- Email: [email protected]
Delay of treatment is associated with advanced stage of rectal cancer but not of colon cancer
Cancer Detection and Prevention 30 (2006) 341–346 www.elsevier.com/locate/cdp
Delay of treatment is associated with advanced stage of rectal cancer but not of colon cancer Marianne Korsgaard MD, PhDa,*, Lars Pedersen MScb, Henrik Toft Sørensen DMScb, Søren Laurberg DMSca b
a Department of Surgery L, Aarhus University Hospital, Tage Hansensgade 2, 8000 Aarhus C, Denmark Department of Clinical Epidemiology, Aarhus University Hospital, Tage Hansensgade 2, 8000 Aarhus C, Denmark
Accepted 20 July 2006
Abstract Background: Dukes’ stage is the most important predictor of prognosis of colorectal cancer, but the association between delay of treatment (DT) and Dukes’ stage is still controversial. Methods: From 1 January 2001 to 31 July 2002, we conducted a population-based prospective observational study based on 733 Danish colorectal cancer patients. DT was determined through questionnaire-interviews, and Dukes’ stage was obtained from medical records and pathological forms. DT was classified into three groups: short (0–60 days), intermediate (61–150 days) and long (>150 days) DT. Dukes’ stage was classified into two groups: non-advanced (Dukes’ stage A or B) and advanced (Dukes’ stage C or D) cancer. Using relative risk (RR) the association between DT and stage was estimated, with short delay as the reference group. Results: The RR of advanced cancer was 1.0 (95% confidence intervals (CI): 0.8–1.3) for colon cancer patients with an intermediate DT, and 1.1 (95% CI: 0.9–1.4) for patients with a long DT. For rectal cancer patients the RR of advanced cancer was 1.9 (95% CI: 1.1–3.1) for patients with an intermediate DT and 2.1 (95% CI: 1.3–3.4) for patients with a long DT. Conclusion: DT was strongly associated with advanced stage of rectal cancer, but not of colon cancer. # 2006 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. Keywords: Colorectal cancer; Delay of treatment; Diagnostic delay; Stage; Duke’s stage; Conflicting results; Validation; Advanced cancer biology; Symptoms; Prognosis
1. Introduction Colorectal cancer is one of the most common cancers in western countries. Although the incidence and mortality are decreasing in many countries [1–3], colorectal cancer is a serious disease, and on average only 50% are alive 5 years after diagnosis [1–3]. It is estimated that worldwide there are about 400,000 deaths from colorectal cancer annually [1–3]. Dukes’ stage is one of the most important predictors of long-term outcome. In Denmark the expected 5-year survival for Dukes’ stage A is about 95%, for Dukes’ stage B is 50%, for Dukes’ stage C is 25%, and for Dukes’ stage D is 0%. At the time of diagnosis, 15% of all colorectal cancer patients are
* Corresponding author. Tel.: +45 2425 3304; fax: +45 8949 7718. E-mail address: [email protected] (M. Korsgaard).
at Dukes’ stage A, 30% are at Dukes’ stage B, 35% are at Dukes’ stage C, and 20% are at Dukes’ stage D [4]. There is increasing evidence that screening for colorectal cancer can result in longer survival due to early detection at a non-advanced Dukes’ stage [5]. Thus, it is likely that we could improve prognosis and increase survival if more patients were to be diagnosed in Dukes’ stage A or B. However, for symptomatic patients, many aspects of the association between delay of treatment (interval between onset of symptoms and surgery or other treatment) and stage are poorly understood. It has been suggested that delay of treatment may be a predictor of the stage of colorectal cancer; this idea, however, is controversial [6–24]. Several studies have indicated no association between diagnostic delay or delay of treatment and stage of colorectal cancer at the time of diagnosis [6–17,23]. However, five studies found delay of
0361-090X/$30.00 # 2006 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.cdp.2006.07.001
342
M. Korsgaard et al. / Cancer Detection and Prevention 30 (2006) 341–346
treatment associated with the stage of cancer for rectal cancer [18–22], and three of them found delay of treatment associated with stage for colon cancer, as well [20–22]. Relatively few patients were included in the existing prospectively obtained studies (between 100 and 554 colorectal cancer patients). Most studies were based on delay data obtained solely from medical records [8,12– 16,20,25]. In some studies, rectal cancer and colon cancer were combined in the analysis [6,8–11,14,15,17,23,24]. Other studies included either rectal cancer patients [12,19] or colon cancer patients [13,16], but not both. The aim of this study was to examine the association between delay of treatment and stage of cancer at time of operation for colorectal cancer, based on prospectively recorded data in a population-based setting.
2. Methods 2.1. Study population and period From 1 January 2001 to 30 June 2002, we conducted this prospective observational study in all the13 hospitals treating colorectal cancer, located in three counties in Denmark: Aarhus County (6 hospitals and about 630,000 inhabitants), and Ringkoebing and Ribe counties (7 hospitals and 475,000 inhabitants). Denmark has a tax-supported public health system for free hospital care in which private hospitals treat very few patients. More than 95% of Danes are registered with their own general practitioner who is paid by the county on a per fee basis with a certain amount of money per registered patient per year. 2.2. Patients Inclusion criteria were: (i) histologically verified, primary adenocarcinoma of the colon or rectum and, (ii) patients residing in our inclusion counties. Exclusion criteria were: (i) dementia/unable to cooperate, (ii) death before interview was possible, and (iii) inability to understand Danish. Further, we excluded patients who were under colonoscopic surveillance because of increased risk of cancer (hereditary non-polyposis colorectal cancer, attenuated familial adenomatous polyposis, familial adenomatous polyposis, or chronic inflammatory bowel disease). During the study period, 951 patients with primary colorectal cancer (598 colon cancer patients and 353 rectal cancer patients) were hospitalized. Of these, 147 patients (15.5%) were compromised by one of the exclusion criteria, and 61 patients (6.4%) refused to participate, leaving 743 patients (78.1%) for inclusion. These were 459 colon cancer patients and 284 rectal cancer patients, of whom 10 patients (3 colon cancer patients and 7 rectal cancer patients) were interviewed, but never had a resection and were not diagnosed as having carcinosis or radiologically verified liver- or lung-metastases. These 10 patients could not be
staged according to Dukes’ classification and thus were excluded from the analyses. This left a total of 733 patients available for the analyses. Surgeons and nurses who conducted the questionnaireinterviews checked daily for colorectal cancer patients in their departments to be included in the study. Further, the pathologists who examined the specimens and ultimately made the diagnosis sent a copy of the description of the histological diagnosis to Aarhus University Hospital. All patients gave informed consent. The study was approved by The Ethical Committees of Aarhus, Ringkoebing and Ribe Counties (1999/4678). 2.3. Delay data Delay of treatment was defined as the period from the date of the onset of symptoms of colorectal cancer until surgery or the start of preoperative radiotherapy or palliative treatment of the cancer. The questionnaire-interviews included questions about symptoms, date of onset of the symptoms, date of consulting a doctor, and how and when the doctor treated the symptoms, the type of symptoms, and dates of further examinations of colon or rectum. Most patients were interviewed prior to operation. If that was not possible, i.e., in case of acute surgery, the interview was done after operation, but always before the stage of the cancer was known from the histological examination. If a patient was unable to remember the exact date of the onset symptoms, but only remembered the month or the season, we chose the median date. If the same symptom had lasted for more than 4 years with no change, and was determined to be benign, the date of the doctor’s examination leading to the diagnosis was set to be the date of the onset symptom (two patients). Any obscurity in a questionnaire was followed-up by a phone call to the interviewer in order to avoid error. 2.4. Dukes’ stage Dukes’ classification, location of the cancer, emergency or elective surgery, along with other clinical data not used in this study, was registered by a local surgeon. In Denmark, the surgeon and the pathologist routinely fill in a form when a patient is operated for colorectal cancer. The form is a description of the specimen macroscopically as well as microscopically. Every electively operated colorectal cancer patient had preoperative ultrasonic examination of the liver, as well as a chest radiograph. 2.5. Validation of Dukes’ stage Dukes’ stage was validated for all patients by reviewing all medical records (after the patients had left the hospital), resulting in the correction of 3.5% of the Dukes’ stages,
M. Korsgaard et al. / Cancer Detection and Prevention 30 (2006) 341–346
343
Table 1 Descriptive data according to type of cancer
often because the medical records contained information about metastases of the lungs or liver, or because carcinosis was described.
Patients Women Men Median age
2.6. Statistical analysis We classified the 456 colon cancer patients and the 277 rectal cancer patients into three groups according to delay of treatment. The first group of patients had delay of treatment of up to 60 days, i.e., ‘‘short delay.’’ Denmark has a guarantee of diagnosis of a maximum of 14 days after referral for further examination, and a guarantee of treatment of another 14 days after diagnosis. We chose the short delay of treatment of up to 60 days as our reference group because it should be possible to diagnose and treat a colorectal cancer patient within this period, if the patient contacts a doctor about their symptoms soon after they appear, and the 14-day guarantees are met. The second group of patients with ‘‘intermediate delay’’ had delays of treatment of 61–150 days, and the third group of patients with a ‘‘long delay’’ had delays of treatment of more than 150 days. The last two delay groups were determined by attempting to equalize the number of patients in each of the groups, and by making the group of intermediate delay containing the median-value of delay of treatment. Dukes’ stage was classified into two groups, i.e., a group of non-advanced cancers including Dukes’ A and B, and a group of advanced cancers including Dukes’ C and D. We analyzed the data first by obtaining contingency tables for the main study variables. From these variables, we calculated the prevalence for advanced cancer according to delay of treatment along with the relative risk, and approximated 95% confidence intervals (CI). Further, we made the same analyses for delay groups with short delay of 90 days, intermediate delay of 91–180 days, and long delay of >180 days. Finally, we stratified the contingency tables according to age, gender, and type of surgery (elective/emergency). We analyzed for colon cancer and rectal cancer, separately. The statistical analyses were performed in the SPSS program (Version 10.0) and the Excel program.
Dukes’ stage A Dukes’ stage B Dukes’ stage C Dukes’ stage D Nonclassified Acutely operated Electively operated Never operated
Colon
Rectum
456 225 (49.0) 234 (51.0) 71 years (range 36–93 years) 63 (13.7) 161 (35.1) 137 (29.8) 95 (20.7) 3 (0.7) 87 (19.1) 366 (80.3) 3 (0.7)
277 119 (41.9) 165 (58.1) 68 years (range 31–92 years) 60 (21.1) 85 (29.9) 86 (30.3) 46 (16.2) 7 (2.5) 6 (2.2) 271 (97.8) 7 (2.5)
Number of patients (%).
3. Results 3.1. Descriptive data Table 1 shows descriptive data of the 456 included colon cancer patients and the 277 included rectal cancer patients. The distribution of main variables was equal between all patients and included patients (data not shown). 3.2. Delay The median delay for colon cancer patients was a little shorter than for rectal cancer patients (116 days versus 134 days). Table 2 shows data of delay of treatment. More colon cancer patients than rectal cancer patients had a short delay (27% versus 18%). Among colon cancer patients, more men than women had a short delay. More male than female rectal cancer patients had a long delay. There were more patients aged 70 years than patients aged >70 years in the group of intermediate or long delay, in particular among rectal cancer patients. 3.3. Delay and Dukes’ stage The frequencies of patients with a non-advanced cancer was higher in the group of short delay than in the groups of
Table 2 Delay of treatment according to localisation of the cancer for all patients and for subgroups Colon (N = 456)
All patients Electively operated Acutely operated Men Women Age 70 Age > 70
Rectum (N = 277)
Short delay
Intermediate delay
Long delay
Short delay
Intermediate delay
Long delay
121 85a 35a 74 47 64 57
150 127 23 69 81 73 77
185 157 28 89 96 96 89
49 46 3 27 22 25 24
115 113a 1a 64 51 73 42
113 112 0 72 41 69 44
(27) (23) (41) (32) (21) (28) (26)
(33) (34) (27) (30) (36) (31) (34)
(41) (43) (32) (38) (43) (41) (40)
Number of patients and for all patients (frequency in %). a One patient was never operated but radiologically classified to Dukes’ stage D.
(18) (17) (75) (11) (19) (15) (22)
(42) (42) (25) (25) (45) (44) (38)
(41) (41) (0) (28) (36) (41) (40)
344
M. Korsgaard et al. / Cancer Detection and Prevention 30 (2006) 341–346
Table 3 Number (frequency in %) of patients in each delay group with the different Dukes’ stages Delay periods (days)
Dukes’ stage
Colon
Rectum
0–60
A B C D
19 40 34 28
(15.7) (33.1) (28.1) (23.1)
12 24 8 5
(24.5) (49.0) (16.3) (10.2)
61–150
A B C D
21 51 46 32
(14.0) (34.0) (30.7) (21.3)
26 32 38 19
(22.6) (27.8) (33.0) (16.5)
>150
A B C D
23 70 28 57
(12.4) (37.8) (15.1) (30.8)
22 29 40 22
(19.5) (25.7) (35.4) (19.5)
intermediate and long delay for rectal cancer patients, whereas there was no substantial difference for colon cancer patients (Table 3). Table 4 shows no significant increased risk of having an advanced cancer (Dukes’ stage C or D) at time of surgery for colon cancer patients having an intermediate delay (relative risk of 1.0 (95% CI: 0.8–1.3)) or a long delay (relative risk of 1.1 (95% CI: 0.9–1.4)) compared with those with a short delay. For rectal cancer patients, there was a reasonably increased risk of having an advanced cancer (Dukes’ stage C or D) at time of surgery for patients having an intermediate Table 4 Relative risk (95% CI) of having an advanced cancer at time of operation with short delay (0–60 days) as reference group Group of patients
Delay periods (days)
Colon
Rectum
All patients
0–60 61–150 >150
1.0 (Reference) 1.0 (0.8–1.3) 1.1 (0.9–1.4)
1.0 (Reference) 1.9 (1.1–3.1) 2.1 (1.3–3.4)
Elective operation
0–60 61–150 >150
1.0 (Reference) 1.1 (0.8–1.5) 1.1 (0.8–1.5)
1.0 (Reference) 1.7 (1.1–2.8) 1.9 (1.2–3.2)
Acute operation
0–60 61–150 >150
1.0 (Reference) 1.0 (0.7–1.4) 0.7 (0.5–1.1)
– – –
Men
0–60 61–150 >150
1.0 (Reference) 1.1 (0.8–1.5) 1.0 (0.7–1.4)
1.0 (Reference) 1.8 (1.0–3.4) 1.9 (1.0–3.5)
Women
0–60 61–150 >150
1.0 (Reference) 1.1 (0.8–1.7) 1.1 (0.7–1.7)
1.0 (Reference) 2.0 (0.9–4.5) 2.5 (1.1–5.6)
Age 70 years
0–60 61–150 >150
1.0 (Reference) 1.1 (0.8–1.5) 1.0 (0.7–1.3)
1.0 (Reference) 1.5 (0.8–2.5) 1.7 (1.0–2.9)
Age > 70 years
0–60 61–150 >150
1.0 (Reference) 0.9 (0.6–1.3) 1.0 (0.7–1.4)
1.0 (Reference) 2.7 (1.0–7.1) 2.7 (1.1–7.1)
Stratified analyses.
delay (relative risk of 1.9 (95% CI: 1.1–3.1)) or a long delay (relative risk of 2.1 (95% CI: 1.3–3.4)) compared with those with a short delay. If the delay groups were made with other intervals, i.e., short delay 90 days, median delay 91–180 days, and long delay >180 days, the relative risks of having an advanced cancer in case of intermediate or long delay were about the same for colon cancer as with the original delay groups (intermediate delay 0.9 CI: 0.7–1.1 and long delay 0.9 CI: 0.8–1.2) indicating no association between delay and stage, and the relative risks of having an advanced cancer were reduced a little for rectal cancer compared to the relative risk estimated with the original delay groups (intermediate delay 1.7 CI: 1.2–2.4 and long delay 1.6 CI: 1.1–2.2), but the association between delay and stage was still significant. Stratifying on age, the association was strongest for older rectal cancer patients (age > 70 years); the relative risk for patients with an intermediate delay was 2.7 (95% CI: 1.0– 7.1) and for patients with a long delay was 2.7 (95%CI: 1.1– 7.1) (Table 4). When stratified on gender, the relative risk estimate was higher for women (2.5 (95% CI: 1.1–5.6)) than for men (1.9 (95% CI: 1.0–3.5)) for patients with a long delay (Table 4).
4. Discussion In this large population-based study, we found delay of treatment strongly associated with the risk of having an advanced cancer at time of operation for rectal cancer, but not for colon cancer. To the best of our knowledge, this study has the largest number of prospective, registered colorectal cancer patients concerning delay in a population-based setting. The main strengths of our study are its large size, the uniformly organized free and tax-supported health care system with primary catchment areas allowing a population-based design, the lack of screening, and the completeness of follow-up. The weakness includes our inability, as in other studies, to obtain prospective data at the onset of symptoms. Therefore, every questionnaire/interview study contains risk of inaccurate reporting. Symptoms of colorectal cancer are often vague and many of the symptoms are the same in benign diseases, increasing the risk of information bias. However, blinding of both the patient and the interviewer before cancer staging is likely to lead to non-differential misclassification. Such misclassification will cause bias of relative risk estimates towards one. A relationship between delay of treatment and stage may be biased by the choice of different ranges of delay, and might have contributed to the strong association between delay of treatment and stage of rectal cancer. Previous studies have examined the association between delay of treatment and stage using different statistical methods. Some of the studies have divided the delay into groups or intervals, and have chosen intervals of different length. To ensure that the strong
M. Korsgaard et al. / Cancer Detection and Prevention 30 (2006) 341–346
association between delay of treatment and stage of rectal cancer was still present, we chose other intervals of delay and analyzed the association between delay of treatment and stage again, in the same way as initially done; as expected, there was still no association between delay of treatment and stage for colon cancer, whereas the relative risk of having an advanced rectal cancer after an intermediate delay and a long delay declined a little, but the association was still significant, thereby fortifying our results. We found a strong association between delay of treatment and stage of rectal cancer. It is evident that the risk of having an advanced cancer after an intermediate and a long diagnostic delay is increased for the group of patients aged >70 years, which is worth remembering when elderly people present with gastrointestinal symptoms suspicious of rectal cancer. That the risk is higher for women with a long delay than for men with a long delay might only, in general, reflect the larger number of elderly women than elderly men. Confirming most former studies [6–17,23], we could not show any association between delay of treatment and Dukes’ stage for colon cancer patients. An explanation as to why an association between delay of treatment and the stage of colon cancer cannot be shown, could be differences in the biology of the cancers, i.e., as suggested by a former study, some being faster growing and thus unable to show an association between delay of treatment and stage of colorectal cancer at the time of operation [6]. This could be suspected for the colon cancer patients, with an acute onset, having a very short delay period, but an advanced cancer. Another explanation could be the vague symptoms (many of them equal to symptoms of benign diseases) [26–36] characterizing colon cancer that is difficult to affirm precisely. If this is the explanation, the prognosis of colon cancer is not likely to be improved in symptomatic patients, focusing on diagnostic delay; the cancer must be diagnosed presymptomatically to improve prognosis, i.e., by screening [5]. Four studies showed an association between delay and stage of colon cancer [20–22,24]. Few patients, between 70 and 445 patients were included in the studies based on statistical analyses using different delay-intervals (6 weeks– 3 months). The patients were prospectively included in only one of the studies [21]. The existing observational studies have given conflicting results about delay and stage of rectal cancer. Most former studies contradict our result [6–17,23]. Holliday and Hardcastle [9], for example, examined 100 colon cancer patients and 100 rectal cancer patients and found no association between delay and stage comparing the average delay periods in the different Dukes’ stages— the median delay periods were not mentioned according to stage. Rectal cancer patients were analyzed combined with the colon cancer patients as in other studies finding no association between delay and stage for rectal cancer [6,8,10,11,14,15,17,23]. In accordance with our study, six studies [18–22,24] found an association between delay of treatment and stage of
345
rectal cancer. None of the studies included as many patients as in this study, and various statistical methods and choices of intervals were used. None of the studies showed the same statistical strength of a significant association between delay of treatment and stage of rectal cancer as we did. Symptoms of rectal cancer tend to be better defined than symptoms of colon cancer and thus easier to affirm [18,27,37–39]. The patients might be more likely to react to these symptoms than the vague symptoms more frequently experienced for colon cancer. The question as to why an association between delay and stage of cancer can be shown so convincingly for rectal cancer but not for colon cancer might be found by associating symptoms with the location of the cancer, as symptoms of colon cancer are known to be more vague than symptoms of rectal cancer. In conclusion, we found a strong association between delay of treatment and stage of the cancer at time of operation for rectal cancer, but not for colon cancer, which is important knowledge in aiming to improve survival. We should focus on early diagnosis for rectal cancer patients, diagnosing more of them when they are in Dukes’ stage A or B. Acknowledgements Contributors include: Aarhus University Hospital— Consultants: Søren Laurberg, Marianne Korsgaard MD, Professor Henrik Toft Sørensen, Chief Biostatistician Lars Pedersen, Department of Clinical Epidemiology, Esbjerg Hospital—Consultants: Karl Erik Juul Jensen, Peiman Poornoroozy, Vagn Juhl Jensen, Grena˚ Hospital—Consultant: Jan Lindholt, Grindsted Hospital—Consultants: Anders Larsen, Vagn Berg, Herning Hospital—Consultants: Mogens Rørbæk, Holstebro Hospital—Consultants: Mads Mark, Erik Skoubo Kristensen, Henrik Sloth, Lemvig Hospital—Consultant: Ivan Nørbæk, Odder Hospital— Consultant: Gunnar Madsen, Randers Hospital—Consultants: Knud Thygesen, Frank Svendsen Jensen, Ringkøbing Hospital—Consultant: Ebbe Fuglsang, Silkeborg Hospital—Consultants: Ole Brandsborg, Erling Østergaard. Sources of support were: Grants from Western Danish Research Forum (Vestdansk Forskningsforum); Danish Medical Research Council (Statens Sundhedsvidenskabelige Forskningsra˚d); Manufacturer Einar Willumsen’s Memorial Award (Fabrikant Einar Willumsens Mindelegat); Dagmar Marshall’s Fund (Dagmar Marshalls Fond); Medical Student Karsten Hansen’s Memorial Award (Stud. med. Karsten Hansens Mindelegat); and Else and Mogens Wedell-Wedellborg’s Fund (Else og Mogens WedellWedellborgs Fond).
References [1] Pisani P, Parkin DM, Ferlay J. stimates of the worldwide mortality from eighteen major cancers in 1985. Implications for prevention and projections of future burden. Int J Cancer 1993;55:891–903.
346
M. Korsgaard et al. / Cancer Detection and Prevention 30 (2006) 341–346
[2] Pisani P, Parkin DM, Bray F, Ferlay J. Estimates of the worldwide mortality from 25 cancers in 1990. Int J Cancer 1999;83:18–29. [3] Wingo PA, Cardinez CJ, Landis SH, Greenlee RT, Ries LA, Anderson RN, et al. Long-term trends in cancer mortality in the United States, 1930–1998. Cancer 2003;97:3133–275. [4] Statens Institut for Medicinsk TeknologivurderingKræft i tyktarm og endetarm. Diagnostik og screening. Medicinsk Teknologivurdering. Medicinsk Teknologivurdering; 2001. [5] Kronborg O. Screening for early colorectal cancer. World J Surg 2000;24:1069–74. [6] Ratcliffe R, Kiff RS, Hoare EM, Kingston RD, Walsh SH, Jeacock J. Early diagnosis in colorectal cancer still no benefit? Ann Chir 1989;43:570–4. [7] Graffner H, Olsson SA. Patient’s and doctor’s delay in carcinoma of the colon and rectum. J Surg Oncol 1986;31:188–90. [8] Pescatori M, Maria G, Beltrani B, Mattana C. Site emergency, and duration of symptoms in the prognosis of colorectal cancer. Dis Colon Rectum 1982;25:33–40. [9] Holliday HW, Hardcastle JD. Delay in diagnosis and treatment of symptomatic colorectal cancer. Lancet 1979;1:309–11. [10] Kyle SM, Isbister WH, Yeong ML. Presentation, duration of symptoms and staging of colorectal carcinoma. Aust N Z J Surg 1991;61: 137–40. [11] Irvin TT, Greaney MG. Duration of symptoms and prognosis of carcinoma of the colon and rectum. Surg Gynecol Obstet 1977; 144:883–6. [12] Devlin HB, Plant JA, Morris D. The significance of the symptoms of carcinoma of the rectum. Surg Gynecol Obstet 1973;137:399–402. [13] Goodman D, Irvin TT. Delay in the diagnosis and prognosis of carcinoma of the right colon. Br J Surg 1993;80:1327–9. [14] Jolly KD, Scott JP, MacKinnon MJ, Clarke AM. Diagnosis and survival in carcinoma of the large bowel. Aust N Z J Surg 1982; 52:12–6. [15] Khubchandani M. Relationship of symptom duration and survival in patients with carcinoma of the colon and rectum. Dis Colon Rectum 1985;28:585–7. [16] McDermott FT, Hughes ES, Pihl E, Milne BJ, Price AB. Prognosis in relation to symptom duration in colon cancer. Br J Surg 1981;68:846–9. [17] Gonzalez-Hermoso F, Perez-Palma J, Marchena-Gomez J, LorenzoRocha N, Medina-Arana V. Can early diagnosis of symptomatic colorectal cancer improve the prognosis? World J Surg 2004;28:716–20. [18] Arbman G, Nilsson E, Storgren-Fordell V, Sjodahl R. A short diagnostic delay is more important for rectal cancer than for colonic cancer. Eur J Surg 1996;162:899–904. [19] Pahlman L, Glimelius B, Enblad P. Clinical characteristics and their relation to surgical curability in adenocarcinoma of the rectum and rectosigmoid. A population-based study on 279 consecutive patients. Acta Chir Scand 1985;151:685–93. [20] Robinson E, Mohilever J, Zidan J, Sapir D. Colorectal cancer: incidence, delay in diagnosis and stage of disease. Eur J Cancer Clin Oncol 1986;22:157–61.
[21] Rowejones DC, Aylett SO. Delay in treatment in carcinoma of colon and rectum. Lancet 1965;2:973–6. [22] Langenbach MR, Schmidt J, Neumann J, Zirngibl H. Delay in treatment of colorectal cancer: multifactorial problem. World J Surg 2003;27:304–8. [23] Bharucha S, Hughes S, Kenyon V, Anderson ID, Carlson GL, Scott NA. Targets and elective colorectal cancer: outcome and symptom delay at surgical resection. Colorectal Dis 2005;7:169–71. [24] Young CJ, Sweeney JL, Hunter A. Implications of delayed diagnosis in colorectal cancer. Aust N Z J Surg 2000;70:635–8. [25] Hackett TP, Cassem NH, Raker JW. Patient delay in cancer. N Engl J Med 1973;289:14–20. [26] Goulston KJ, Cook I, Dent OF. How important is rectal bleeding in the diagnosis of bowel cancer and polyps? Lancet 1986;2:261–5. [27] MacDonald L, Freeling P. Bowels: beliefs and behaviour. Fam Pract 1986;3:80–4. [28] Kay L, Jorgensen T, Schultz-Larsen K. Abdominal pain in a 70year-old Danish population. An epidemiological study of the prevalence and importance of abdominal pain. J Clin Epidemiol 1992; 45:1377–82. [29] Selvachandran SN, Hodder RJ, Ballal MS, Jones P, Cade D. Prediction of colorectal cancer by a patient consultation questionnaire and scoring system: a prospective study. Lancet 2002;360:278–83. [30] Byles JE, Redman S, Hennrikus D, Sanson-Fisher RW, Dickinson J. Delay in consulting a medical practitioner about rectal bleeding. J Epidemiol Community Health 1992;46:241–4. [31] Curless R, French J, Williams GV, James OF. Comparison of gastrointestinal symptoms in colorectal carcinoma patients and community controls with respect to age. Gut 1994;35:1267–70. [32] Silman AJ, Mitchell P, Nicholls RJ, Macrae FA, Leicester RJ, Bartram CI, et al. Self-reported dark red bleeding as a marker comparable with occult blood testing in screening for large bowel neoplasms. Br J Surg 1983;70:721–4. [33] Goulston K, Chapuis P, Dent O, Bokey L. Significance of bowel symptoms. Med J Aust 1987;146:631–3. [34] Farrands PA, Hardcastle JD, Chamberlain J, Moss S. Factors affecting compliance with screening for colorectal cancer. Community Med 1984;6:12–9. [35] Chapuis PH, Goulston KJ, Dent OF, Tait AD. Predictive value of rectal bleeding in screening for rectal and sigmoid polyps. Br Med J (Clin Res Ed) 1985;290:1546–8. [36] Dent OF, Goulston KJ, Zubrzycki J, Chapuis PH. Bowel symptoms in an apparently well population. Dis Colon Rectum 1986;29:243–7. [37] MacArthur C, Smith A. Factors associated with speed of diagnosis, referral, and treatment in colorectal cancer. J Epidemiol Community Health 1984;38:122–6. [38] Wobbes T. Carcinoma of the colon and rectum in geriatric patients. Age Ageing 1985;14:321–6. [39] Majumdar SR, Fletcher RH, Evans AT. How does colorectal cancer present? Symptoms, duration, and clues to location. Am J Gastroenterol 1999;94:3039–45.
Delay of treatment is associated with advanced stage of rectal cancer but not of colon cancer Marianne Korsgaard MD, PhDa,*, Lars Pedersen MScb, Henrik Toft Sørensen DMScb, Søren Laurberg DMSca b
a Department of Surgery L, Aarhus University Hospital, Tage Hansensgade 2, 8000 Aarhus C, Denmark Department of Clinical Epidemiology, Aarhus University Hospital, Tage Hansensgade 2, 8000 Aarhus C, Denmark
Accepted 20 July 2006
Abstract Background: Dukes’ stage is the most important predictor of prognosis of colorectal cancer, but the association between delay of treatment (DT) and Dukes’ stage is still controversial. Methods: From 1 January 2001 to 31 July 2002, we conducted a population-based prospective observational study based on 733 Danish colorectal cancer patients. DT was determined through questionnaire-interviews, and Dukes’ stage was obtained from medical records and pathological forms. DT was classified into three groups: short (0–60 days), intermediate (61–150 days) and long (>150 days) DT. Dukes’ stage was classified into two groups: non-advanced (Dukes’ stage A or B) and advanced (Dukes’ stage C or D) cancer. Using relative risk (RR) the association between DT and stage was estimated, with short delay as the reference group. Results: The RR of advanced cancer was 1.0 (95% confidence intervals (CI): 0.8–1.3) for colon cancer patients with an intermediate DT, and 1.1 (95% CI: 0.9–1.4) for patients with a long DT. For rectal cancer patients the RR of advanced cancer was 1.9 (95% CI: 1.1–3.1) for patients with an intermediate DT and 2.1 (95% CI: 1.3–3.4) for patients with a long DT. Conclusion: DT was strongly associated with advanced stage of rectal cancer, but not of colon cancer. # 2006 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. Keywords: Colorectal cancer; Delay of treatment; Diagnostic delay; Stage; Duke’s stage; Conflicting results; Validation; Advanced cancer biology; Symptoms; Prognosis
1. Introduction Colorectal cancer is one of the most common cancers in western countries. Although the incidence and mortality are decreasing in many countries [1–3], colorectal cancer is a serious disease, and on average only 50% are alive 5 years after diagnosis [1–3]. It is estimated that worldwide there are about 400,000 deaths from colorectal cancer annually [1–3]. Dukes’ stage is one of the most important predictors of long-term outcome. In Denmark the expected 5-year survival for Dukes’ stage A is about 95%, for Dukes’ stage B is 50%, for Dukes’ stage C is 25%, and for Dukes’ stage D is 0%. At the time of diagnosis, 15% of all colorectal cancer patients are
* Corresponding author. Tel.: +45 2425 3304; fax: +45 8949 7718. E-mail address: [email protected] (M. Korsgaard).
at Dukes’ stage A, 30% are at Dukes’ stage B, 35% are at Dukes’ stage C, and 20% are at Dukes’ stage D [4]. There is increasing evidence that screening for colorectal cancer can result in longer survival due to early detection at a non-advanced Dukes’ stage [5]. Thus, it is likely that we could improve prognosis and increase survival if more patients were to be diagnosed in Dukes’ stage A or B. However, for symptomatic patients, many aspects of the association between delay of treatment (interval between onset of symptoms and surgery or other treatment) and stage are poorly understood. It has been suggested that delay of treatment may be a predictor of the stage of colorectal cancer; this idea, however, is controversial [6–24]. Several studies have indicated no association between diagnostic delay or delay of treatment and stage of colorectal cancer at the time of diagnosis [6–17,23]. However, five studies found delay of
0361-090X/$30.00 # 2006 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.cdp.2006.07.001
342
M. Korsgaard et al. / Cancer Detection and Prevention 30 (2006) 341–346
treatment associated with the stage of cancer for rectal cancer [18–22], and three of them found delay of treatment associated with stage for colon cancer, as well [20–22]. Relatively few patients were included in the existing prospectively obtained studies (between 100 and 554 colorectal cancer patients). Most studies were based on delay data obtained solely from medical records [8,12– 16,20,25]. In some studies, rectal cancer and colon cancer were combined in the analysis [6,8–11,14,15,17,23,24]. Other studies included either rectal cancer patients [12,19] or colon cancer patients [13,16], but not both. The aim of this study was to examine the association between delay of treatment and stage of cancer at time of operation for colorectal cancer, based on prospectively recorded data in a population-based setting.
2. Methods 2.1. Study population and period From 1 January 2001 to 30 June 2002, we conducted this prospective observational study in all the13 hospitals treating colorectal cancer, located in three counties in Denmark: Aarhus County (6 hospitals and about 630,000 inhabitants), and Ringkoebing and Ribe counties (7 hospitals and 475,000 inhabitants). Denmark has a tax-supported public health system for free hospital care in which private hospitals treat very few patients. More than 95% of Danes are registered with their own general practitioner who is paid by the county on a per fee basis with a certain amount of money per registered patient per year. 2.2. Patients Inclusion criteria were: (i) histologically verified, primary adenocarcinoma of the colon or rectum and, (ii) patients residing in our inclusion counties. Exclusion criteria were: (i) dementia/unable to cooperate, (ii) death before interview was possible, and (iii) inability to understand Danish. Further, we excluded patients who were under colonoscopic surveillance because of increased risk of cancer (hereditary non-polyposis colorectal cancer, attenuated familial adenomatous polyposis, familial adenomatous polyposis, or chronic inflammatory bowel disease). During the study period, 951 patients with primary colorectal cancer (598 colon cancer patients and 353 rectal cancer patients) were hospitalized. Of these, 147 patients (15.5%) were compromised by one of the exclusion criteria, and 61 patients (6.4%) refused to participate, leaving 743 patients (78.1%) for inclusion. These were 459 colon cancer patients and 284 rectal cancer patients, of whom 10 patients (3 colon cancer patients and 7 rectal cancer patients) were interviewed, but never had a resection and were not diagnosed as having carcinosis or radiologically verified liver- or lung-metastases. These 10 patients could not be
staged according to Dukes’ classification and thus were excluded from the analyses. This left a total of 733 patients available for the analyses. Surgeons and nurses who conducted the questionnaireinterviews checked daily for colorectal cancer patients in their departments to be included in the study. Further, the pathologists who examined the specimens and ultimately made the diagnosis sent a copy of the description of the histological diagnosis to Aarhus University Hospital. All patients gave informed consent. The study was approved by The Ethical Committees of Aarhus, Ringkoebing and Ribe Counties (1999/4678). 2.3. Delay data Delay of treatment was defined as the period from the date of the onset of symptoms of colorectal cancer until surgery or the start of preoperative radiotherapy or palliative treatment of the cancer. The questionnaire-interviews included questions about symptoms, date of onset of the symptoms, date of consulting a doctor, and how and when the doctor treated the symptoms, the type of symptoms, and dates of further examinations of colon or rectum. Most patients were interviewed prior to operation. If that was not possible, i.e., in case of acute surgery, the interview was done after operation, but always before the stage of the cancer was known from the histological examination. If a patient was unable to remember the exact date of the onset symptoms, but only remembered the month or the season, we chose the median date. If the same symptom had lasted for more than 4 years with no change, and was determined to be benign, the date of the doctor’s examination leading to the diagnosis was set to be the date of the onset symptom (two patients). Any obscurity in a questionnaire was followed-up by a phone call to the interviewer in order to avoid error. 2.4. Dukes’ stage Dukes’ classification, location of the cancer, emergency or elective surgery, along with other clinical data not used in this study, was registered by a local surgeon. In Denmark, the surgeon and the pathologist routinely fill in a form when a patient is operated for colorectal cancer. The form is a description of the specimen macroscopically as well as microscopically. Every electively operated colorectal cancer patient had preoperative ultrasonic examination of the liver, as well as a chest radiograph. 2.5. Validation of Dukes’ stage Dukes’ stage was validated for all patients by reviewing all medical records (after the patients had left the hospital), resulting in the correction of 3.5% of the Dukes’ stages,
M. Korsgaard et al. / Cancer Detection and Prevention 30 (2006) 341–346
343
Table 1 Descriptive data according to type of cancer
often because the medical records contained information about metastases of the lungs or liver, or because carcinosis was described.
Patients Women Men Median age
2.6. Statistical analysis We classified the 456 colon cancer patients and the 277 rectal cancer patients into three groups according to delay of treatment. The first group of patients had delay of treatment of up to 60 days, i.e., ‘‘short delay.’’ Denmark has a guarantee of diagnosis of a maximum of 14 days after referral for further examination, and a guarantee of treatment of another 14 days after diagnosis. We chose the short delay of treatment of up to 60 days as our reference group because it should be possible to diagnose and treat a colorectal cancer patient within this period, if the patient contacts a doctor about their symptoms soon after they appear, and the 14-day guarantees are met. The second group of patients with ‘‘intermediate delay’’ had delays of treatment of 61–150 days, and the third group of patients with a ‘‘long delay’’ had delays of treatment of more than 150 days. The last two delay groups were determined by attempting to equalize the number of patients in each of the groups, and by making the group of intermediate delay containing the median-value of delay of treatment. Dukes’ stage was classified into two groups, i.e., a group of non-advanced cancers including Dukes’ A and B, and a group of advanced cancers including Dukes’ C and D. We analyzed the data first by obtaining contingency tables for the main study variables. From these variables, we calculated the prevalence for advanced cancer according to delay of treatment along with the relative risk, and approximated 95% confidence intervals (CI). Further, we made the same analyses for delay groups with short delay of 90 days, intermediate delay of 91–180 days, and long delay of >180 days. Finally, we stratified the contingency tables according to age, gender, and type of surgery (elective/emergency). We analyzed for colon cancer and rectal cancer, separately. The statistical analyses were performed in the SPSS program (Version 10.0) and the Excel program.
Dukes’ stage A Dukes’ stage B Dukes’ stage C Dukes’ stage D Nonclassified Acutely operated Electively operated Never operated
Colon
Rectum
456 225 (49.0) 234 (51.0) 71 years (range 36–93 years) 63 (13.7) 161 (35.1) 137 (29.8) 95 (20.7) 3 (0.7) 87 (19.1) 366 (80.3) 3 (0.7)
277 119 (41.9) 165 (58.1) 68 years (range 31–92 years) 60 (21.1) 85 (29.9) 86 (30.3) 46 (16.2) 7 (2.5) 6 (2.2) 271 (97.8) 7 (2.5)
Number of patients (%).
3. Results 3.1. Descriptive data Table 1 shows descriptive data of the 456 included colon cancer patients and the 277 included rectal cancer patients. The distribution of main variables was equal between all patients and included patients (data not shown). 3.2. Delay The median delay for colon cancer patients was a little shorter than for rectal cancer patients (116 days versus 134 days). Table 2 shows data of delay of treatment. More colon cancer patients than rectal cancer patients had a short delay (27% versus 18%). Among colon cancer patients, more men than women had a short delay. More male than female rectal cancer patients had a long delay. There were more patients aged 70 years than patients aged >70 years in the group of intermediate or long delay, in particular among rectal cancer patients. 3.3. Delay and Dukes’ stage The frequencies of patients with a non-advanced cancer was higher in the group of short delay than in the groups of
Table 2 Delay of treatment according to localisation of the cancer for all patients and for subgroups Colon (N = 456)
All patients Electively operated Acutely operated Men Women Age 70 Age > 70
Rectum (N = 277)
Short delay
Intermediate delay
Long delay
Short delay
Intermediate delay
Long delay
121 85a 35a 74 47 64 57
150 127 23 69 81 73 77
185 157 28 89 96 96 89
49 46 3 27 22 25 24
115 113a 1a 64 51 73 42
113 112 0 72 41 69 44
(27) (23) (41) (32) (21) (28) (26)
(33) (34) (27) (30) (36) (31) (34)
(41) (43) (32) (38) (43) (41) (40)
Number of patients and for all patients (frequency in %). a One patient was never operated but radiologically classified to Dukes’ stage D.
(18) (17) (75) (11) (19) (15) (22)
(42) (42) (25) (25) (45) (44) (38)
(41) (41) (0) (28) (36) (41) (40)
344
M. Korsgaard et al. / Cancer Detection and Prevention 30 (2006) 341–346
Table 3 Number (frequency in %) of patients in each delay group with the different Dukes’ stages Delay periods (days)
Dukes’ stage
Colon
Rectum
0–60
A B C D
19 40 34 28
(15.7) (33.1) (28.1) (23.1)
12 24 8 5
(24.5) (49.0) (16.3) (10.2)
61–150
A B C D
21 51 46 32
(14.0) (34.0) (30.7) (21.3)
26 32 38 19
(22.6) (27.8) (33.0) (16.5)
>150
A B C D
23 70 28 57
(12.4) (37.8) (15.1) (30.8)
22 29 40 22
(19.5) (25.7) (35.4) (19.5)
intermediate and long delay for rectal cancer patients, whereas there was no substantial difference for colon cancer patients (Table 3). Table 4 shows no significant increased risk of having an advanced cancer (Dukes’ stage C or D) at time of surgery for colon cancer patients having an intermediate delay (relative risk of 1.0 (95% CI: 0.8–1.3)) or a long delay (relative risk of 1.1 (95% CI: 0.9–1.4)) compared with those with a short delay. For rectal cancer patients, there was a reasonably increased risk of having an advanced cancer (Dukes’ stage C or D) at time of surgery for patients having an intermediate Table 4 Relative risk (95% CI) of having an advanced cancer at time of operation with short delay (0–60 days) as reference group Group of patients
Delay periods (days)
Colon
Rectum
All patients
0–60 61–150 >150
1.0 (Reference) 1.0 (0.8–1.3) 1.1 (0.9–1.4)
1.0 (Reference) 1.9 (1.1–3.1) 2.1 (1.3–3.4)
Elective operation
0–60 61–150 >150
1.0 (Reference) 1.1 (0.8–1.5) 1.1 (0.8–1.5)
1.0 (Reference) 1.7 (1.1–2.8) 1.9 (1.2–3.2)
Acute operation
0–60 61–150 >150
1.0 (Reference) 1.0 (0.7–1.4) 0.7 (0.5–1.1)
– – –
Men
0–60 61–150 >150
1.0 (Reference) 1.1 (0.8–1.5) 1.0 (0.7–1.4)
1.0 (Reference) 1.8 (1.0–3.4) 1.9 (1.0–3.5)
Women
0–60 61–150 >150
1.0 (Reference) 1.1 (0.8–1.7) 1.1 (0.7–1.7)
1.0 (Reference) 2.0 (0.9–4.5) 2.5 (1.1–5.6)
Age 70 years
0–60 61–150 >150
1.0 (Reference) 1.1 (0.8–1.5) 1.0 (0.7–1.3)
1.0 (Reference) 1.5 (0.8–2.5) 1.7 (1.0–2.9)
Age > 70 years
0–60 61–150 >150
1.0 (Reference) 0.9 (0.6–1.3) 1.0 (0.7–1.4)
1.0 (Reference) 2.7 (1.0–7.1) 2.7 (1.1–7.1)
Stratified analyses.
delay (relative risk of 1.9 (95% CI: 1.1–3.1)) or a long delay (relative risk of 2.1 (95% CI: 1.3–3.4)) compared with those with a short delay. If the delay groups were made with other intervals, i.e., short delay 90 days, median delay 91–180 days, and long delay >180 days, the relative risks of having an advanced cancer in case of intermediate or long delay were about the same for colon cancer as with the original delay groups (intermediate delay 0.9 CI: 0.7–1.1 and long delay 0.9 CI: 0.8–1.2) indicating no association between delay and stage, and the relative risks of having an advanced cancer were reduced a little for rectal cancer compared to the relative risk estimated with the original delay groups (intermediate delay 1.7 CI: 1.2–2.4 and long delay 1.6 CI: 1.1–2.2), but the association between delay and stage was still significant. Stratifying on age, the association was strongest for older rectal cancer patients (age > 70 years); the relative risk for patients with an intermediate delay was 2.7 (95% CI: 1.0– 7.1) and for patients with a long delay was 2.7 (95%CI: 1.1– 7.1) (Table 4). When stratified on gender, the relative risk estimate was higher for women (2.5 (95% CI: 1.1–5.6)) than for men (1.9 (95% CI: 1.0–3.5)) for patients with a long delay (Table 4).
4. Discussion In this large population-based study, we found delay of treatment strongly associated with the risk of having an advanced cancer at time of operation for rectal cancer, but not for colon cancer. To the best of our knowledge, this study has the largest number of prospective, registered colorectal cancer patients concerning delay in a population-based setting. The main strengths of our study are its large size, the uniformly organized free and tax-supported health care system with primary catchment areas allowing a population-based design, the lack of screening, and the completeness of follow-up. The weakness includes our inability, as in other studies, to obtain prospective data at the onset of symptoms. Therefore, every questionnaire/interview study contains risk of inaccurate reporting. Symptoms of colorectal cancer are often vague and many of the symptoms are the same in benign diseases, increasing the risk of information bias. However, blinding of both the patient and the interviewer before cancer staging is likely to lead to non-differential misclassification. Such misclassification will cause bias of relative risk estimates towards one. A relationship between delay of treatment and stage may be biased by the choice of different ranges of delay, and might have contributed to the strong association between delay of treatment and stage of rectal cancer. Previous studies have examined the association between delay of treatment and stage using different statistical methods. Some of the studies have divided the delay into groups or intervals, and have chosen intervals of different length. To ensure that the strong
M. Korsgaard et al. / Cancer Detection and Prevention 30 (2006) 341–346
association between delay of treatment and stage of rectal cancer was still present, we chose other intervals of delay and analyzed the association between delay of treatment and stage again, in the same way as initially done; as expected, there was still no association between delay of treatment and stage for colon cancer, whereas the relative risk of having an advanced rectal cancer after an intermediate delay and a long delay declined a little, but the association was still significant, thereby fortifying our results. We found a strong association between delay of treatment and stage of rectal cancer. It is evident that the risk of having an advanced cancer after an intermediate and a long diagnostic delay is increased for the group of patients aged >70 years, which is worth remembering when elderly people present with gastrointestinal symptoms suspicious of rectal cancer. That the risk is higher for women with a long delay than for men with a long delay might only, in general, reflect the larger number of elderly women than elderly men. Confirming most former studies [6–17,23], we could not show any association between delay of treatment and Dukes’ stage for colon cancer patients. An explanation as to why an association between delay of treatment and the stage of colon cancer cannot be shown, could be differences in the biology of the cancers, i.e., as suggested by a former study, some being faster growing and thus unable to show an association between delay of treatment and stage of colorectal cancer at the time of operation [6]. This could be suspected for the colon cancer patients, with an acute onset, having a very short delay period, but an advanced cancer. Another explanation could be the vague symptoms (many of them equal to symptoms of benign diseases) [26–36] characterizing colon cancer that is difficult to affirm precisely. If this is the explanation, the prognosis of colon cancer is not likely to be improved in symptomatic patients, focusing on diagnostic delay; the cancer must be diagnosed presymptomatically to improve prognosis, i.e., by screening [5]. Four studies showed an association between delay and stage of colon cancer [20–22,24]. Few patients, between 70 and 445 patients were included in the studies based on statistical analyses using different delay-intervals (6 weeks– 3 months). The patients were prospectively included in only one of the studies [21]. The existing observational studies have given conflicting results about delay and stage of rectal cancer. Most former studies contradict our result [6–17,23]. Holliday and Hardcastle [9], for example, examined 100 colon cancer patients and 100 rectal cancer patients and found no association between delay and stage comparing the average delay periods in the different Dukes’ stages— the median delay periods were not mentioned according to stage. Rectal cancer patients were analyzed combined with the colon cancer patients as in other studies finding no association between delay and stage for rectal cancer [6,8,10,11,14,15,17,23]. In accordance with our study, six studies [18–22,24] found an association between delay of treatment and stage of
345
rectal cancer. None of the studies included as many patients as in this study, and various statistical methods and choices of intervals were used. None of the studies showed the same statistical strength of a significant association between delay of treatment and stage of rectal cancer as we did. Symptoms of rectal cancer tend to be better defined than symptoms of colon cancer and thus easier to affirm [18,27,37–39]. The patients might be more likely to react to these symptoms than the vague symptoms more frequently experienced for colon cancer. The question as to why an association between delay and stage of cancer can be shown so convincingly for rectal cancer but not for colon cancer might be found by associating symptoms with the location of the cancer, as symptoms of colon cancer are known to be more vague than symptoms of rectal cancer. In conclusion, we found a strong association between delay of treatment and stage of the cancer at time of operation for rectal cancer, but not for colon cancer, which is important knowledge in aiming to improve survival. We should focus on early diagnosis for rectal cancer patients, diagnosing more of them when they are in Dukes’ stage A or B. Acknowledgements Contributors include: Aarhus University Hospital— Consultants: Søren Laurberg, Marianne Korsgaard MD, Professor Henrik Toft Sørensen, Chief Biostatistician Lars Pedersen, Department of Clinical Epidemiology, Esbjerg Hospital—Consultants: Karl Erik Juul Jensen, Peiman Poornoroozy, Vagn Juhl Jensen, Grena˚ Hospital—Consultant: Jan Lindholt, Grindsted Hospital—Consultants: Anders Larsen, Vagn Berg, Herning Hospital—Consultants: Mogens Rørbæk, Holstebro Hospital—Consultants: Mads Mark, Erik Skoubo Kristensen, Henrik Sloth, Lemvig Hospital—Consultant: Ivan Nørbæk, Odder Hospital— Consultant: Gunnar Madsen, Randers Hospital—Consultants: Knud Thygesen, Frank Svendsen Jensen, Ringkøbing Hospital—Consultant: Ebbe Fuglsang, Silkeborg Hospital—Consultants: Ole Brandsborg, Erling Østergaard. Sources of support were: Grants from Western Danish Research Forum (Vestdansk Forskningsforum); Danish Medical Research Council (Statens Sundhedsvidenskabelige Forskningsra˚d); Manufacturer Einar Willumsen’s Memorial Award (Fabrikant Einar Willumsens Mindelegat); Dagmar Marshall’s Fund (Dagmar Marshalls Fond); Medical Student Karsten Hansen’s Memorial Award (Stud. med. Karsten Hansens Mindelegat); and Else and Mogens Wedell-Wedellborg’s Fund (Else og Mogens WedellWedellborgs Fond).
References [1] Pisani P, Parkin DM, Ferlay J. stimates of the worldwide mortality from eighteen major cancers in 1985. Implications for prevention and projections of future burden. Int J Cancer 1993;55:891–903.
346
M. Korsgaard et al. / Cancer Detection and Prevention 30 (2006) 341–346
[2] Pisani P, Parkin DM, Bray F, Ferlay J. Estimates of the worldwide mortality from 25 cancers in 1990. Int J Cancer 1999;83:18–29. [3] Wingo PA, Cardinez CJ, Landis SH, Greenlee RT, Ries LA, Anderson RN, et al. Long-term trends in cancer mortality in the United States, 1930–1998. Cancer 2003;97:3133–275. [4] Statens Institut for Medicinsk TeknologivurderingKræft i tyktarm og endetarm. Diagnostik og screening. Medicinsk Teknologivurdering. Medicinsk Teknologivurdering; 2001. [5] Kronborg O. Screening for early colorectal cancer. World J Surg 2000;24:1069–74. [6] Ratcliffe R, Kiff RS, Hoare EM, Kingston RD, Walsh SH, Jeacock J. Early diagnosis in colorectal cancer still no benefit? Ann Chir 1989;43:570–4. [7] Graffner H, Olsson SA. Patient’s and doctor’s delay in carcinoma of the colon and rectum. J Surg Oncol 1986;31:188–90. [8] Pescatori M, Maria G, Beltrani B, Mattana C. Site emergency, and duration of symptoms in the prognosis of colorectal cancer. Dis Colon Rectum 1982;25:33–40. [9] Holliday HW, Hardcastle JD. Delay in diagnosis and treatment of symptomatic colorectal cancer. Lancet 1979;1:309–11. [10] Kyle SM, Isbister WH, Yeong ML. Presentation, duration of symptoms and staging of colorectal carcinoma. Aust N Z J Surg 1991;61: 137–40. [11] Irvin TT, Greaney MG. Duration of symptoms and prognosis of carcinoma of the colon and rectum. Surg Gynecol Obstet 1977; 144:883–6. [12] Devlin HB, Plant JA, Morris D. The significance of the symptoms of carcinoma of the rectum. Surg Gynecol Obstet 1973;137:399–402. [13] Goodman D, Irvin TT. Delay in the diagnosis and prognosis of carcinoma of the right colon. Br J Surg 1993;80:1327–9. [14] Jolly KD, Scott JP, MacKinnon MJ, Clarke AM. Diagnosis and survival in carcinoma of the large bowel. Aust N Z J Surg 1982; 52:12–6. [15] Khubchandani M. Relationship of symptom duration and survival in patients with carcinoma of the colon and rectum. Dis Colon Rectum 1985;28:585–7. [16] McDermott FT, Hughes ES, Pihl E, Milne BJ, Price AB. Prognosis in relation to symptom duration in colon cancer. Br J Surg 1981;68:846–9. [17] Gonzalez-Hermoso F, Perez-Palma J, Marchena-Gomez J, LorenzoRocha N, Medina-Arana V. Can early diagnosis of symptomatic colorectal cancer improve the prognosis? World J Surg 2004;28:716–20. [18] Arbman G, Nilsson E, Storgren-Fordell V, Sjodahl R. A short diagnostic delay is more important for rectal cancer than for colonic cancer. Eur J Surg 1996;162:899–904. [19] Pahlman L, Glimelius B, Enblad P. Clinical characteristics and their relation to surgical curability in adenocarcinoma of the rectum and rectosigmoid. A population-based study on 279 consecutive patients. Acta Chir Scand 1985;151:685–93. [20] Robinson E, Mohilever J, Zidan J, Sapir D. Colorectal cancer: incidence, delay in diagnosis and stage of disease. Eur J Cancer Clin Oncol 1986;22:157–61.
[21] Rowejones DC, Aylett SO. Delay in treatment in carcinoma of colon and rectum. Lancet 1965;2:973–6. [22] Langenbach MR, Schmidt J, Neumann J, Zirngibl H. Delay in treatment of colorectal cancer: multifactorial problem. World J Surg 2003;27:304–8. [23] Bharucha S, Hughes S, Kenyon V, Anderson ID, Carlson GL, Scott NA. Targets and elective colorectal cancer: outcome and symptom delay at surgical resection. Colorectal Dis 2005;7:169–71. [24] Young CJ, Sweeney JL, Hunter A. Implications of delayed diagnosis in colorectal cancer. Aust N Z J Surg 2000;70:635–8. [25] Hackett TP, Cassem NH, Raker JW. Patient delay in cancer. N Engl J Med 1973;289:14–20. [26] Goulston KJ, Cook I, Dent OF. How important is rectal bleeding in the diagnosis of bowel cancer and polyps? Lancet 1986;2:261–5. [27] MacDonald L, Freeling P. Bowels: beliefs and behaviour. Fam Pract 1986;3:80–4. [28] Kay L, Jorgensen T, Schultz-Larsen K. Abdominal pain in a 70year-old Danish population. An epidemiological study of the prevalence and importance of abdominal pain. J Clin Epidemiol 1992; 45:1377–82. [29] Selvachandran SN, Hodder RJ, Ballal MS, Jones P, Cade D. Prediction of colorectal cancer by a patient consultation questionnaire and scoring system: a prospective study. Lancet 2002;360:278–83. [30] Byles JE, Redman S, Hennrikus D, Sanson-Fisher RW, Dickinson J. Delay in consulting a medical practitioner about rectal bleeding. J Epidemiol Community Health 1992;46:241–4. [31] Curless R, French J, Williams GV, James OF. Comparison of gastrointestinal symptoms in colorectal carcinoma patients and community controls with respect to age. Gut 1994;35:1267–70. [32] Silman AJ, Mitchell P, Nicholls RJ, Macrae FA, Leicester RJ, Bartram CI, et al. Self-reported dark red bleeding as a marker comparable with occult blood testing in screening for large bowel neoplasms. Br J Surg 1983;70:721–4. [33] Goulston K, Chapuis P, Dent O, Bokey L. Significance of bowel symptoms. Med J Aust 1987;146:631–3. [34] Farrands PA, Hardcastle JD, Chamberlain J, Moss S. Factors affecting compliance with screening for colorectal cancer. Community Med 1984;6:12–9. [35] Chapuis PH, Goulston KJ, Dent OF, Tait AD. Predictive value of rectal bleeding in screening for rectal and sigmoid polyps. Br Med J (Clin Res Ed) 1985;290:1546–8. [36] Dent OF, Goulston KJ, Zubrzycki J, Chapuis PH. Bowel symptoms in an apparently well population. Dis Colon Rectum 1986;29:243–7. [37] MacArthur C, Smith A. Factors associated with speed of diagnosis, referral, and treatment in colorectal cancer. J Epidemiol Community Health 1984;38:122–6. [38] Wobbes T. Carcinoma of the colon and rectum in geriatric patients. Age Ageing 1985;14:321–6. [39] Majumdar SR, Fletcher RH, Evans AT. How does colorectal cancer present? Symptoms, duration, and clues to location. Am J Gastroenterol 1999;94:3039–45.