Colorectal cancer (CRC) is a significant cause of mortality, with over 40,000 new cases diagnosed annually in the UK contributing to over 16,000 deaths (Bowel Cancer UK).^1,2 Up to 20% of colon cancers (CC) present as an emergency (eCC) necessitating emergent surgery.^3,4 Although eCC has been shown to be associated with poorer overall survival (OS), much discrepancy exists in the literature regarding its association with disease-free survival (DFS).⁵⁻⁷

Studies reporting the oncologic outcomes of CRC presenting as an emergency consist of heterogeneous populations of patients with colon and rectal cancers.^3,5 Colon and rectal cancers are 2 distinct entities with different molecular, clinical, pathologic, and biologic characteristics and treatment modalities.^8,4,9−11 Since the incorporation of combined multimodal treatment and total mesorectal excision the disparity in OS and DFS between colon and rectal cancer has increased.¹²⁻¹⁷ Rectal cancer patients may alter the impression of outcomes in emergency presenting colon cancer. Consequently, previous studies assessing outcomes in eCC may be flawed. Furthermore, the negative impact of eCC has previously been attributed to immediate postoperative complications with an inpatient hospital mortality of approximately 15%. Inclusion of such cases in studies assessing long-term outcomes may have overestimated the negative impact of eCC.¹⁸

The aim of the current study was to determine the association between eCC and disease-free/overall survival.

Materials and Methods

This is a retrospective comparative observational study of all consecutive patients undergoing colon cancer resection with curative intent between January 2000 and December 2010 at a single institution. Data was collected from HIPE (Hospital Inpatient Inquiry System), histology/endoscopy reports and direct contact with primary care physicians.

Data regarding patient demographics (age, gender), tumor characteristics (stage/AJCC, lymphovascular invasion, grade, size, anatomic location), details of surgery and adjuvant therapy were collected and compared between CC and eCC. The number of lymph nodes retrieved and margin status were noted and used as metrics of adequacy of oncological resection. Long-term oncologic outcomes including locoregional recurrence (LR), systemic recurrence (SR), OS and DFS were recorded.

The aim of this study was to evaluate the association, if any, between eCC and OS/DFS. Inclusion criteria were as follows: histologic diagnosis of colonic adenocarcinoma and adequate follow-up. Exclusion criteria included rectal adenocarcinoma, patients who died within 30 days of surgery and patients with inadequate follow-up. Patients with metastasis at diagnosis (stage IV disease) were excluded from the survival (univariate and multivariate) analysis. Patients were categorized as (1) emergency presenting colon cancer (eCC; e.g., obstruction, perforation, abscess) and (2) elective colon cancer (CC). Primary endpoints included overall and disease-free survival (locoregional and systemic recurrence). Difference in distribution of clinical, demographic and pathological data was evaluated using a chi-square test for categoric variables and student's t test for continuous variables. Survival/recurrence rates were plotted and compared in Kaplan-Meier estimates (eCC versus CC). Differences in distribution of survival/recurrence rates were assessed using a log-rank test. Factors significant on univariate analysis were incorporated into a forward conditional cox proportional hazards multivariate model to determine the independent association between eCC and OS/DFS. P < 0.05 was considered statistically significant. All calculations were performed in SPSS (version 15, Chicago, Illinois).

Results

Patient and tumor characteristics

A total of 439 patients underwent surgery with a diagnosis of colonic adenocarcinoma during the study period; 65 patients were diagnosed with distant metastasis (stage IV) at presentation (40 in the CC group and 25 in the eCC group). Stage IV cases were included in the analysis of clinical and demographic data and excluded from the survival analysis. A total of 439 patients underwent curative-intent surgery during the study interval. Of these cases, 97 (22.1%) presented as an emergency. The mode of presentation within the emergency group included intestinal obstruction (n = 67, 69.1%), perforation (n = 20, 20.6%) and local abscess (n = 10, 10.3%). Patient and tumour characteristics are summarized in Table 1. There were more left-sided tumors in the eCC group (splenic flexure, P = 0.017; descending colon, P = 0.004). Overall, eCC displayed features of more advanced disease with a significantly higher proportion of T4 (P = 0.009), N2 tumors (P < 0.01) with lymphovascular invasion (P < 0.01). There were fewer stage I tumors (P = 0.04) and more stage IV tumors in the eCC group.

Table 1 Patient and tumor characteristics (all patients, stages I, II, III, and IV

View Fullscreen

Association between eCC and disease-free survival

On univariate analysis, eCC (P < 0.01), T4 tumors (P < 0.01), lymphovascular invasion (P < 0.01), N1 (P= 0.04), and N2 nodal status (P < 0.01) were associated with adverse DFS (Table 2). On multivariate analysis, eCC persisted as an independent predictor of adverse DFS (HR 1.30, 95% CI 0.88–1.92, P = 0.05; Table 2). The role of eCC in DFS was further assessed in a Kaplan-Meier estimate and log-rank test (Fig. 1A). There was a significant difference in distribution of curves representing eCC and CC (P = 0.01). Modes of presentation within eCC (obstruction, perforation and abscess) were also plotted and compared in a Kaplan-Meier estimate. Curves representing each mode differed significantly from the CC group (P = 0.04; Fig. 1B).

Table 2 Univariate (log rank) and multivariate analysis (forward conditional Cox proportional hazards model) of variables associated with disease-free survival (DFS; stage IV patients excluded).

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

Association between eCC and disease-free-survival in stage II

The incidence of stage II disease was similar between CC (42.4%) and eCC (35.1%), P = 0.28 (Table 1). In this cohort (stage II only), eCC was associated with adverse DFS on univariate (P = 0.01) and multivariate analysis (HR 1.4, 95% CI 1.31–3.44, P = 0.03). A Kaplan-Meier estimate was generated comparing DFS between eCC and CC. There was a significant difference in distribution of curves representing eCC and CC (P = 0.01; Fig. 2A). Modes of presentation within eCC (obstruction, perforation, and abscess) were compared with respect to DFS in stage II alone. There was a significant association between the mode of emergency presentation and DFS (P = 0.04; Fig. 2B).

View Figure

• Download as Powerpoint
• Download Figure

Discussion

This study characterizes the effect of emergency presentation on long-term oncologic outcomes in patients undergoing resection with curative intent for colon cancer. Discrepancy exists within the literature regarding the role of eCC in DFS. A number of issues have hampered the development of a consensus regarding this topic including (1) the combination of colon and rectal cancers, (2) failure to incorporate significant findings into a multivariate analysis, (3) inclusion of stage IV cases/postoperative mortality and (4) the relative rarity of this presentation.¹⁹⁻²² With the exclusion of rectal cancer and in-hospital mortality, this study specifically evaluates the impact of emergency surgery on the long-term oncologic outcomes of colon cancer.

The current study showed that eCC was an independent predictor of locoregional recurrence and adverse DFS, but not systemic recurrence and OS. These findings suggest that patients who developed early locoregional recurrence may be amenable to repeated curative surgical resection and this cohort of patients may subsequently survive beyond 5 years, therefore not influencing OS. Similar findings have been reported by others.²³ Furthermore, the negative impact of eCC on locoregional recurrence and DFS remained in patients with stage II disease in the subgroup analysis. These findings have important implications on the use of adjuvant chemotherapy and surveillance strategy.

Up to 20% of stage II (node negative) CRC patients develop disease recurrence on follow-up, reflecting limitations of the current staging system.²⁴ However, there is still a lack of strong evidence for adjuvant chemotherapy in this cohort of patients despite multiple large international trials. Several poor prognostic factors including lymphovascular invasion, T4 tumors, poor differentiation, and emergency presentation are currently used as adjuncts to the TNM staging to guide adjuvant chemotherapy.²⁵⁻²⁷ The findings of this study further validated the prognostic value of emergency presentation and its indication for adjuvant therapy in patients with stage II disease.²⁸

Moreover, this study showed a significant association between eCC and locoregional recurrence. This implies a more stringent surveillance strategy including computed tomography and endoscopy to detect early recurrence. Early locoregional recurrences may be amenable to further curative resections, thus conferring overall survival benefit as shown in this study.

There are several limitations to this study. It is a retrospective study that incorporates patients over a 10-year period. Varying surgeons, pathologists, and radiologists were involved during this time. Surgical techniques and treatment modalities have evolved during the study period. Despite these limitations, this study reaffirms that eCC is associated with adverse pathologic features, and is an independent predictor of locoregional recurrence and DFS. Patients with eCC may be considered for adjuvant chemotherapy following curative surgery and should undergo more stringent surveillance strategy.