By William B. Ershler, MD, Editor
SYNOPSIS: In a well-controlled retrospective analysis of patients who presented with metastatic colorectal cancer, primary tumor resection was associated with improved overall survival. By careful multivariate analysis controlling for confounding clinical variables that might result in selection bias (age, performance status, disease burden, etc.), primary tumor resection remained significantly associated with improved survival.
SOURCE: Ahmed S, Leis A, Fields A, et al. Survival impact of surgical resection of primary tumor in patients with stage IV colorectal cancer. Results from a large population-based cohort study. Cancer 2014;120:683–691.
Surgical resection of primary colorectal tumors in patients who present with stage IV disease is commonly undertaken,1 but without firm evidence for benefit in terms of survival.2-5 Despite a recent meta-analysis of 15 reported observational studies that revealed a 31% reduction in mortality (hazard ratio [HR] = 0.69, 95% CI = 0.61-79) with surgical excision of the primary tumor and an absolute difference in median survival of approximately four months,6 there remains concern because of potential selection bias. Certainly, younger patients with fewer comorbidities and better functional status would seem more likely to receive surgery, and short of a prospective randomized trial, these concerns will linger. In the current report, Ahmed and colleagues from Saskatchewan attempt to address the selection bias issue by examining a large cohort of newly diagnosed stage IV colorectal cancer (CRC), controlling for several of these potentially confounding variables and assessing survival in those who did or did not receive surgical resection of the primary tumor (SRPT).
This was a large retrospective, population-based cohort study including patients with stage IV CRC diagnosed between 1992 and 2005 in the province of Saskatchewan, Canada. Survival was estimated by using the Kaplan-Meier method and compared by log-rank test. Cox proportional multivariate regression analysis was performed to determine survival benefit of SRPT by controlling for recorded prognostic variables.
Of a total of 1378 eligible stage IV CRC patients, 944 (68.5%) underwent SRPT. For the whole cohort, the median age was 70 years (range, 22-98 years), the male:female ratio was 1.3:1, and 29.5% had rectal or rectosigmoid tumor (rectal, 20.1%, rectosigmoid, 9.4%). A total of 544 patients (39.5%) were symptomatic, primarily from obstruction (83%), bowel perforation (16%), or major bleeding (10%). Of the total, 1038 (75.3%) had liver metastases and 698 (50.7%) had extrahepatic metastases. Among 698 patients with extrahepatic disease, 217 (31%) had lung metastases, 205 (29.3%) had peritoneal involvement, 31 (4.4%) had bony metastases, and nine (1.3%) patients had documented brain metastases.
Of the 1378 patients, 42.3% received chemotherapy. For those who received SRPT and chemotherapy, the median overall survival was 18.3 months (95% CI = 16.6-20 months), compared to 8.4 months (95% CI = 7.1-9.7 months) for those treated with chemotherapy alone (p < .0001).
On univariate analysis, a number of clinicopathological factors were correlated with survival, including ECOG performance status > 1, high CEA, advanced age (poor survival), and use of chemotherapy (better survival). Tests for interaction between surgical resection of primary tumor and age, performance status, CEA level, second-line therapy, or more than one metastatic site were significant. In the proportional hazard model, SRPT was associated with better survival in younger patients, patients with good performance status, normal CEA level, patients treated with second-line therapy, and patients with one metastatic site. By multivariate regression analysis, the use of chemotherapy, SRPT, and surgical resection of metastases were correlated with a favorable survival, whereas older age, poor performance status, low albumin, elevated bilirubin, elevated alkaline phosphatase, anemia, leukocytosis, colonic primary (as opposed to rectal or rectosigmoid), and grade 3 tumor were correlated with inferior survival. After controlling for these clinically important variables, only the interactions between the SRPT and second-line chemotherapy, or more than two metastatic sites, were significant. After adjusting for other important prognostic variables in a Cox proportional multivariate model, the HR for survival with surgical resection of primary tumor was 0.54 (95% CI = 0.48-0.62).
Several uncontrolled studies had previously demonstrated a survival benefit for SRPT in the management of stage IV CRC, but it is difficult to generalize from these reports because of selection bias. The current analysis is of a population-based cohort sufficiently large to allow for control of several potentially confounding variables (e.g., age, functional status, comorbidities). The authors found that despite significant differences in the baseline characteristics between the operated and non-operated groups when the known prognostic variables were included in a multivariate model, resection of the primary tumor remained an important prognostic factor. SRPT was associated with 51% relative reduction in mortality after adjustment for age, performance status, comorbid illnesses, chemotherapy, excision of metastases, use of newer chemotherapy regimens, and disease burden. The authors found significant survival differences between the two groups at large (resection versus non-resection), or when examining various subgroups (e.g., those who were asymptomatic or minimally symptomatic at presentation), with the surgical intervention group having significantly better survival in each analysis (range, 7.6-13.6 months).
Clinical oncologists commonly recommend a surgical approach for patients who present with symptomatic disease (i.e., obstruction, perforation, or bleeding), but the current data would suggest that an operative approach should be considered in all, even those who present without symptoms referable to the primary tumor. Under such circumstances, it is unclear why removing the primary tumor would provide survival advantage. It may relate to reduced tumor burden, avoidance of local complications, or an abscopal effect as occasionally observed in patients with renal carcinoma or malignant melanoma. However, despite the large sample size and careful methodology, the current findings are retrospective and need confirmation. In this light, we await the findings from a randomized study currently underway in the Netherlands (CAIRO 4) examining the role of primary resection in patients presenting with metastatic CRC.