Randomized Clinical Trials in Colon Cancer

Surgery remains the mainstay of treatment for colon cancer and surgical resection alone results in 5-year survival in more than 60% of patients. However, the use of fluorouracil (5-FU)-based adjuvant chemotherapy for patients at high risk of recurrence further prolongs disease-free survival and has become the standard of care. New areas of research focus on decreasing the surgical trauma with minimally invasive approaches, improving the surgical staging of patients with colon cancer, and improving adjuvant treatment regimens. We review those randomized controlled trials that have most impacted the clinical management of patients with colon cancer in 2009.

Significant advances have been made in the treatment of colon cancer in recent years. Surgery remains the mainstay of treatment and surgical resection alone results in 5-year survival in more than 60% of patients. However, the use of fluorouracil (5-FU)-based adjuvant chemotherapy for patients at high risk of recurrence further prolongs disease-free survival (DFS) and has become the standard of care. New areas of research focus on decreasing the surgical trauma with minimally invasive approaches, improving the surgical staging of patients with colon cancer, and improving adjuvant treatment regimens. In this article, we review those randomized controlled trials (RCTs) that have most impacted the clinical management of patients with colon cancer in 2009 ( Table 1 ).

Table 1

Key randomized controlled trials establishing role of adjuvant therapy in stage III and high-risk stage II colon cancer

Author	Year of Study	Experimental Arms	Trial Outcome	Contribution to Patient Care
Windle et al	1987	5-FU vs 5-FU + lev vs observation	Improved DSS (52% vs 32%) with 5-FU + lev	Suggested an additive effect of levamisole when used in combination with 5-FU
Moertel et al	1995	5-FU + lev vs lev alone vs observation	33% reduction in death rate with 5-FU + lev	Established benefit of adjuvant therapy in stage III colon cancer
Wolmark et al (NSABP C-03)	1993	5-FU + LV vs MOF	32% reduction in mortality and 30% improvement in DFS with 5-FU + LV	Contributed to establishment of 5-FU based adjuvant therapy as preferred regimen
O’Connell et al	1997	5-FU + LV vs observation	Improved OS (74% vs 63%) with 5-FU + LV	Established that 6 months of adjuvant therapy improves outcomes and defined a standard of care
O’Connell et al	1998	5-FU + lev vs 5-FU + LV + lev, 6 vs 12 months	Improved OS (70% vs 60%) with 6 months of 5-FU + LV + lev	Established the added benefit associated with LV; confirmed survival benefit seen with 6 months of therapy
Wolmark et al (NSABP C-04)	1999	5-FU + LV vs 5-FU + lev vs 5-FU + LV + lev	Improved DFS (65% vs 60%) favoring regimens with 5-FU + LV, no added benefit with lev	Established 5-FU + LV as the preferred adjuvant regimen

Abbreviations: DFS, disease-free survival; DSS, disease-specific survival; 5-FU, 5-fluorouracil; lev, levamisole; LV, leucovorin; MOF, lomustine, vincristine, and 5-FU; OS, overall survival.

Surgery

Surgical resection is the mainstay of treatment for colon cancer. At the time of the last chapter , only four level Ia randomized controlled surgical trials had been published in the literature. These trials provide some of the rationale supporting current surgical decision making in colon cancer. Early vascular ligation with a no-touch isolation technique has long been advocated as a means of improving outcomes. Although the only RCT to address this question failed to demonstrate a statistically significant difference with the technique when compared with standard practices, a trend toward improved DFS with the no-touch technique was observed. The no-touch isolation technique was used in another RCT evaluating the extent of dissection necessary for patients presenting with descending colon cancer. In this trial with a minimum of 5-year follow-up, survival after segmental colectomy and more radical left hemicolectomy with routine inferior mesenteric artery ligation was equivalent, indicating high ligation of the inferior mesenteric artery is not necessary. In the third trial highlighted in the last review, the question of hand-sewn versus stapled anastomosis for colon cancer was addressed. This RCT demonstrated that hand-sewn anastomosis was associated with an increased rate of radiologically detected leak. No conclusions could be drawn with regard to likelihood of local recurrence with either anastomotic technique. The final RCT evaluated the necessity of nasogastric tube usage after colon resection and concluded that the routine use of nasogastric tube drainage was unnecessary.

Four RCTs were highlighted in the last review that addressed the optimal postoperative surveillance schedule for patients with completely resected colon cancer. These trials evaluated intensive versus less intensive surveillance strategies and attempted to identify a benefit with early detection and intervention for either local or distant recurrence. Although no survival difference was observed in any of the studies, all were underpowered and therefore no definitive conclusions could be made. No new data have been published and current postoperative surveillance schedules are based on expert opinion. A large, multicenter RCT is currently ongoing which, if accrual goals are met, will be adequately powered to answer this question.

More than 24 RCTs addressing the surgical care of patients with colon cancer have been published; this includes five level Ia RCTs. Most of these address the role of laparoscopy in the treatment of colon cancer. Although it quickly became clear within the surgical community that laparoscopic colon resections were technically feasible, the oncologic merits of laparoscopy were less certain. The adequacy of lymph node staging with a laparoscopic approach, the effect of abdominal insufflation on tumor dissemination, and risk of trochar site implants were unknown. To address these concerns, a number of RCTs have been conducted comparing laparoscopic and open colon resections. The first large, single-institution RCT to report results for laparoscopy-assisted colectomy in patients with colon cancer was initially reported in 2002. In this trial, short-term outcomes such as return of bowel function and length of stay were improved in the laparoscopy cohort and cancer-related survival was also higher. This trial suggested that oncologic outcomes associated with laparoscopic surgery are at least equivalent, if not better, to those observed with open procedures.

Since this initial report, several other multi-institutional RCTs have reported both short- and long-term outcomes. On pathologic review, lymph node retrieval with laparoscopy was adequate, with 10 to 12 nodes per specimen harvested, and adequate margins were also achievable. Only the United Kingdom Medical Research Council Conventional versus Laparoscopic-Assisted Surgery in Colorectal Cancer (CLASICC) trial, which included patients with rectal cancer, noted a slightly increased rate of positive circumferential margins in the laparoscopic cohort, although this did not reach the level of statistical significance. This did not correlate to an increase in tumor recurrence with longer follow-up, however, and may have minimal clinical significance.

Long-term oncologic outcomes evaluated in trials of laparoscopic colectomy have varied, but include DFS, cancer-related survival, and time to recurrence. Most, because of power-size calculations, have been designed as noninferiority trials. No other trial has reproduced the survival benefit with minimally invasive colectomy initially reported by Lacy and colleagues. Additionally, the survival benefit observed by Lacy and colleagues was not maintained with longer follow-up. All RCTs reported to date with long-term follow-up have found laparoscopic and open resections to be equivalent with regard to survival and recurrence.

Minimally invasive surgery is associated with short-term benefits. Blood loss is less after laparoscopic resection. Additionally, patients tend to require fewer narcotics and return to bowel function is quicker; this correlates to a 1- to 2-day shorter length of stay. Despite these observations, very little difference in quality of life (QOL) has been observed between the two procedures. Three trials have used validated questionnaires to assess QOL. Short-term QOL after laparoscopy was slightly improved in two of the studies, with one study demonstrating a slight difference in global QOL and the other a benefit in social and role functioning 2 weeks following surgery. These findings were not observed with longer follow-up (2- and 3-month). The third study evaluated both short-term and long-term (3-year) QOL with no differences seen. It is noteworthy that in one of these RCTs, patients with body mass index (BMI) higher than 30 were excluded. As obese patients may be those most likely to have short-term QOL benefits from laparoscopic resection, exclusion of these patients from the trial may have diluted any potential QOL benefit. Alternatively, the failure to demonstrate a QOL benefit with laparoscopy may relate to the inherent difficulty in measuring postoperative QOL and limited responsiveness of the utilized instruments to important postoperative factors. Some drawbacks to laparoscopic resection exist, as the procedure is associated with longer operative times.

In summary, five large, multicenter trials have demonstrated the equivalency of laparoscopic and open colectomy in patients with colon cancer. Laparoscopic colectomy in the setting of colon cancer should currently be considered an acceptable and equivalent alternative to an open resection in the hands of experienced surgeons. However, little is known about the learning curve surrounding laparoscopic colectomy or how this technique (and the outcomes observed in the setting of a RCT setting) will translate into community practice.

Sentinel lymph node (SLN) biopsy has dramatically changed nodal staging for melanoma and breast cancer by allowing the identification of the first lymph node to receive lymphatic drainage, and potential metastases, from a tumor. The SLN concept has been applied to colon cancer in an attempt to enhance pathologic identification of nodal metastases. One RCT has compared the likelihood of pathologic upstaging when standard versus enhanced pathologic evaluation through SLN mapping was performed. In this trial, the false negative rate of SLN biopsy was 9.8% and enhanced pathologic evaluation upstaged 10.7% of patients; most of these patients had micrometastases. Although the clinical significance of micrometastases is currently unknown, this may represent a cohort of patients who would previously have been classified as stage II but may be at increased risk of recurrence. Unfortunately, the low false negative rate observed in this study has not been consistently reported in the literature; in four other prospective, nonrandomized multicenter trials, the false negative rate of SLN in colon cancer ranged from 7% to 54%. Therefore, further study is required to confirm both the validity of the SLN concept in colon cancer as well as the clinical relevance of micrometastases identified through this technique.

Adjuvant therapy

It is widely accepted that adjuvant chemotherapy is associated with improved survival in patients with stage III and possibly high-risk stage II colon cancer (see Table 1 ). The efficacy of adjuvant therapy, specifically 5-FU–based regimens, was first established in a multicenter, cooperative group trial evaluating 5-FU plus levamisole versus levamisole alone versus observation alone. A series of RCTs followed, which have further defined the role of 5-FU modulators such as leucovorin (LV) or levamisole, as well as the optimal duration of therapy. In NSABP C-04, the efficacy of 5-FU + LV versus 5-FU + levamisole versus 5-FU + LV + levamisole was evaluated. The outcome of this trial established 6 months of 5-FU + LV as the preferred adjuvant regimen for patients with stage III or high-risk stage II disease.

Although the benefit of adjuvant therapy for patients with stage III disease is well accepted, the role of chemotherapy for patients with completely resected stage II disease is controversial. At the time of the last review, one RCT addressing the role of adjuvant therapy in patients with stage II colon cancer had been reported but was underpowered to detect small survival differences. Three meta-analyses addressing the question were subsequently reported with conflicting findings. The first, a review of four National Surgical Adjuvant Breast and Bowl Project (NSABP) adjuvant studies, concluded that a survival advantage existed for stage II disease that was similar in scope to that for stage III disease. In contrast, the International Multicentre Pooled Analysis of B2 Colon Cancer Trials (IMPACT B2) study failed to identify a benefit in survival or recurrence. A final meta-analysis, which included only studies comparing 5-FU + LV or levamisole to observation alone, suggested a benefit in stage II disease, although smaller in magnitude than that seen in patients with stage III disease.

Two RCTs have recently been reported that were specifically designed to address the question of adjuvant therapy for stage II colon cancer. One, unfortunately, was closed early because of difficulty in accrual and, like the prior RCT, was inadequately powered. The second trial, however, is a multicenter trial that accrued more than 3000 patients who, in the opinion of each patient’s clinician, did not have a clear indication for chemotherapy. These patients consisted primarily of patients with stage II disease. The trial was powered to have an 80% chance of detecting a 5% improvement in survival. A small but statistically significant advantage to the use of adjuvant chemotherapy for stage II disease was observed. Unfortunately, central pathologic review was not a part of the study design and further exploratory analyses to identify characteristics of patients who may benefit most from adjuvant therapy are unable to be performed. Adjuvant treatment for patients with stage II colon cancer remains controversial.

Two new chemotherapeutic agents, both with efficacy in the metastatic setting, have been explored as adjuvant treatments in colon cancer. The first, irinotecan, was evaluated in a number of RCTs as an addition to 5-FU and LV. Disappointingly, in a multicenter trial coordinated through the Cancer and Leukemia Group B (CALGB), no survival benefit was associated with the addition of irinotecan to the adjuvant treatment of patients with stage III disease. These findings have been preliminarily confirmed in two other large, multicenter trials that have been presented only in abstract form; we await the final results of these trials.

The second agent considered as an addition to regimens of 5-FU and LV was oxaliplatin. Two large RCTs have been performed evaluating these regimens and long-term results reported. In both trials, 3-year DFS with the addition of oxaliplatin is improved by an estimated 5%. This benefit was observed in both stage II and III colon cancer, but the magnitude of the benefit was most significant in patients with stage III disease. As a result of these trials, oxaliplatin has been established as the new standard of care in the adjuvant treatment of patients with surgically resected stage III colon cancer.

In addition to exploring agents to add in combination with intravenous fluorouracil regimens, several RCTs have explored oral fluorouracil analogues as a means of simplifying chemotherapy administration for patients and minimizing toxicity. Capecitabine is currently the only oral agent approved by the Food and Drug Administration (FDA) for the adjuvant treatment of colon cancer and has been demonstrated to be an effective alternative to intravenous regimens. Similar findings were found with oral uracil and tegafur used in combination with LV ; this regimen, however, is not currently approved in the United States.

RCTs have been conducted evaluating the role of regional adjuvant systemic therapy as an addition to systemic therapy. Regional administrations have included both intraperitoneal and intraportal infusions. Consistently, these trials have all failed to demonstrate a benefit in survival or local recurrence with the use of regional therapies, although most were underpowered to detect small survival differences.

The prospective RCTs that are highlighted in the text that follows do not represent all of the trials that have contributed to our current knowledge of the optimal care of the patient with colon cancer. Rather, we have highlighted those trials that have served to define the standard of care in 2009. Only trials that have been published in a peer-reviewed format are considered.