Munich Cancer Registry, Ludwig-Maximillians-University, Munich, Germany
* Correspondence to: Dr J. Engel, Tumorregister München, Institut für Med. Informationsverarbeitung, Biometrie und Epidemiologie, Ludwig-Maximilians-Universität München, Marchioninistraße 15, D-81377 München, Germany. Tel: +49-89-7095-4489; Fax: +49-89-7095-4753; Email: engel{at}ibe.med.uni-muenchen.de
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Patients and methods: Details of all cancer patients in the Munich region are recorded by the Munich Cancer Registry. Rectal cancer patients with an invasive primary tumor diagnosed between 1996 and 1998 were included in this analysis (n=936). Observed and relative survival are presented. Observed survival was also investigated with a Cox proportional hazards regression model.
Results: Median follow-up time of survivors was 5.7 years. Five-year relative survival for the whole sample was 62.2%. International Union Against Cancer (UICC) stage was the most important prognostic factor in the multivariate analysis. Compared with the 19921999 Surveillance Epidemiology and End Results (SEER) data (62.4%), relative survival for each disease stage and the whole sample were very similar. In comparison with other European registries, Munich patients had slightly higher survival rates per stage (for example, 5-year relative survival in UICC III was 58.3% in Munich, 54.6% in South East Netherlands, 33.3% in Modena and 47.4% in Cote d'Or); however, more patients in Munich were in higher disease stages with worse prognoses, indicating poorer early detection.
Conclusions: These results indicate that treatment of rectal cancer in Munich is good, but early detection could be improved. Cancer registries should publish their population-based stage data to ensure quality of care and provide regular feedback to health-care workers and decision makers. Comparisons between countries without stage data should be conducted cautiously.
Key words: cancer registry, epidemiology, rectal cancer, stage, survival
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
It is particularly important to compare population outcomes of routine care with clinical trial data, which are more frequently published. Randomized control trials (RCTs) may present impressive results, but they often assess the latest treatment (perhaps not yet available to all) in exclusive patient groups in specialized settings. Furthermore, treatment guidelines are often based on such trials, but without older patients or those with comorbid conditions being included the guidelines have limited application. Cancer registries, in contrast, can show which treatments are routinely prescribed and effective in all patient groups. Furthermore, while trials tend to give the impression that we are making progress in treating patients, epidemiological data may show that this is not always the case in reality [1, 4
, 5
].
The quality of health care in different countries is also frequently debated, somewhat unscientifically, in the popular media. This can then impact on public perception and health-care policy [6]. League tables of cancer survival rates, however, are uninformative if stage-specific results are not considered. A high survival rate may be due to good treatment or early diagnosis or both; only additional stage data can clarify this. Without stage data it is also difficult to argue for the effectiveness of cancer screening programs and the introduction of new treatments. Although cancer registries collect population data for cancer incidence and mortality, many do not have specific disease stage information. For example, out of 40 EUROCARE registries in the 1980s, only seven had stage data in more than 75% of cases [7
]; in another EUROCARE publication stage data from 11 registries are available [8
].
This paper presents a detailed epidemiological view of rectal cancer from a population-based sample of patients diagnosed between 1996 and 1998 in Munich. This paper will also compare the Munich Cancer Registry (MCR) data with other cancer registry and trial data, where possible. This paper aims to demonstrate the importance and practical application of population-based data and the need for treatment and stage details for international comparisons.
![]() |
Methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Data quality and register coverage
The cooperation of all pathology laboratories provides the basis for incidence. Documentation from different sources leads to a high completeness of data. In comparison with the German Saarland Cancer Registry, where for 1998 age standardized (world) incidence rates (per 100 000 population) were 18.6 (male) and 11.3 (female), the Munich rates are comparable with 16.8 (male) and 8.9 (female). Completeness concerning progression-events is different. Estimated by relative survival there is probably underdocumentation of metastases. Local recurrences, however, are almost complete following surgical removal and documentation of pathological reports. At present, follow-up, in terms of life status, is available in 95% of the catchment area for patients registered since 1978, and for patients registered in the field study almost 99%. The Bavarian Cancer Registration Law came into force at the beginning of 1998, allowing the MCR to legally process all death certificates from the catchment area. A death certificate only (DCO) rate, therefore, for the years 19961998 would not be meaningful. The proportion of microscopically verified cases of rectal cancer is over 90%.
Patient sample
Tumors found 16 cm or less from the anal margin were recorded as rectal cancer. A total of 1038 patients (living and treated in the Munich region) were diagnosed with rectal cancer between 1996 and 1998. Patients with an in situ carcinoma (n=5) or evidence of another previous or synchronous cancer (n=89; 8.6%) were excluded from this analysis. The analysis was therefore performed on the 936 patients with an invasive primary rectal cancer tumor. Survival status was assessed in June 2003, allowing at least 5 years of follow-up.
Prognostic variables
The following prognostic data were available: stage [International Union Against Cancer (UICC) and TNM], disease spread (local, regional and distant), grading, age, gender, tumor location, local recurrence, residual tumor status, adjuvant therapy and operative technique. Table 1 shows the comparison between UICC or American Joint Committee on Cancer (AJCC) stage, pT category, and the now redundant Duke's stage [10]. The Surveillance Epidemiology and End Results (SEER) data are divided into local, regional and distant spread, and this system is also illustrated in Table 1, in comparison with the other stage definitions. Age was divided into <65 and
65 years, to allow comparison with other registry data, which often use this age categorization. There were four tumor location classifications: <4 cm, 4 to <8 cm, 8 to <12 cm and >12 cm (to 16 cm). Although published studies speak of local recurrence, the definitions given indicate that, strictly speaking, they mean locoregional recurrence. For continuity, the term local recurrence will also be used throughout this paper but is defined as any recurrence of rectal cancer within the pelvis. Residual tumor (R) status was R0 (no residual tumor) compared with RX (presence of residual tumor cannot be assessed), R1(microscopic residual tumor) and R2 (macroscopic residual tumor) combined. Five adjuvant therapy possibilities were recorded: no therapy, preoperative neo-adjuvant therapy, radiation therapy only, chemotherapy only, and combined radiation and chemotherapy. Patients who received preoperative therapy were not accorded a traditional postoperative UICC or TNM stage. These patients were therefore excluded from all stage analyses. Patients with disease progression who received preoperative therapy were coded as M1 or UICC IV and included in the analyses. The operative technique, total mesorectal excision (TME), was recorded in the original operation reports and extracted from these. This may have resulted in under documentation of TME, as clinicians appear to have been reluctant to use this term, despite removing the mesorectum, because they had been performing such an operation before TME was reported.
|
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
|
|
|
|
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Figure 5 shows a summary of rectal cancer 5-year relative survival rates from five population-based reports over the last 25 years [13
, 7
, 14
]. Of the two studies that showed no significant improvement since 1990, the US study also reported no improvement in disease stage during this time, suggesting no change in screening practice [1
]. The improvement seen in the French study was attributed to an increase in surgical and adjuvant radiotherapy treatment and an increase in the percentage of Dukes' A patients, with better prognoses [2
]. In the south-east Netherlands, relative survival improved from 1980 to 2000, but there was no stage shift in this study since good access to sigmoidoscopy has been available in the south-east Netherlands since the early 1980s. Increased survival in this group was attributed to improved surgical treatment with the TME technique and increased use of preoperative radiotherapy [3
]. The 62% 5-year relative survival rate in Munich from 1996 to 1998 is comparable with these international studies. This overview of the literature on long-term studies shows that there has been some improvement due to screening and treatment in rectal cancer, but variations in survival between countries exist.
|
To conduct an up to date comparison, we have collected data from US and European registries with stage data during a similar time period to the Munich study. Compared with the south-east Netherlands, Modena and Côte d'Or, 5-year relative survival rates were slightly higher in Munich for each disease stage. However, 52% of Munich patients were in UICC stage III or IV. This indicates that treatment in Munich is good but, in comparison with other regions in Europe, our rectal cancer early detection rates are poor. Differences in treatment between The Netherlands and Germany may explain the difference in relative survival rates. For example, adjuvant chemotherapy is not recommended in The Netherlands [3]. In Germany, combined chemo- and radiotherapy is recommended in UICC II and III patients, and in this study this treatment provided a significant survival benefit for some patients. Another difference is that the Netherlands study indicated greater use of preoperative radiotherapy (30% versus 9%) [3
]. Preoperative radiotherapy was recommended in this region of The Netherlands; it was not strongly recommended in Munich at this time. The greatest differences in 5-year relative survival, however, were in UICC I and II patients (8%), where adjuvant therapy is not used as much. The difference in stage distribution is likely to be due to better access to screening in south-east Netherlands [3
]. Age differences may also play a role. Patients from The Netherlands had different 5-year relative survival rates depending on age: <60 years, 69%; 6074 years, 63%; and 75 + years, 49% [3
]. For the same age groups our 5-year relative survival rates were 64%, 63% and 63%. Moreover, only 25% of Munich patients were in the oldest age group compared with 36% in Eindhoven.
Our relative survival rates were more similar to the US registries in the SEER report across all stages [12]. It is not known whether patients undergoing neo-adjuvant therapy were included in the SEER data. If so, this may have inflated the SEER results. In Munich there were 8% fewer patients with regional spread, but 4% more M1 patients. The decrease in the number of SEER patients with distant spread (19% in 1989 [15
] to 15% in 1999 [12
]) would indicate improved early detection in the USA. Indeed, 48.4% (male) and 42.5% (female) of over 65 years olds had a blood test within the last year or colorectal endoscopy within the last 5 years [16
]. One registry, in California, not included in the 19752000 SEER report, however, had lower incidence of UICC III tumors than Munich [17
]. The 2-year unadjusted observed survival rates for a comparable sample in Munich were, however, higher than those reported in California. Since the 1989 comparison between Europe and the USA [15
], when European relative survival rates were almost 20% lower, relative survival rates have not improved significantly in the USA [1
]. In contrast, nowadays, the relative survival rates in Munich (62.2) and The Netherlands (61) [3
] are comparable with the USA (62.4) [12
]. Although there was less room for improvement in America, the European nations should be pleased to be making progress and catching up, so to speak.
Our analysis, as expected, showed that UICC stage was the strongest prognostic factor, even when controlling for age, operation method and treatment modalities. This indicates that more early detection of rectal cancer could improve survival rates. Fecal occult blood testing and sigmoidoscopy, for example, have been shown to reduce colorectal cancer mortality and most importantly incidence [1, 16
, 18
]. Currently in Germany, fecal occult blood testing is paid for yearly by health insurers for over 50 year olds and colonoscopy is recommended at 55 and 65 years. No program exists, however, to ensure that the population is systematically offered these tests or invited to attend screening. There is a vast literature that indicates which groups should be targeted and how attendance can be improved. If screening is to increase, however, diagnostic facilities and funding are required [18
].
In the Munich population, operative technique and adjuvant therapy treatment were also significant predictors of survival. Such details are often not available in population-based studies. The 5-year relative survival for patients in the Munich population treated with TME (76.2%) was comparable with the survival rate (80%) achieved by Heald himself, the founder of TME [19]. The 5-year relative survival benefits of combined therapy were also comparable with a national US sample [20
] and the Munich data for 5-year observed survival compared well with data from RCTs investigating adjuvant therapy treatment [13
]. The recurrence rates in these trials were high, however, indicating that good surgical control may also have contributed to the high survival rates in Munich. Nonetheless, this would suggest that when treatments are implemented outside of trials they achieve similar survival benefits. For improvements in survival to be seen, however, more patients would have to receive the recommended therapy. In this sample only one-third of patients received the recommended combined adjuvant therapy and less than one-quarter of surgeons reported using the TME technique.
While it is useful for doctors to know how they are performing in comparison with other countries and hospitals, it is equally important for cancer patients to know their prognosis. With improved treatment and earlier diagnosis, survival rates in some cancers are now relatively high. Some stages have relative survival rates comparable with expected survival rates in the normal population, for example. For rectal cancer patients in UICC stage I, 97% were likely to live for >5 years after their diagnosis. Patients should hear this good news when applicable, especially since cancer is still considered a death sentence. Furthermore, to enable informed treatment decisions, patients must also be aware of their survival chances with different therapy alternatives. Patients should at least be informed and given the opportunity to discuss treatment options, even if ultimately they prefer the physician to choose or make a recommendation. In response to this trend, many useful decision-making models and tools have been developed to help the doctor and patient [21]. Although these tools may be valuable, without accurate population-based survival data to compare in the models, their use is limited. Population data, provided by cancer registries, is therefore extremely important for both doctors and patients.
In conclusion, we can say that in comparison with Europe, the survival rates in Munich were high but the early detection rates poor. In comparison with RCTs, our treatment effects were good, but the application of recommended treatment in the population could be improved. To help to improve routine care, the MCR must provide regular feedback to doctors on their performance, for example through the Internet. Of greatest importance, however, is the implementation of a systematic colorectal cancer screening program, drawing on the vast literature investigating how to increase screening acceptance and uptake. In addition, doctors can ease the burden and fear of cancer by giving patients detailed information about their chances of survival. Finally, this paper has shown that it is unwise to compare data across countries without stage data: the media and politicians should take note. Even with stage data, comparisons are difficult because of population variations, and different analysis and treatment strategies. Despite limitations of some missing data for prognostic factors and of missing covariates for some process quality indicators, e.g. for exact adjuvant treatment, it is useful to communicate distributions of prognostic factors and survival, thus describing whether results of innovative clinical trials have transferred to population level care [22].
![]() |
Acknowledgements |
---|
Received for publication July 26, 2004. Revision received November 19, 2004. Accepted for publication November 23, 2004.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2. Finn-Faivre C, Maurel J, Benhamiche AM et al. Evidence of improving survival of patients with rectal cancer in France: a population based study. Gut 1999; 44: 377381.
3. Martijn H, Voogd AC, van de Poll-Franse LV et al. Improved survival for patients with rectal cancer since 1980: a population-based study. Eur J Cancer 2003; 39: 20732079.[CrossRef][ISI][Medline]
4. Engel J, Eckel R, Schubert-Fritschle G et al. Moderate progress for ovarian cancer in the last 20 years: prolongation of survival, but no improvement in the cure rate. Eur J Cancer 2002; 38: 24352445.[CrossRef][ISI][Medline]
5. Peppercorn JM, Weeks JC, Cook EF, Joffe S. Comparison of outcomes in cancer patients treated within and outside clinical trials: conceptual framework and structured review. Lancet 2004; 363: 263270.[CrossRef][ISI][Medline]
6. Harrabin R, Coote A, Allen J. Health in the News: Risk, Reporting and Media Influence. London: The Kings Fund Report 2003.
7. Gatta G, Faivre J, Capocaccia R, Ponz de Leon M. Survival of colorectal cancer patients in Europe during the period 19781989. Eur J Cancer 1998; 34: 21762183.[CrossRef][ISI]
8. Gatta G, Capocaccia R, Sant M et al. Understanding variations in survival for colorectal cancer in Europe: a EUROCARE high resolution study. Gut 2000; 47: 533538.
9. World Health Organization. International Classification of Diseases for Oncology (ICD-O), 2nd edition. Geneva: World Health Organization 1990.
10. American Joint Committee on Cancer. Cancer Staging HandbookTNM Classification of Malignant Tumors, 6th edition. New York: Springer 2002.
11. Hakulinen T. Cancer survival corrected for heterogeneity in patient withdrawal. Biometrics 1982; 38: 933942.[ISI][Medline]
12. Ries LAG, Eisner MP, Kosary CL et al. SEER Cancer Statistics Review, 19752000. Bethesda, MD: National Cancer Institute 2003. Available on-line: http://seer.cancer.gov/csr/1975_2000 (19 January, 2003, date last accessed).
13. Rödel C, Sauer R. Perioperative radiotherapy and concurrent radiochemotherapy in rectal cancer. Semin Surg Oncol 2001; 20: 312.[CrossRef][ISI][Medline]
14. Brenner H, Hakulinen T. Long-term cancer patient survival achieved by the end of the 20th century: most up-to-date estimates from the nationwide Finnish cancer registry. Br J Cancer 2001; 85: 367371.[CrossRef][ISI][Medline]
15. Gatta G, Ciccolallo L, Capocaccia R et al. Differences in colorectal cancer survival between European and US populations: the importance of sub-site and morphology. Eur J Cancer 2003; 39: 22142222.[CrossRef][ISI][Medline]
16. Swan J, Breen N, Coates RJ et al. Progress in cancer screening practices in the United States: results from the 2000 National Health Interview Survey. Cancer 2003; 97: 15281540.[CrossRef][ISI][Medline]
17. Hodgson DC, Zhang W, Zaslavsky AM et al. Relation of hospital volume to colostomy rates and survival for patients with rectal cancer. J Natl Cancer Inst 2003; 95: 708716.
18. Rhodes JM. Colorectal cancer screening in the UK: Joint Position Statement by the British Society of Gastroenterology. The Royal College of Physicians, and The Association of Coloproctology of Great Britain and Ireland. Gut 2000; 46: 746748.
19. Heald RJ, Moran BJ, Ryall RD et al. Rectal cancer: the Basingstoke experience of total mesorectal excision, 19781997. Arch Surg 1998; 133: 894899.
20. Jessup JM, Stewart AK, Menck HR. The National Cancer Data Base report on patterns of care for adenocarcinoma of the rectum, 198595. Cancer 1998; 83: 24082418.[CrossRef][ISI][Medline]
21. O'Connor AM, Legare F, Stacey D. Risk communication in practice: the contribution of decision aids. BMJ 2003; 327: 736740.
22. Burton A, Altman DG. Missing covariate data within cancer prognostic studies: a review of current reporting and proposed guidelines. Br J Cancer 2004; 91: 48.[CrossRef][ISI][Medline]
|