Rectal cancer surgery in patients more than 80 years of age

Patients and methods 

From 1980 to 1998, 368 patients were operated on for adenocarcinoma of the rectum in our institution. For the purpose of the study, they were divided into two groups: group 1 included 92 patients (25%) who were 80 years of age or older on first presentation (age range 80 to 91 years; mean age 83.3 years); group 2 included 276 patients (75%) who were younger than 80 years (age range 15 to 79 years; mean age 62.6 years).

The collected data included the sex and operative risk factors (according to the classification of the American Society of Anesthesiologists [ASA] [11]), the location of the tumours and Dukes’ classification [12], the incidence of emergency or elective operations, the incidence of radical or palliative resections, the operative procedures, the associated treatments, the short-term and long-term outcomes. Operative mortality was defined as death inside or outside the hospital during the first 30 days after surgery. All patients were followed up for 5 years or to the end of the study period (February 2000). Causes of deaths were recorded.

The chi-square test or Fisher’s exact test were used for the comparisons of categorical variables between the two groups. Multiple logistic regression analysis was used to investigate the independent effects of variables on operative mortality. Crude and cancer-specific survival curves were constructed for the two groups of patients, using the Kaplan-Meier method, and statistical comparisons were made using the log rank test. Statistical significance was defined as P < 0.05. Statistical analysis was performed on SPSS Software 10.0 (SPSS Inc, Chicago, Illinois).

Results 

Characteristics of patients and tumors are given in Table 1. There were significant differences with regard to the sex-ratio and the operative risk of patients: among patients over 80 years of age (group 1), 41% were male compared with 59% in patients under 80 years (group 2; chi-square = 8.41, 1 d.f., P = 0.004); 33% of patients were classified as ASA III/IV in group 1 compared with 16% in group 2 (chi-square = 11.32, 1 d.f., P < 0.001). There was no significant difference between the two groups with regard to the location and Dukes’ classification of the tumours.

Table 1. Sex and operative risk factors of patients, location and Dukes’ classification of tumors

		Group 1 Age >80 years (n = 92)	Group 2 Age <80 years (n = 276)	P
	Sex
	Male	38 (41)	162 (59)	0.004*
	Female	54 (59)	114 (41)
	ASA
	I and II	62 (67)	231 (84)	<0.001*
	III and IV	30 (33)	45 (16)
	Location
	Upper rectum	28 (30)	88 (32)	0.73*
	Mid rectum	50 (55)	138 (50)
	Low rectum	14 (15)	50 (18)
	Stage
	Dukes A	7 (8)	21 (8)	0.87*
	Dukes B	43 (47)	136 (49)
	Dukes C	25 (27)	78 (28)
	Dukes D	17 (18)	41 (15)

Values in parentheses are percentages. ASA = American Society of Anesthesiologists.

* Chi-square test.

Operative and nonoperative treatments are given in Table 2. Twenty percent of patients (n = 18) underwent emergency surgery for obstruction (n = 15) or perforation (n = 3) in group 1, compared with 7% in group 2 (n = 18 patients; 14 for obstruction and 4 for perforation; chi-square = 13.30, 1 d.f., P < 0.001). Seventy-eight percent of patients underwent resection with a curative intent in group 1, compared with 87% in group 2 (chi-square = 4.04, 1 d.f., P = 0.04). There was no significant difference between the two groups with regard to the type of procedures and to chemotherapy. Radiotherapy was more frequently proposed in group 2 (chi-square = 10.23, 1 d.f., P < 0.001).

Table 2. Operative and associated treatments

		Group 1 (n = 92)	Group 2 (n = 276)	P
	Operation
	Emergency	18 (20)	18 (7)	<0.001*
	Obstruction	15	14
	Perforation	3	4
	Elective	74 (80)	258 (93)
	Resection
	Curative	72 (78)	240 (87)	>0.04*
	Palliative/no resection	14/6 (22)	29/7 (13)
	Type of procedure
	Anterior resection	38 (44)	138 (51)	>0.06*
	Abdominoperineal resection	31 (36)	104 (39)
	Endoanal resection	6 (7)	14 (5)
	Hartmann’s operation	11 (13)	13 (5)
	Chemotherapy	10 (11)	47 (17)	>0.16*
	Preoperative	1	6
	Postoperative	9	41
	Radiotherapy	5 (5)	54 (20)	<0.001*
	Preoperative	1	12
	Postoperative	4	42

Values in parentheses are percentages.

* Chi-square test.

The overall operative mortality rate was 8% in group 1 compared with 4% in group 2 (chi-squareYates = 1.25, 1 d.f., P = 0.26). In patients aged 80 years or more, 4% died after elective surgery compared with 22% after emergency surgery; 3% of ASA I and II patients died compared with 17% of ASA III and IV patients; 7% died after curative resection compared with 10% after palliative or no resection. There was no significant difference with the corresponding rates in patients younger than 80 (Table 3). Multiple logistic regression analysis showed that ASA staging and emergency surgery were independently related to operative mortality but age was not. The causes of operative mortality are summarized in Table 4; most patients (n = 12) died of cardiac or respiratory complications rather than from complications directly attributable to the surgery (n = 5).

Table 3. Operative mortality rates

		Group 1 (n = 92)	Group 2 (n = 276)	P
	Total number of deaths	7 (8)	11 (4)	0.26*
	ASA
	I and II	2/62 (3)	5/231 (2)	0.64†
	III and IV	5/30 (17)	6/45 (13)	0.94*
	Surgery
	Elective	3/74 (4)	8/258 (3)	0.71†
	Emergency	4/18 (22)	3/18 (17)	1.00*
	Resection
	Curative	5/72 (7)	8/240 (3)	0.19†
	Palliative/no resection	2/20 (10)	3/36 (8)	1.00†

Values in parentheses are percentages.

* Chi-square test. † Fisher’s exact test.

Table 4. Causes of operative mortality

		Group 1 (n = 7)	Group 2 (n = 11)
	Pulmonary embolism	2	4
	Cardiac failure	2	3
	Respiratory failure	0	1
	Peritoneal sepsis (preoperative condition)	0	1
	Anastomotic leak	2	2
	Colonic necrosis	1	0

Excluding operative mortality, the overall 2-year and 5-year survival rates were, respectively, 57% and 35% in group 1 (patients over 80 years of age), compared with 77% and 53% in group 2 (log-rank test = 12.44, 1 d.f., P = 0.0004; Fig. 1). The cancer-specific 2-year and 5-year survival rates were, respectively, 63% and 45% in group 1, compared with 78% and 54% in group 2 (log-rank test = 4.06, 1 d.f., P = 0.04).

View Large Image Download to PowerPoint

Fig. 1. Overall survival of patients (excluding operative mortality). Solid line represents group 1 (patients aged >80 years); dashed line represents group 2 (patients aged <80 years).

In patients operated on with a curative intent, the overall 2-year and 5-year survival rates were, respectively, 63% and 40% in group 1, compared with 80% and 57% in group 2 (log-rank test = 9.43, 1 d.f., P = 0.002). In patients over 80 years of age, there were 32 deaths within 5 years of surgery; of these, 6 (18%) were unrelated to cancer. In younger patients, there were 77 deaths; of these, 5 (6%) were unrelated to cancer (chi-squareYates = 2.51, 1 d.f., P = 0.07). The cancer-specific 2-year and 5-year survival rates were, respectively, 68% and 50% in group 1, compared with 82% and 59% in group 2 (log-rank test = 3.05, 1 d.f., P = 0.08; Fig. 2); for Dukes’ B patients, these rates were 84% and 58% in group 1, compared with 93% and 70% in group 2 (log-rank test = 2.72, 1 d.f., P = 0.10); for Dukes’ C patients, these rates were 39% and 26% in group 1, compared with 59% and 31% in group 2 (log-rank test = 0.92, 1 d.f., P = 0.34).

View Large Image Download to PowerPoint

Fig. 2. Cancer-specific survival of patients operated on with a curative intent.Solid line represents group 1 (patients aged >80 years); dashed line represents group 2 (patients aged <80 years).

Comments 

Previous studies about colorectal cancer surgery in the elderly were performed in patients older than 70 [5], [10], 75 [4], [13], 80 [6], [8], [9], [14], [15], [16], [17], or 90 years of age [18]. In the present study, limited to rectal cancer surgery, 80 years was selected as a limit of age for two reasons. Firstly, patients over 80 years of age represent more than 20% of patients [6], [14], this rate being verified in our series. Secondly, some studies have suggested that 80 years represents a barrier beyond which the postoperative mortality rate is significantly increased [17], [19], [20].

Most of the available studies suggested that colorectal cancer surgery in the elderly is associated with a significantly higher perioperative mortality rate than in younger patients [5], [14], [19], [21]. In some studies, old age is even an independent operative risk factor [14], [21]. The conclusion that can be drawn from the present series is different. The observed 8% perioperative mortality is within the reported range of 6% to 20% for elderly patients [4], [5], [9], [10], [13], [15], [22], [23], [24]. It was higher than the rate observed in younger patients (4%), but the difference was not statistically significant. Moreover, two confounding factors increasing the difference between the two groups of patients were identified by analysis of perioperative mortality rates: the ASA distribution of patients and the incidence of elective and emergent surgery. The proportions of patients with ASA high classes (III and IV) and those receiving emergent surgery were significantly higher among patients aged 80 years or older. The relationships between the preoperative general state of health and postoperative mortality [13], [15], [18] as well as the lower mortality rate among patients undergoing an elective procedure [9], [10], [19], [25] have been well demonstrated and were also encountered in our series in both groups of patients. After stratification on the ASA distribution, the mortality rates were quite similar in both groups of patients. For ASA classes I and II, they were 3% in the older group 2% in the younger group. For ASA classes III and IV, they were 17% and 13%, respectively. After stratification on elective or emergent surgery, the mortality rates were also quite similar. In patients who received elective surgery, they were 4% in the older group and 3% in the younger group. In patients who underwent emergent surgery, they were 22% and 17%, respectively. These findings were confirmed by multivariate analysis. Therefore, age does not increase the mortality rate by itself but only by its association with higher ASA classes and emergent surgery.

The overall 5-year survival rate in patients over 80 years of age (37%) was lower than that observed in younger patients (52%). But this difference appeared to be mainly explained by the lower proportion of curative operations and the higher proportion of deaths from other causes in the elderly group. For patients operated on with a curative intent, the 5-year cancer-specific survival rates were not significantly different in both groups of patients, whatever the tumour stage. Thus, age does not affect adversely the prognosis of rectal cancer by itself, this conclusion being consistent with the available literature [10], [26].

The present study clearly supports the policy of having the same surgical indications for rectal cancer in elderly patients as in younger patients; the operative mortality rate is not significantly increased and the 5-year cancer-specific survival rate is almost the same for patients operated on with a curative intent. It also suggests that systematic screening for rectal cancer in elderly patients (digital examination, rectoscopy) might improve the overall prognosis by reducing the incidence of emergency operations and increasing the curative resection rate.