified lymph nodes. Analysis revealed that lymph node ratio was a determining factor for overall survival, disease-free survival, and cancer-specific survival if more than 10 lymph nodes were removed. These and many other studies have been used to justify the use of lymph node yield as a surgical end point and surgical quality indicator. As discussed, this is generally held to be important to adequately stage the tumor, thereby preventing understaging, although this direct causation has never been formally shown. From several large studies it is proposed that a minimum of 12 lymph nodes need to be examined to accurately stage the cancer. To acquire and examine adequate numbers of lymph nodes, a complete lymphovascular pedicle is removed en bloc with the specimen. These arterially defined pedicles guide the colonic resection margins to enable restoration of intestinal continuity by anastomosis.

How can we improve oncological outcomes in colon cancer surgery?

Despite some regional differences, over the past 4 decades oncological outcomes for patients with colonic cancer have significantly improved (CRUK Cancer Stats). These have occurred on a background of earlier diagnosis and improvements in adjuvant treatment. However, despite good surgical practice, large numbers of patients with potentially curable colon cancer present with local or distant recurrence following stage I and stage II resections. It has been suggested that circulating tumor cells, micrometastases, and tumor cells residing in immune-privileged areas, such as distinct areas of the bone marrow, may account for some of the unexpected recurrences; it is also possible that surgical practice could be improved. The fundamental issue is whether more aggressive resection might result in more accurate staging, and better oncological outcomes. The newly described technique of complete mesocolic excision (CME) with central vascular ligation (CVL) provides this advantage. The CME approach, from a lymph node retrieval perspective per se, is no different from what is currently done with a “high-tie.”

SLN biopsy is used routinely in breast cancer and melanoma surgery. The guiding principle in the technique is that lymph node spread is a marker of systemic disease (Cady-Fisher) and that sensitivity can be improved by sampling solely the first node that a tumor may spread to: the SLN. Should this node be found positive for metastases, adjuvant treatment is necessary. Several studies have attempted to identify whether SLN biopsy can improve identification of lymphatic spread in colorectal cancer above what is currently being performed. The identification of the node is performed in vivo or ex vivo by injecting blue dye, technetium-labeled colloid (99mTc), or indocyanine green fluorescent dye into the normal adjacent bowel. The results from SLN identification in colorectal cancer have been mixed. Two recent systematic reviews of the technique have suggested that the technique may improve staging at the expense of increased workload for either surgeon or pathologist.

Experimental approaches to improve staging include serum sampling either before or after surgery to quantify circulating tumor cells (CTC) or tumor DNA and intraoperative bone marrow sampling. There is evidence to support the relationship of both CTCs and bone marrow tumor cells with outcome in colorectal cancer, but reliable and sensitive methodology is not currently available. Identification of circulating tumor DNA may be more promising; it is reported to have greater sensitivity and, by offering individualized mutational analysis, this technique may facilitate targeted adjuvant therapy.

CME and CVL: background and supportive evidence

Rectal cancer surgery was revolutionized by the work of Bill Heald, who reported in 1986 that local recurrence rates could be vastly improved by using the technique of TME. This technique not only removes the primary rectal cancer with an adequate circumferential resection margin, but also removes the mesorectum. The technique of TME improved local recurrence rates from 30% to 40% to as low as 3.7% and is now regarded as the “gold standard” in rectal cancer surgery.

Based on the TME experience, the group from Erlangen in Germany have advocated for CME in conjunction with CVL for colon cancer. CME is reported to differ from traditional colon cancer surgery by achieving a far more radical excision of the lymphovascular pedicle and mesocolon. In addition, the CME technique promotes resection of the specimen with an intact visceral peritoneum together with proximal and distal resection margins of at least 10 cm. Arterial supply to the affected segment of bowel is taken at its origin from the superior mesenteric artery (right and transverse colon) and the aorta (left colon), described as CVL. CME has been shown to lead to increased lymph node harvest and more mesocolic tissue. In a comparison between the Leeds and Erlangen units, it was shown that CME led to an almost doubling in both the number of lymph nodes retrieved and area of mesentery resected. However, a Danish study showed only a 9% increase in lymph node yield.

The Leeds group has since gone on to show that mesocolic plane resections alone result in improved 5-year survival outcomes, most apparent in stage III resections. The Erlangen group has also shown that CME improves 5-year survival rates and locoregional recurrence rates for their patients. The improvements are not as marked as with TME, but the improvement in local recurrence, from 6.5% to 3.6%, and in 5-year survival, from 82.1% to 89.1%, are striking.

CME has been shown to be technically feasible in both open and laparoscopic colon surgery. However, a Greek study of 90 patients (41 open and 49 laparoscopic) showed that laparoscopic CME resulted in a marginally shorter distance from tumor to the high tie and slightly fewer lymph nodes in the specimen. An RCT with long-term follow-up remains to be performed to ascertain whether laparoscopic CME can be recommended.

Whether CME is any different from traditional practice in which “high ties” of vascular pedicles are advocated may be debatable. Some of the concepts described in CME are not new; both Enker and colleagues and Turnbull and colleagues stressed the importance of radical lymphadenectomy in improving oncological outcomes. For left-sided tumors, the dissections described by the Erlangen group appear generally similar to traditional descriptions. For right-sided and transverse colon tumors, the Erlangen description involves a much higher tie and dissection, including full mobilization of the duodenum, head of pancreas, and, on occasion, dissection of the gastroepiploic arteries. Interestingly, the reported Erlangen complication rates from their more radical dissections are no different than the traditional approach. Others have identified an increase in genitourinary complications, including ejaculatory dysfunction, with CVL for left-sided tumors. The outcomes of the Erlangen group have since been replicated by others, suggesting that high vascular ligation shows oncological benefit in both local recurrence and 5-year mortality rates. It has also been shown that the practice of CME can be standardized, taught, and implemented with reproducible results.

The apparent improved outcomes with CME are yet to be confirmed with a formal RCT. Proposed explanations for the apparent improvements are that increasing lymph node yield permits stage migration, that increased lymph node yield removes a source of metastases, and that it has nothing to do with lymphatics but is due to the preservation of an intact peritoneum.