PURPOSE: To compare the outcomes of primary surgery for colorectal cancer in an older patient population consisting of the young old (65-74), older old (75-84), and oldest old (greater than or equal to 85) with those in younger patient groups. BACKGROUND: Colorectal cancer continues to be a significant cause of cancer-related deaths in the United States, particularly in older people. Age remains the most significant risk factor for colorectal cancer.(1) Studies have shown that over 60% of patients requiring surgical intervention for this disease are age 70 and older.(2) Furthermore, for every 7 years over the age of 50, the risk of developing colorectal cancer nearly doubles.' The prevalence of colorectal cancer in older people is likely to increase further as the size of the geriatric population increases. Surgical therapy is the cornerstone for treatment of colorectal cancer. Attempts at more conservative and less definitive treatment are associated with at least a twofold increase in death rate.(4) Given the propensity of older people to develop colorectal cancer and the current status of surgery as the standard for treatment, the clinician is faced with the dilemma of how aggressive to be in treating this population of patients while being respectful of their coexisting comorbidities, life expectancy, and quality of life issues. Several studies have shown that age as an isolated factor has minimal or no effect on mortality after colorectal surgery for cancer.(5-9) This systematic review analyzed postoperative mortality, morbidity, length of hospital stay, overall survival, and cancer-specific survival as measures of outcome in older patients with colorectal cancer compared with younger (< 65 years) patients. DATA SOURCES: The following electronic databases were searched from when they were started to July 1998: Medline, Embase, CancerLit, Cochrane Controlled Trials Register, Cinahl, Healthstar, Science Citation Index, Edina Biosis, National Health Service Economic Evaluation Database, Index to scientific and technical proceedings, and Pascal. Manual searches were performed of conference abstracts from annual meetings of the Association of Surgeons of Great Britain and Ireland, European Congress of Surgery, American Society of Colon and Rectal Surgeons, 1996-1998, and the First European Conference on the Economics of Cancer, 1997. The National research register, Medical Research Council trials directory, current research in Britain, United Kingdom Coordinating Committee on Cancer Research trials register, center watch clinical trials listing, physician data query, Nation Institutes of Health inventory of clinical trials and studies, trial amnesty on Cochrane library, system for information on grey literature in Europe, index of UK theses, Department of Health and Social Services (DHSS) data CD, and the International Network of Agencies for Health Technology Assessment (INAHTA) database were also searched for relevant citations and continuing or recently completed studies. Cancer registries in the United Kingdom in which prospective audits were believed to have taken place were also contacted. STUDY SELECTION CRITERIA: Prospective, longitudinal studies of adults undergoing primary treatment for Duke's stage A-D colon or rectal cancer were eligible. Population-based studies (including all patients with colorectal cancer) and consecutive and nonconsecutive surgical series were included. Only those studies published after January 1, 1988, and with more than 100 participants were included; studies in which patients without symptoms were identified by screening and that focused on prognostic markers or blood transfusion in relation to outcome of surgery and randomized controlled trials of follow-up methods were excluded. DATA EXTRACTION: Data were collected from 28 independent studies, which included 34,194 patients. These studies allowed for the data to be broken down by age. Three of the studies documented the progress of all patients with colorectal cancer within a geographical area, five documented data on all patients presenting to the hospital irrespective of whether they underwent surgery, and the remaining studies were prospective series of surgical cases (14 of which included consecutive patients undergoing surgery for colorectal cancer, six of which included only patients undergoing curative resection). Of the 28 studies, 22 included both colon cancer and rectal cancer patients, five included only rectal cancer patients, and one included only colon cancer patients. Data were extracted by one reviewer from published papers and verified by a second reviewer. The following data were recorded: demographic characteristics of patients, site and stage of the tumors, preoperative comorbidities, surgical interventions, and outcomes (including postoperative morbidity, mortality, recurrence, survival, quality of life, and cost effectiveness). Rate ratios were calculated for each of the older patient subgroups with respect to those patients less than age 65 for mortality and survival data. Postoperative complications and prognostic factors, such as stage of disease and type of surgery (elective vs emergent) in different age groups, were compared by calculating an overall rate per group by summing the number of events and the denominators from each individual study. The rates of postoperative complications were then assessed for trends in incidence using the chi-squared test, as were the distribution of prognostic factors. MAIN RESULTS: The study represented 34,194 patients, of which 34% were < 65, 32% were 65 to 74, 27% were 75 to 84, and 8% were greater than or equal to 85. The postoperative mortality rate ratios were 1.8, 3.2, and 6.2 in the 65 to 74, 75 to 84, and greater than or equal to 85 age groups, respectively, when compared with the < 65 age group. The frequencies of various postoperative complications are shown in Table 1. There was a significant trend toward increased rates of pneumonia/respiratory failure, cardiovascular complications, cerebrovascular accidents, and thromboembolism in older people, whereas rates of anastomotic leak were not significantly different. Table 2 shows the median 2-year and 5-year overall survival rates for each of the age groups. In those undergoing surgery with curative intent, the median survival rate ratios at 2 years were 0.92, 0.82, and 0.65 for the 65 to 74, 75 to 84, and greater than or equal to 85 age groups, respectively, compared with those <65. However, this trend was not as clear as the trend in overall survival secondary to the greater variability between studies. Data from consecutive patients presenting with colorectal cancer regardless of whether they underwent surgery suggested a decline in cancer-specific survival with age. However, cancer-specific survival in those undergoing curative surgery yielded rate ratios close to I at both 2 years and 5 years in all three older age groups. Median values for cancer-specific survival at 2-year and 5-year follow-up for those undergoing curative resection are shown in Table 3. A linear relation with respect to age and stage of disease at presentation (P = .0014) was evident in those who were staged, with older patients presenting with more advanced disease. In addition, the frequency of unstaged cancers increased with age, with 3.9% in those <greater than>65, 6.1% in those 65 to 74, 9.0% in those 75 to 84, and 17.3% in those greater than or equal to 85. Five studies reported data on prevalence of comorbidities. Several of the comorbidities were more prevalent in the older age groups, but these comorbidities were assessed differently in the various studies, and therefore pooled estimates were not obtainable. There was a statistically significant (P < .0001) trend toward increasing rates of emergent versus elective surgery with advancing age (Table 4). There was also a significant trend toward decreasing rates of curative operations with advancing age (P < .0001). The percentage of patients undergoing curative surgery were as follows: 76% in the <65 age group (n = 10,772), 75% in the 65 to 74 age group (n = 9,710), 73% in the 75 to 84 age group (n = 7,805), and 67% in the 85 age group ( n = 1,932). Data from studies reporting on the number of patients not undergoing surgery for colorectal cancer demonstrate that older people are less likely to undergo surgery. The rates of no operation were 4%, 6%, 11%, and 21% in the <less than>65, 65 to 74, 75 to 84, and = 85 age groups, respectively. CONCLUSION: This systematic review concludes that the relationship between outcomes with colorectal cancer surgery and age of patients is complex and confounded by variables including stage at presentation, tumor site, preexisting comorbidities, and type of treatment administered. Furthermore, selected older patients can achieve cancer-specific survival rivaling that in young patients, as demonstrated by a relative cancer-specific survival of nearly 1 for all older groups undergoing curative surgery. As such, surgery should not be withheld from the older patient based on age alone.
