Colon Ca 5.pdf

Original article

Colon cancer management and outcome in relation to individual hospitals in a defined population 1 ¨ A. Sjovall , T. Holm1 , T. Singnomklao2 , F. Granath3 , B. Glimelius4 and B. Cedermark1 1 Department of Surgery, Karolinska University Hospital, Solna, Karolinska Institutet, 2 Oncologic Centre, 3 Department of Medicine, Clinical Epidemiology Unit, Karolinska Institutet and 4 Department of Oncology and Pathology, Karolinska University Hospital, Solna, Karolinska Institutet, Stockholm, Sweden ¨ Correspondence to: Dr A. Sjovall, Department of Surgery, P9:03, Karolinska Hospital, S171 76 Stockholm, Sweden (e-mail: [email protected])

Background: The Stockholm and Gotland region in Sweden has a common management protocol for

the treatment of colon cancer. The aim of this study was to assess the management and treatment of colon cancer in the region and to try to identify ways to improve the outcome further. Methods: Clinical data on all patients diagnosed with colon cancer in the region’s nine hospitals between January 1996 and December 2000 were prospectively collected. Patients were followed until December 2004, and their management and outcome analysed. Results: Colon cancer was diagnosed in 2775 patients. An elective operation was performed in 2116 (76·3 per cent) patients and an emergency procedure in 590 (21·3 per cent). Emergency surgery was an independent risk factor for death. The crude overall cumulative 5-year survival was 46·2 per cent. A multivariable analysis of risk of dying and risk of local recurrence showed significant differences between hospitals. The number of lymph nodes examined in the specimens also differed between hospitals. Conclusion: Differences in the management and outcome of colon cancer in the nine hospitals, despite a common management protocol, indicate a need for improving collaboration between hospitals and multidisciplinary management. Paper accepted 20 December 2006 Published online 29 January 2007 in Wiley InterScience (www.bjs.co.uk). DOI: 10.1002/bjs.5455

Introduction

Since 1980, The Stockholm Colorectal Cancer Study Group has been committed to improving the management and outcome of colorectal cancer. It has accomplished two large randomized trials on preoperative radiotherapy in rectal cancer (Stockholm I and II trials), an adjuvant chemotherapy trial in colorectal cancer and an educational project on multidisciplinary management in patients with rectal cancer, including the concept of total mesorectal excision1 – 4 . In addition, the group has established management protocols for colorectal cancer since 1980. Since 1996, as part of these protocols, all patients within the Stockholm population with colorectal cancer have been prospectively registered, and their follow-up data recorded. As a result of these efforts, local control and survival in patients with rectal cancer has improved in the region as well as in other parts of Sweden in the past decade. A similar effect on outcome has become evident in other countries as a result of the widespread implementation Copyright  2007 British Journal of Surgery Society Ltd Published by John Wiley & Sons Ltd

of total mesorectal excision and the introduction of preoperative radiotherapy5 – 7 . Recent population-based data from Sweden show that patients with rectal cancer now have a better prognosis than those with colon cancer8,9 . From a population of 1·9 million, about 550 new patients with colon cancer are registered in the region annually. Although the management protocol established in Stockholm includes patients with colon cancer, no specific efforts have been made to improve outcome in these patients. Thus in 2004 the Stockholm Colorectal Cancer Study Group initiated the Colon Cancer Project in Stockholm in an attempt to improve the outcome for this large group of patients. The initial task was to analyse the prospective data on almost 3000 patients with colon cancer registered between 1996 and 2000 and followed until December 2004. The aim of this study was to assess the overall management and outcome in these patients, to establish whether the management and prognosis was related to British Journal of Surgery 2007; 94: 491–499

492

¨ A. Sjovall, T. Holm, T. Singnomklao, F. Granath, B. Glimelius and B. Cedermark

individual hospitals and to try to identify ways of improving management and outcome. Patients and methods

In Sweden, all patients with cancer are reported to the National Cancer Registry by the physician and the pathologist at diagnosis. Causes of death are reported to the Cause of Death Registry at the National Board of Health and Welfare. The Stockholm County Council keeps record of all healthcare consumption including diagnoses according to the International Classification of Diseases of the World Health Organization. The registration uses an identification number unique to each resident of Sweden. Since 1996, all patients with colon cancer in the Stockholm region have been registered with the Oncologic Centre in Stockholm. The database includes information on age, sex, tumour location and stage, emergency or elective surgery, type of surgery performed, postoperative mortality, radiotherapy, chemotherapy, the histopathology of the tumour and follow-up data on recurrence and survival. The colon is defined as the large bowel above 15 cm from the anal verge, excluding the appendix. Tumours are classified as being in the right, transverse or left colon or as being multiple (more than one synchronous tumour or later diagnosis of a new tumour) or unknown location (when the location of the primary tumour is impossible to establish). The right colon includes the caecum, ascending colon and hepatic flexure, while the left colon includes the splenic flexure, descending colon and the sigmoid. Surgery is defined as ‘curative’ if no distant metastases are present and the primary resection is locally complete according to both the surgeon and the pathologist, as ‘uncertain’ when there are no distant metastases but the local completeness of the resection is uncertain according to either the surgeon or the pathologist, and as ‘incomplete’ if the resection is locally incomplete according to the surgeon or the pathologist, or if they both assess the completeness of the resection as uncertain. A fourth patient group is defined for those who have distant metastases at primary surgery or in whom the primary tumour is not resected. The Oncologic Centre database is continuously validated and updated using the registries above, and all patients with an invasive adenocarcinoma of the colon are included. This study included all patients from the Stockholm population diagnosed with a primary invasive adenocarcinoma of the colon between January 1996 and December 2000. The patients were followed until December 2004. The region had nine hospitals and five pathology departments. Copyright  2007 British Journal of Surgery Society Ltd Published by John Wiley & Sons Ltd

Hospitals 1 and 2 were university hospitals associated with Karolinska Institutet, 3 and 4 were large community hospitals that also had medical students and 5–9 were community hospitals without medical students (see Table 1). All had emergency rooms except Hospital 6. In addition to the information in the Oncologic Centre database, details of healthcare consumption for all patients after the diagnosis were taken from the Stockholm County Council registry. In patients without a reported recurrence at the end of follow-up, an analysis of medical records was performed to ensure that no patient with a diagnosed cancer recurrence had been missed. Data on patients from the Stockholm population who had been treated in hospitals outside the region were also collected through medical records. In the few patients who moved out of the region during the study, data on tumour recurrence were not available. However, in patients who died during follow-up, information on the date and cause of death was retrieved and recorded. In addition to the analysis of management and outcome in the total number of patients within the region, results were analysed separately for the nine different hospitals. In these analyses, only abdominal operations were included. The institutions outside the region where patients had had emergency surgery were for the purpose of this study merged and defined as hospital 10. Patients treated in this group were excluded from the comparative analyses regarding outcome in relation to hospital, but they were included in the descriptive part of this study, to avoid selection of patients. During the study, a randomized trial assessed the value of adjuvant chemotherapy to patients with stage II or III colon cancer, and some of the patients in this study were included in that trial4 .

Statistical analysis The survival time was calculated from the date of the primary surgery until the date of death or the end of follow-up. In non-operated patients, the survival time was calculated from the date of diagnosis until the time of death or end of follow-up. Comparisons of survival times between patient groups were made with the Kaplan–Meier method and log rank test. The χ2 test was used for comparisons of differences in proportions. A proportional hazards regression model was used to assess survival after elective versus emergency abdominal surgery. The risk of death and recurrence in the different hospitals was also calculated in a proportional hazards regression model. In these multivariable analyses, www.bjs.co.uk

British Journal of Surgery 2007; 94: 491–499

Copyright  2007 British Journal of Surgery Society Ltd Published by John Wiley & Sons Ltd

www.bjs.co.uk

44 138 (38·9) 85 (23·9) 83 (23·4) 5

29 108 (34·6) 84 (26·9) 78 (25·0) 13 6 9

156 (43·9) 35 155 (43·7) 8 1 4 270 (76·0) 81 (22·8) 208 (58·6)

151 (48·4) 33 124 (39·7) 4 0 13 246 (78·8) 53 (17·0) 187 (59·9)

4 11

184 (51·8) 171 (48·2) 74 (22–93)

156 (50·0) 156 (50·0) 72 (30–95)

2

20 12

39 153 (33·8) 145 (32·1) 108 (23·9) 7

182 (40·3) 42 214 (47·3) 13 1 9 297 (65·7) 146 (32·3) 288 (63·7)

216 (47·8) 236 (52·2) 75 (30–94)

3

17 6

57 (13·8) 163 (39·4) 112 (27·0) 77 (18·6) 5

184 (44·4) 42 180 (43·5) 7 1 14 300 (72·5) 100 (24·2) 296 (71·5)

206 (49·8) 208 (50·2) 75 (29–97)

4

10 10

37 165 (40·0) 125 (30·3) 79 (19·2) 6

187 (45·4) 59 (14·3) 157 (38·1) 9 0 4 295 (71·6) 113 (27·4) 297 (72·1)

193 (46·8) 219 (53·2) 76 (35–97)

5

0 8

82 (18·3) 171 (38·1) 113 (25·2) 61 (13·6) 22

186 (41·4) 37 215 (47·9) 11 0 6 435 (96·9) 8 362 (80·6)

210 (46·8) 239 (53·2) 74 (27–94)

6

5 3

9 46 (40·4) 27 (23·7) 31 (27·2) 1

42 6 59 (51·8) 7 0 2 83 (72·8) 29 77 (67·5)

55 (48·2) 59 (51·8) 73 (32–93)

7

3 1

17 53 (47·3) 11 (9·8) 25 (22·3) 6

46 10 51 (45·6) 3 2 11 80 (71·4) 21 78 (69·6)

55 (49·1) 57 (50·9) 75 (24–97)

8

1 2

8 32 41 18 6

41 13 49 2 0 1 79 (75·2) 25 76 (72·4)

48 57 74 (32–90)

9

0 1

10 14 10 12 4

20 4 24 1 1 5 31 14 32

22 28 71 (29–91)

10†

Total

66 (11·2) 63 (3·1)

332 (12·0) 1043 (37·6) 753 (27·1) 572 (20·6) 75 (2·7)

1195 (43·1) 281 (10·1) 1228 (44·3) 65 (2·3) 6 69 (2·5) 2116 (76·3) 590 (21·3) 1901 (68·5)

1345 (48·5) 1430 (51·5) 74 (22–97)

Values in parentheses are percentages or *ranges. †Refers to patients who had surgery outside the Stockholm and Gotland region or those who did not visit a department of surgery.

Male Female Median age (years)* Tumour location Right Transverse Left Multiple Unknown No surgery Elective surgery Emergency surgery Curative surgery Tumour stage I II III IV Unknown 30-day mortality Emergency surgery Elective surgery

1

Hospital

Patient characteristics of 2775 patients diagnosed with colon cancer in nine hospitals in the Stockholm and Gotland region, 1996–2000

Characteristics

Table 1

Colon cancer management 493

British Journal of Surgery 2007; 94: 491–499

494

¨ A. Sjovall, T. Holm, T. Singnomklao, F. Granath, B. Glimelius and B. Cedermark

adjustment was made for case mix including age, sex and tumour node metastasis stages. For comparisons between hospitals, adjustment was also made for elective and emergency surgery. Results

A total of 2855 patients were diagnosed with colon cancer during the study period. In 80, the tumour was diagnosed at autopsy, and these patients were excluded from further analysis. Clinical characteristics of the remaining 2775 in relation to treatment hospital are shown in Table 1. There were 1345 (48·5 per cent) men and 1430 (51·5 per cent) women, and the median age at diagnosis was 73 years in men and 76 years in women. There were no significant differences between hospitals regarding sex, but hospital 1 patients had a lower median age and hospital 5 patients a higher median age. Patient and tumour characteristics did not change over time (data not shown).

Diagnosis and tumour location The diagnosis was verified by histopathology in 2745 patients (98·9 per cent) and by unequivocal radiology or endoscopy findings alone with no subsequent colon resection in 30 (1·1 per cent). Right-sided tumours were significantly more common in women: 669 of 1430 (46·8 per cent) versus 526 of 1345 (39·1 per cent) in men (P < 0·001), while left-sided tumours were more common in men; 651 of 1345 (48·4 per cent) versus 577 of 1430 (40·3 per cent) in women (P < 0·001). The proportion of tumours in the transverse colon was similar in men and in women, 9·9 and 10·6 per cent respectively. Synchronous distant metastases were found in 572 patients (20·6 per cent). Liver metastases were present in 398 patients and peritoneal carcinomatosis in 130. Other less common locations of distant spread were lung, intra-abdominal lymph nodes and ovaries. Synchronous metastases in multiple locations were diagnosed in 89 patients.

Table 2 Surgical procedures in 2706 patients with colon cancer in the Stockholm and Gotland region, 1996–2000 Procedure

n

Ileocaecal resection Right hemicolectomy Resection of transverse colon Left hemicolectomy Resection of sigmoid colon Anterior resection* Hartmann’s procedure Subtotal colectomy† Endoscopic polypectomy Exploratory laparotomy

58 (2·1) 1177 (43·5) 115 (4·2) 242 (8·9) 694 (25·6) 114 (4·2) 48 84 (3·1) 57 (2·1) 117 (4·3)

Values in parentheses are percentages. *Patients with sigmoid cancer; †patients with multiple tumours, previous colorectal cancer or inflammatory bowel disease.

53 of 57 patients who had an endoscopic polypectomy, the tumour was located in the sigmoid colon. Laparotomy without bowel resection was performed in 117 patients. The operation was considered potentially curative in 1901 patients (70·3 per cent). Of those, 1549 (81·5 per cent) had an elective and 352 (18·5 per cent) an emergency operation. In 184 patients, the local completeness was uncertain. Of those, 132 (71·7 per cent) had an elective and 52 (28·3 per cent) an emergency procedure. In 621 patients, the operation was considered non-curative owing to locally incomplete resection or synchronous distant metastases. Of these, 435 (70·0 per cent) had an elective operation. Reoperation within 30 days of surgery was performed in 88 (4·2 per cent) patients after elective and 32 (5·4 per cent) patients after emergency surgery. The two most common causes for reoperations were anastomotic leakage and abdominal wall dehiscence. Mortality within 30 days of surgery was significantly higher after emergency surgery than elective surgery: 66 of 590 (11·2 per cent) versus 63 of 2059 (3·1 per cent) (P < 0·001).

Histopathology Surgery Surgery was performed in 2706 patients (97·5 per cent) (Table 2). An abdominal procedure was performed in 2649, which was elective in 2059 (77·7 per cent) and an emergency in 590 (22·2 per cent). Of patients who had a bowel resection, right hemicolectomy and sigmoid resection were the most common procedures. In 188 patients, another organ was resected at the same time as the bowel, most commonly the spleen or an ovary. In Copyright  2007 British Journal of Surgery Society Ltd Published by John Wiley & Sons Ltd

The tumours were classified as stage I in 332 (12·0 per cent) patients, stage II in 1043 (37·6 per cent) patients, stage III in 753 (27·1 per cent) patients and stage IV in 572 (20·6 per cent) patients (Table 1). In 75 patients, the tumour stage was unknown because no surgery had been performed or no lymph nodes had been found or examined. Hospital 6 had a significantly higher proportion of stage I tumours and a lower proportion of stage IV tumours than the other hospitals, whereas hospital 8 had fewer stage III www.bjs.co.uk

British Journal of Surgery 2007; 94: 491–499

Colon cancer management

495

tumours and hospital 9 significantly more stage III tumours (Table 1). The median number of lymph nodes examined was six (range 0–36), and 12 or more lymph nodes were examined in only 385 (15·4 per cent) of the specimens (Table 3). The numbers of specimens having at least 12 lymph nodes examined improved over time: in 2000, 12 or more lymph nodes were examined in 22 per cent of patients compared with 8 per cent in 1996. The median number of lymph nodes examined ranged from five to seven in the nine hospitals. The number of patients having 12 or more nodes examined ranged from 19 of 365 (5·2 per cent) in hospital 4 to 95 of 418 (22·7 per cent) in hospital 6.

Curative Uncertain Incomplete Palliative

100 90

Cumulative survival (%)

80 70 60 50 40 30 20

Neoadjuvant and adjuvant treatment Preoperative chemotherapy or radiotherapy or both was given to 25 patients, all because of advanced disease. In the 1901 patients operated on with a curative intent, 246 had postoperative adjuvant chemotherapy, 198 of whom had stage III tumours. During the study period there was a significant increase in the use of adjuvant treatment to patients with stage III disease and a decrease to those with stage II disease. The proportion of patients with stage III tumours, younger than 76 years and having adjuvant treatment after curative surgery ranged from 40 per cent in hospital 5 to 69 per cent in hospital 6 (data not shown).

10 0

0

1

2

3

4

5

Time after primary surgery or diagnosis (years) No. at risk Curative Uncertain Incomplete Palliative

1901 184 99 591

1732 133 59 187

1550 104 39 80

1409 87 32 38

1146 72 24 23

798 53 17 8

Recurrence and survival

Crude survival in 2775 patients diagnosed with colon cancer in Stockholm and Gotland, 1996–2000, analysed by definition of surgery (curative procedure, uncertain, incomplete resection or palliative procedure in patients with distant metastases or non-resected tumour) (P < 0·001, log rank test)

The median survival time for all 2775 patients, including patients who died within 30 days of surgery, was 50 (range 0–110) months. The crude cumulative 5-year survival was 46·2 per cent. At the end of follow-up, 1145 (41·3 per cent) patients were alive after a median follow-up time of 75 (range 49–110) months. The overall survival times for all 2775 patients in relation to the definition of the surgery are shown in Fig. 1.

Distant metastases, locoregional recurrence or both were found after a median follow-up of 18 months in 386 of 1901 (20·3 per cent) patients operated on for cure. The crude cumulative 5-year survival after curative surgery was 61·1 per cent. In patients where the local completeness of the resection was uncertain, 87 of 184 (47·3 per cent) patients were

Table 3

Fig. 1

Number of lymph nodes examined after abdominal colon resection in five pathology departments and nine hospitals (n = 2494)

Pathology Department

Hospital

1–6 nodes

7–11 nodes

1 2 3 7 5 6 9 4 8

122 (43·9) 172 (51·3) 185 (44·7) 58 (55·8) 221 (56·2) 184 (44·0) 38 254 (69·6) 60 (67)

80 (28·8) 75 (22·4) 127 (30·7) 23 107 (27·2) 125 (29·9) 36 86 (23·6) 22

A B

C D E

> 11 nodes 46 59 (17·6) 86 (20·8) 16 38 95 (22·7) 19 19 7

Data missing

Total

30 29 16 7 27 14 4 6 1

278 335 414 104 393 418 97 365 90

Values in parentheses are percentages.

Copyright  2007 British Journal of Surgery Society Ltd Published by John Wiley & Sons Ltd

www.bjs.co.uk

British Journal of Surgery 2007; 94: 491–499

¨ A. Sjovall, T. Holm, T. Singnomklao, F. Granath, B. Glimelius and B. Cedermark

diagnosed with a tumour recurrence after a median time of 10 months. The cumulative 5-year survival was 39·3 per cent. In 99 patients with an incomplete primary tumour resection, 50 (51 per cent) were diagnosed with progressive disease after a median of 10 months and the cumulative 5-year survival was 24·2 per cent. In 591 patients, including those with synchronous distant metastases and those where the tumour was not resected owing to extensive tumour growth, the cumulative 5-year survival was 2·9 per cent. Excluding patients who died within 30 days of surgery and after adjusting for age, sex and tumour stage, the relative risk of overall death was significantly higher after emergency than after elective surgery (hazard ratio (HR) 1·68, 95 per cent confidence interval (c.i.) 1·45 to 1·96, P < 0·001). After adjusting for local completeness of the surgery, the risk of death was still significantly higher after emergency surgery (HR 1·62, 95 per cent c.i. 1·37 to 1·90, P < 0·001). In order to exclude all deaths related to postoperative complications, an analysis was performed for patients who survived for more than 180 days. This still showed a higher relative risk of dying after emergency than elective surgery, with an HR of 1·79 (95 per cent c.i. 1·50 to 2·13, P < 0·001).

100 90 80 70 Cumulative survival (%)

496

Copyright  2007 British Journal of Surgery Society Ltd Published by John Wiley & Sons Ltd

50 40 30

10

0

Hospital 5 Hospital 6 Hospital 7 Hospital 8 Hospital 9

Hospital 1 Hospital 2 Hospital 3 Hospital 4

20

1

2

3

4

Time after surgery (years) No. at risk 292 347 439 397 404 424 111 95 99

Hospital 1 Hospital 2 Hospital 3 Hospital 4 Hospital 5 Hospital 6 Hospital 7 Hospital 8 Hospital 9

Outcome differences between hospitals The crude survival after an abdominal operation in relation to treating hospital is shown in Fig. 2. The proportion of patients having emergency surgery differed between the hospitals, but the differences in survival persisted also when only patients who had an elective procedure were included. The pattern of hospital performance was consistent throughout the study period. The population in the catchment area of hospital 1 was considered representative of the region and this hospital was chosen as the reference (HR 1·0). In the multivariable analysis, the relative risk for death was significantly lower in hospital 6 and higher in hospital 2, as shown in Table 4. These results were also consistent throughout the study period. In patients who had an abdominal resection with curative intent, the local recurrence rate was 6·2 per cent (113 of 1826). The local recurrence rate was also analysed in relation to individual hospitals (Table 5), and patients who had surgery in hospital 4 had a relatively high risk of local recurrence (HR 2·78, 95 per cent c.i. 1·20 to 6·43, P = 0·017). There was no significant difference in the risk of having distant metastases between the nine different hospitals (data not shown).

60

223 258 314 314 303 373 80 76 73

164 171 224 229 218 310 55 61 56

185 204 254 265 255 339 61 68 59

134 141 186 180 168 259 38 51 43

Crude survival in 2608 patients after abdominal operation for colon cancer in nine hospitals in the Stockholm and Gotland region

Fig. 2

Table 4 Overall death after abdominal surgery for colon cancer in the nine hospitals in the Stockholm and Gotland region, 1996–2000, adjusted for age, sex, tumour node metastasis (TNM) stage and type of surgery (emergency or elective) (n = 2608)

Hospital

No. of events

No. of patients

Hazard ratio

95% confidence interval

170 220 267 234 264 185 73 49 60

292 347 439 397 404 424 111 95 99

1·00 1·24 0·98 1·06 0·96 0·79 1·20 0·93 1·11

1·00–1·54 0·79–1·20 0·86–1·32 0·78–1·19 0·63–0·99 0·90–1·61 0·66–1·31 0·82–1·51

1 2 3 4 5 6 7 8 9

www.bjs.co.uk

British Journal of Surgery 2007; 94: 491–499

Colon cancer management

497

Locoregional recurrence after abdominal potentially curative surgery for colon cancer in the nine hospitals in the Stockholm and Gotland region, 1996–2000, adjusted for age, sex, tumour node metastasis (TNM) stage and type of surgery (emergency or elective) (n = 1826)

Table 5

Hospital 1 2 3 4 5 6 7 8 9

No. of events

No. of patients

Hazard ratio

95% confidence interval

8 15 15 27 25 10 7 2 4

181 206 285 293 294 346 76 73 72

1.00 2.32 1.29 2.78 2.23 0.89 2.81 0.96 1.30

0·94–5·74 0·52–3·20 1·20–6·43 0·95–5·24 0·34–2·38 0·98–8·09 0·20–4·68 0·38–4·50

Discussion

The management of rectal cancer within defined populations has been studied extensively in recent years, mainly thanks to randomized radiotherapy trials, educational projects and quality audits, and the treatment results have improved accordingly1 – 3,5 – 7,10,11 . However, few studies have addressed colon cancer treatment specifically, and recent large population-based audits are scarce12 . Owing to the prospective registration of all patients with colorectal cancer within the Stockholm population since 1996, the outcome has now been assessed in almost 3000 patients with colon cancer followed for at least 4 years. The outcome in this large group of patients has improved in Sweden during the past four decades, but it is notable that survival after treatment for rectal cancer is now better than for colon cancer8 . The proportion of non-operated patients, emergency or elective procedures and potentially curative surgery has not changed markedly from previous population-based Swedish studies13,14 . The worse prognosis for patients who have emergency rather than elective surgery has been reported15 – 18 . It is not clear whether this is caused by patient-related factors or by poor surgery in the emergency setting. Biondo et al.19 show no differences in overall survival between elective and emergency patients after curative resection of stage II colon cancer. However, in the Stockholm and Gotland population, the proportion of patients in whom a locally complete resection could be accomplished was lower in emergency than in elective procedures. Hospital 6, which has essentially no emergency surgery, showed the best results in elective surgery compared with the other hospitals. This might be because hospital 6 is small, with dedicated colorectal surgeons and a large colorectal cancer Copyright  2007 British Journal of Surgery Society Ltd Published by John Wiley & Sons Ltd

volume per surgeon. The lack of an emergency room in this hospital could also have resulted in the selection of fewer patients with concomitant diseases that might have affected the outcome. In the current management protocol, the recommended number of examined lymph nodes for correct staging is at least 12 nodes20,21 . The number of lymph nodes examined after abdominal colon resections in this study is low and signals a suboptimal histopathological examination of the specimen and possibly less than optimal surgery. Previous authors have reported similar problems with harvesting of lymph nodes22 – 24 . The management protocol, introduced in 1996, aimed at achieving similar results in all hospitals in the region. Most of the differences in this study were not significant, but were consistent throughout the study period, indicating that they were not random. However, there are potential confounders to consider. The database lacks information on concomitant diseases; differences in survival could also be affected by socioeconomic differences. Regarding the risk for local recurrence, the low number of events makes it hard to draw firm conclusions from the interhospital differences. It is questionable whether a region of this size should have colon cancer surgery performed in nine different hospitals. The six larger institutions had a volume of 60–90 patients with colon cancer per year, while three hospitals had volumes of less than 25 per year. This study did not show a difference in survival related to hospital volume as reported by previous authors25 – 28 , so a high hospital volume is not a guarantee for good results. The standard of the surgical team as well as pre- and postoperative management are very important29 , although the present study cannot determine the most important factors. After surgery for rectal cancer, the proportion of local recurrences is considered a measure of the quality of the surgery. As for colon cancer, the risk of local recurrence in a population-based survey has not, to the authors’ knowledge, been previously assessed. Indeed, there is no established definition of local recurrence after colon cancer surgery, and in the evaluation in this study, data on locoregional recurrence as reported to the database by the surgeon or oncologist was used. The different risks of locoregional recurrences among the hospitals may reflect a true difference in surgical quality. Data on the competence and experience of the surgeon in charge of each operation were not available for this study, and therefore a more detailed study of the factors responsible for differences in outcome was not possible. During the study period, the management protocol for colon cancer did not include a compulsory preoperative www.bjs.co.uk

British Journal of Surgery 2007; 94: 491–499

498

¨ A. Sjovall, T. Holm, T. Singnomklao, F. Granath, B. Glimelius and B. Cedermark

assessment of the local extent of the primary tumour. This could have an impact on the risk of unexpected findings during colon resection, and might increase the risk of the resection being incomplete during primary surgery. It is generally accepted that a complete resection is a crucial factor in predicting the outcome for patients with colorectal cancer, and the results of this study support this. As can be seen in Fig. 1, the survival of patients where both the surgeon and the pathologist agreed on a locally complete resection was significantly better than if either expressed uncertainty as to the resection margins or, even worse, if both expressed uncertainty or one reported an incomplete resection. This supports a previous report on prognosis after rectal cancer surgery30 . Colon cancer surgery is considered less demanding than rectal cancer surgery, appropriate for general surgeons without a specific colorectal interest and often for residents. There is no established policy in the Stockholm and Gotland region on supervision of trainees in surgery, and a right hemicolectomy has generally been considered suitable for residents with or without the help of a colorectal surgeon. As survival for patients with rectal cancer in Sweden is now better than for those with colon cancer, after intense efforts in clinical research on adjuvant radiotherapy and educational initiatives on surgery, it is time for action to improve management of patients with colon cancer. It seems crucial to increase the multidisciplinary collaboration between surgeons and radiologists for better preoperative staging. In addition, educational initiatives to improve surgery are needed. For example, a better survival than in the current study is reported after extensive and meticulous dissection of the mesocolic lymph nodes31 . A complete removal of the intact mesocolon to the affected colonic segment may be important in the case of colon cancer, as is now accepted in the total mesorectal excision technique in rectal cancer surgery. Also, the histopathological examination of the specimen should be optimized for quality control of the surgery performed and for correct staging. Adjuvant treatment should always be considered in multidisciplinary conferences, and given according to continuously updated clinical guidelines when indicated. If these measures can be adopted, more patients with colon cancer should be cured. Acknowledgements

The Colon Cancer Project in Stockholm is supported by Tyco Healthcare. The present study was supported by the Stockholm Cancer Society. The authors also thank the Stockholm Colorectal Cancer Study Group who made the registers available for this study. Copyright  2007 British Journal of Surgery Society Ltd Published by John Wiley & Sons Ltd

References 1 Preoperative short-term radiation therapy in operable rectal carcinoma. A prospective randomized trial. Stockholm Rectal Cancer Study Group. Cancer 1990; 66: 49–55. 2 Martling AL, Holm T, Rutqvist LE, Moran BJ, Heald RJ, Cedermark B. Effect of a surgical training programme on outcome of rectal cancer in the County of Stockholm. Stockholm Colorectal Cancer Study Group, Basingstoke Bowel Cancer Research Project. Lancet 2000; 356: 93–96. 3 Martling A, Holm T, Johansson H, Rutqvist LE, Cedermark B. The Stockholm II trial on preoperative radiotherapy in rectal carcinoma: long-term follow-up of a population-based study. Stockholm Colorectal Study Group. Cancer 2001; 92: 896–902. 4 Glimelius B, Dahl O, Cedermark B, Jakobsen A, Bentzen SM, Starkhammar H et al. Adjuvant chemotherapy in colorectal cancer: a joint analysis of randomised trials by the Nordic Gastrointestinal Tumour Adjuvant Therapy Group. Acta Oncol 2005; 44: 904–912. 5 Improved survival with preoperative radiotherapy in resectable rectal cancer. Swedish Rectal Cancer Trial. N Engl J Med 1997; 336: 980–987. 6 Kapiteijn E, Putter H, van de Velde CJ. Impact of the introduction and training of total mesorectal excision on recurrence and survival in rectal cancer in The Netherlands. Br J Surg 2002; 89: 1142–1149. 7 Wibe A, Eriksen MT, Syse A, Myrvold HE, Soreide O. Total mesorectal excision for rectal cancer – what can be achieved by a national audit? Colorectal Dis 2003; 5: 471–477. 8 Talback M, Stenbeck M, Rosen M, Barlow L, Glimelius B. Cancer survival in Sweden 1960–1998 – developments across four decades. Acta Oncol 2003; 42: 637–659. 9 Birgisson H, Talback M, Gunnarsson U, Pahlman L, Glimelius B. Improved survival in cancer of the colon and rectum in Sweden. Eur J Surg Oncol 2005; 31: 845–853. 10 Martling A, Holm T, Rutqvist LE, Johansson H, Moran BJ, Heald RJ et al. Impact of a surgical training programme on rectal cancer outcomes in Stockholm. Br J Surg 2005; 92: 225–229. 11 Peeters KC, van de Velde CJ. Surgical quality assurance in rectal cancer treatment: the key to improved outcome. Eur J Surg Oncol 2005; 31: 630–635. 12 McLeish JA, Thursfield VJ, Giles GG. Survival from colorectal cancer in Victoria: 10-year follow up of the 1987 management survey. ANZ J Surg 2002; 72: 352–356. 13 Berge T, Ekelund G, Mellner C, Pihl B, Wenckert A. Carcinoma of the colon and rectum in a defined population. An epidemiological, clinical and postmortem investigation of colorectal carcinoma and coexisting benign polyps in Malmo, Sweden. Acta Chir Scand Suppl 1973; 438: 1–86. 14 Arbman G, Nilsson E, Storgren-Fordell V, Sjodahl R. Outcome of surgery for colorectal cancer in a defined population in Sweden from 1984 to 1986. Dis Colon Rectum 1995; 38: 645–650.

www.bjs.co.uk

British Journal of Surgery 2007; 94: 491–499

Colon cancer management

15 Jestin P, Nilsson J, Heurgren M, Pahlman L, Glimelius B, Gunnarsson U. Emergency surgery for colonic cancer in a defined population. Br J Surg 2005; 92: 94–100. 16 Carraro PG, Segala M, Cesana BM, Tiberio G. Obstructing colonic cancer: failure and survival patterns over a ten-year follow-up after one-stage curative surgery. Dis Colon Rectum 2001; 44: 243–250. 17 McArdle CS, Hole DJ. Emergency presentation of colorectal cancer is associated with poor 5-year survival. Br J Surg 2004; 91: 605–609. 18 Fazio VW, Tekkis PP, Remzi F, Lavery IC. Assessment of operative risk in colorectal cancer surgery: the Cleveland Clinic Foundation colorectal cancer model. Dis Colon Rectum 2004; 47: 2015–2024. 19 Biondo S, Marti-Rague J, Kreisler E, Pares D, Martin A, Navarro M et al. A prospective study of outcomes of emergency and elective surgeries for complicated colonic cancer. Am J Surg 2005; 189: 377–383. 20 Compton CC, Fielding LP, Burgart LJ, Conley B, Cooper HS, Hamilton SR et al. Prognostic factors in colorectal cancer. College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med 2000; 124: 979–994. 21 Nelson H, Petrelli N, Carlin A, Couture J, Fleshman J, Guillem J et al. Guidelines 2000 for colon and rectal cancer surgery. J Natl Cancer Inst 2001; 93: 583–596. 22 Johnson PM, Malatjalian D, Porter GA. Adequacy of nodal harvest in colorectal cancer: a consecutive cohort study. J Gastrointest Surg 2002; 6: 883–888; discussion 889–890. 23 Maurel J, Launoy G, Grosclaude P, Gignoux M, Arveux P, Mathieu-Daude H et al. Lymph node harvest reporting in

Copyright  2007 British Journal of Surgery Society Ltd Published by John Wiley & Sons Ltd

499

24

25

26

27

28

29

30

31

patients with carcinoma of the large bowel: a French population-based study. Cancer 1998; 82: 1482–1486. Jestin P, Pahlman L, Glimelius B, Gunnarsson U. Cancer staging and survival in colon cancer is dependent on the quality of the pathologists’ specimen examination. Eur J Cancer 2005; 41: 2071–2078. Dimick JB, Cowan JA Jr, Upchurch GR Jr, Colletti LM. Hospital volume and surgical outcomes for elderly patients with colorectal cancer in the United States. J Surg Res 2003; 114: 50–56. Marusch F, Koch A, Schmidt U, Pross M, Gastinger I, Lippert H. Hospital caseload and the results achieved in patients with rectal cancer. Br J Surg 2001; 88: 1397–1402. Schrag D, Cramer LD, Bach PB, Cohen AM, Warren JL, Begg CB. Influence of hospital procedure volume on outcomes following surgery for colon cancer. JAMA 2000; 284: 3028–3035. Harmon JW, Tang DG, Gordon TA, Bowman HM, Choti MA, Kaufman HS et al. Hospital volume can serve as a surrogate for surgeon volume for achieving excellent outcomes in colorectal resection. Ann Surg 1999; 230: 404–411; discussion 411–413. Hermanek P, Hohenberger W. The importance of volume in colorectal cancer surgery. Eur J Surg Oncol 1996; 22: 213–215. Martling A, Singnomklao T, Holm T, Rutqvist LE, Cedermark B. Prognostic significance of both surgical and pathological assessment of curative resection for rectal cancer. Br J Surg 2004; 91: 1040–1045. Hohenberger W, Reingruber B, Merkel S. Surgery for colon cancer. Scand J Surg 2003; 92: 45–52.

www.bjs.co.uk

British Journal of Surgery 2007; 94: 491–499