Screening controversies
Department of Public Health Sciences, University of Toronto, Toronto, Canada
For breast cancer, several trials have shown efficacy of mammography screening in women age 50-69. For women age 40-49 the benefit seen in some trials after prolonged follow-up may be largely due to screening over the age of 50. With improving therapy, it is unclear whether modern programmes can deliver the benefit suggested from the trials. A trial among women age 50-59 in Canada found no benefit from mammography over and above skilled clinical breast examinations (CBE); the detection by mammography of impalpable cancers had no impact upon breast cancer mortality. Equivalent benefit may thus be achieved by regular CBE, using carefully trained examiners. The International Agency for Research on Cancer has endorsed trials of CBE. For cervix cancer, although there is evidence to support cytology screening for women age 25-60 every 3 –5 years, there is a tendency to screen younger women and all women too frequently. There is no good evidence that either HPV testing, nor visual inspection, will be superior. For colon cancer, trials support screening by the fecal occult blood test every 1-2 years, yet the level of efficacy (of the order of 20% reduction in mortality) is unlikely to be achieved in populations because of poor compliance. Endoscopy has not yet been established. For prostate cancer, many advocate PSA testing, but its efficacy in reducing mortality has not been demonstrated. However, widespread application of PSAs has affected prostate cancer incidence, without a dramatic effect on mortality. With improving prostate cancer therapy, even short term prolongation of life will reduce mortality, because of competing causes of death in those affected by the disease. For lung cancer, the trials conducted 20 or more years ago of chest x-ray and cytology screening were negative. However, a small benefit was not excluded. A large trial in the US is therefore re-evaluating chest X-ray screening. The major controversy for lung cancer now relates to whether screening using low dose (helical) CT will prove effective. Advocates of a single arm approach to evaluation ignore the fact that over-diagnosis of lung cancer would confound a historical comparison. Fortunately a large trial in the US is nearing its planned accrual, and trials in Europe are being initiated. For maximal effectiveness in the population, good organization of screening is essential. Further, compliance with the program must be high. However the contribution of screening relative to other approaches to cancer control is probably no more than a 3% decrease in mortality from cancer, so that decisions resulting in the initiation of screening programs must not lead to less support for other potentially more effective cancer control approaches, such as dietary modification for colon cancer prevention.
Paper presented at the International Symposium on Predictive Oncology and Intervention Strategies; Nice, France; February 7 - 10, 2004; in plenary session 803 (Risk & assessment).