Aspects of survival from colorectal cancer in Denmark
Lene Hjerrild Iversen
Abstract:This thesis has reported survival among Danish colorectal cancer patients over the past decades and it has explored different aspects of the inferior short-term and long-term survival of Danish patients in relation to (i) patient factors: old age and comorbidity; (ii) disease factors: prognostic factors for early death after emergency surgery; (iii) diagnostic factors: impact of diagnostic delay; (iv) treatment factors: seasonal variation in postoperative mortality and the benefit of a new approach for management of obstructive cancer; and (v) structural factors: hospital volume and surgeon characteristics. Short-term survival. For colonic cancer, the 30-day relative survival rose from 86% in 1977-1982 to 90% in 1995-1999, and for rectal cancer it rose from 90% to 94% (I). Data from regional hospital discharge registries show that the overall 30-day mortality rates of colonic and rectal cancer remained stable at about 11% and 4-5%, respectively, during 1985-2004 (II). Stratifying for urgency of surgery, but not for tumour site, the 30-day mortality rate was 6.2% after elective surgery and 22.1% after emergency surgery in the period 2001-2008 according to DCCG data (17). Nevertheless, the 30-day mortality was about twice as high in Denmark as in Norway, Sweden and Scotland, even if the data for these countries are older than the Danish data. Mortality rates after palliative surgery are three times higher than the rates following curative surgery (115). The stage distribution at the time of diagnosis is more unfavourable in Denmark than in the other Nordic countries (114). Comparison of survival among countries is, however, encumbered by several methodological issues related to completeness, and data quality of cancer registries, among others, biases the survival estimates. Like most western populations, the Danish population is ageing and the proportion of elderly colorectal cancer patients aged >75 years has therefore risen from 37% in 1977-1982 to 42% in 1995-1999 (III). Disparity in cancer treatment between elderly and younger patients exists on a number of counts, e.g., the former's curative resection rate is lower, their emergency presentation rate is higher and they are moreover more likely to present with later-stage disease than are younger patients. However, in Denmark the curative resection rate among elderly patients aged >75 years rose from 36% in 1977-1982 to 49% in 1995-1999 (III). This trend was paralleled by an increase in 30-day and 6-month relative survival. Patients aged >70 years have a 30-day mortality rate of 13.1%, but their younger counterparts' mortality rate is only 3.5% (17). A mortality rate at least two to three times higher among the elderly than among younger patients has been reported repeatedly in various populations (95,96,116,127,128,130,131,135,136,144,145). In Denmark there is an inverse relationship between the comorbidity level and the resection rate in colorectal cancer. In the period 1995-2006, surgical treatment of patients with colonic cancer and severe comorbidity became progressively more aggressive, whereas surgical treatment of patients with rectal cancer apparently became more cautious or differentiated (VI). Nevertheless, the overall 30-day mortality rate after resectional surgery remained stable at about 8% in colonic cancer and at about 6% in rectal cancer. Almost every fourth patient had severe comorbidity as determined by an ASA of III or more and their 30-day mortality rate was at least 18% in 2001-2008. Any reduction in their short-term mortality will therefore have a substantial impact on the overall mortality rate. Despite the impact of comorbidity on postoperative mortality, the distinct seasonal variation seen in mortality from cardiovascular and respiratory diseases, with excess mortality in the winter months, has not been observed in postoperative mortality from colorectal cancer (VII). Postoperative mortality from colonic cancer was non-significantly higher in July than in other months of the year (VII). Evidence reveals a volume-outcome relationship regarding postoperative mortality in colonic cancer (IV) and the most recent literature suggests that it probably also is so in rectal cancer. However, volume may be a surrogate marker or proxy for other important structural factors such as quality and capacity of intensive care units, the availability of other clinical services like cardiac care units, multiple medical specialties, multidisciplinary infrastructure and nurse staffing, etc. Postoperative mortality after emergency surgery for colonic cancer was as high as 22% in 2001-2005 and mortality was significantly associated with the postoperative course. Patients developing medical complications had a mortality rate of 57.8%. Independent risk factors for death within 30 days after surgery were age ≥ 71 years, male gender, ASA grade ≥ III, palliative outcome, free or iatrogenic tumour perforation, splenectomy, intraoperative surgical adverse events and postoperative medical complications (VIII). SEMS placement performed on the indication acute bowel obstruction in patients with potentially curable disease can be accomplished with high technical and clinical success rates. The perforation rate, however, may reach 12%. Even so, the mortality rate within 30 days after a SEMS attempt and later surgery may, irrespective of its timing, by very low (3%) relative to the mortality seen after emergency surgery (IX). Long-term survival. The 5-year relative survival improved by 9% for both colonic and rectal cancer from 1977-1982 to 1995-1999 (I). Further improvement has been observed and in 2004-2006, the 5-year relative survival from colonic cancer was 52% (95% CI 51-54) for men and 57% (95% CI 55-58) for women. For rectal cancer the corresponding percentages were 55% (95% CI 53-57) and 57% (95% CI 55-59) (202). Overall, from 1977 until 2006, 1-year and 5-year survival increased almost 0.5-1% annually. Long-term survival has improved more in rectal cancer than in colonic cancer and survival from rectal cancer surpassed that of colonic cancer in the 2000s (202,204). Elderly patients aged >75 years experienced a marked 13-16% increase in relative survival from 1977-1982 to 1995-1999, i.e., a period during which the rate of curative surgery increased pronouncedly among the elderly (III). The survival improvement among their younger counterparts in that period only reached 7%. Mortality from colorectal cancer was only excessive in the elderly during the first two years after surgery. In 1995-2006, about 30-43% of colorectal cancer patients had moderate and severe comorbidity as determined by a Charlson Comorbidity score of 1-2 and 3+, respectively. These comorbid patients had a long-term survival inferior to that of patients with no comorbidity. In colonic cancer, the 5-year survival in 1998-2000 was 43% in patients with no comorbidity and only 20% in patients with severe comorbidity. Comorbidity had an even stronger impact in rectal cancer (VI). Evidence repeatedly demonstrates a volume-outcome effect on long-term survival from colonic and rectal cancer with improved survival being significantly associated with increasing hospital caseload and surgeon's education/specialty (V). In addition, the most recent evidence reveals that surgeon caseload may have a stronger impact on long-term survival than hospital volume which reflects the complexity in the interaction between hospital caseload and surgeon caseload. A total therapeutic delay ≥ 60 day has been shown to have a negative impact on the long-term survival from rectal cancer, but not from colon cancer, given that stage is an intermediate step in the causal pathway between delay and survival (X). Neither provider delay ≥ 60 days, nor hospital delays ≥ 30 days or ≥ 60 days had any prognostic impact on long-term survival from colorectal cancer. Emergency surgery for colonic cancer is associated with an inferior long-term survival. The 5-year survival after acute curative surgery in Denmark is 39% (16). However, the use of SEMS as bridge to elective curative surgery makes it possible to achieve 3-year survival rates similar to those of 75% seen after elective curative surgery for colonic cancer (IX) - despite an unexpectedly high perforation rate.