Treatment of local-regional prostate cancer detected by PSA screening: benefits and harms according to prognostic factors.

BACKGROUND: Men with screen-detected prostate cancer can choose to undergo immediate curative treatment or enter into an expectant management programme. We quantified how the benefits and harms of immediate treatment vary according to the prognostic factors of clinical T-stage, Gleason score, and patient age. METHODS: A microsimulation model based on European Randomized Study of Screening for Prostate Cancer data was used to predict the benefits and harms of immediate treatment versus delayed treatment of local-regional prostate cancer in men aged 55-74 years. Benefits included life-years gained and reduced probability of death from prostate cancer. Harms included lead time and probability of overdiagnosis. RESULTS: The ratio of mean lead time to mean life-years gained ranged from 1.8 to 31.2, and the additional number of treatments required per prostate cancer death prevented ranged from 0.3 to 11.6 across the different prognostic groups. Both harm-benefit ratios were lowest, most favourable, for men aged 55-59 years and diagnosed with moderate-risk prostate cancer. Ratios were high for men aged 70-74 years regardless of clinical T-stage and Gleason score. CONCLUSION: Men aged 55-59 years with moderate-risk prostate cancer are predicted to derive greatest benefit from immediate curative treatment. Immediate treatment is least favourable for men aged 70-74 years with either low-risk or high-risk prostate cancer.

To be screened or not to be screened? Modeling the consequences of PSA screening for the individual.

BACKGROUND: Screening with prostate-specific antigen (PSA) can reduce prostate cancer mortality, but may advance diagnosis and treatment in time and lead to overdetection and overtreatment. We estimated benefits and adverse effects of PSA screening for individuals who are deciding whether or not to be screened. METHODS: Using a microsimulation model, we estimated lifetime probabilities of prostate cancer diagnosis and death, overall life expectancy and expected time to diagnosis, both with and without screening. We calculated anticipated loss in quality of life due to prostate cancer diagnosis and treatment that would be acceptable to decide in favour of screening. RESULTS: Men who were screened had a gain in life expectancy of 0.08 years but their expected time to diagnosis decreased by 1.53 life-years. Of the screened men, 0.99% gained on average 8.08 life-years and for 17.43% expected time to diagnosis decreased by 8.78 life-years. These figures imply that the anticipated loss in quality of life owing to diagnosis and treatment should not exceed 4.8%, for screening to have a positive effect on quality-adjusted life expectancy. CONCLUSION: The decision to be screened should depend on personal preferences. The negative impact of screening might be reduced by screening men who are more willing to accept the side effects from treatment.

Prediction of higher mortality reduction for the UK Breast Screening Frequency Trial: a model-based approach on screening intervals.

BACKGROUND: The optimal interval between two consecutive mammograms is uncertain. The UK Frequency Trial did not show a significant difference in breast cancer mortality between screening every year (study group) and screening every 3 years (control group). In this study, the trial is simulated in order to gain insight into the results of the trial and to predict the effect of different screening intervals on breast cancer mortality. METHODS: UK incidence, life tables and information from the trial were used in the microsimulation model MISCAN-Fadia to simulate the trial and predict the number of breast cancer deaths in each group. To be able to replicate the trial, a relatively low sensitivity had to be assumed. RESULTS: The model simulated a larger difference in tumour size distribution between the two groups than observed and a relative risk (RR) of 0.83 of dying from breast cancer in the study group compared with the control group. The predicted RR is lower than that reported from the trial (RR 0.93), but within its 95% confidence interval (0.63-1.37). CONCLUSION: The present study suggests that there is benefit of shortening the screening interval, although the benefit is probably not large enough to start annual screening.

Population-based mammography screening below age 50: balancing radiation-induced vs prevented breast cancer deaths.

INTRODUCTION: Exposure to ionizing radiation at mammography screening may cause breast cancer. Because the radiation risk increases with lower exposure age, advancing the lower age limit may affect the balance between screening benefits and risks. The present study explores the benefit-risk ratio of screening before age 50. METHODS: The benefits of biennial mammography screening, starting at various ages between 40 and 50, and continuing up to age 74 were examined using micro-simulation. In contrast with previous studies that commonly used excess relative risk models, we assessed the radiation risks using the latest BEIR-VII excess absolute rate exposure-risk model. RESULTS: The estimated radiation risk is lower than previously assessed. At a mean glandular dose of 1.3 mGy per view that was recently measured in the Netherlands, biennial mammography screening between age 50 and 74 was predicted to induce 1.6 breast cancer deaths per 100,000 women aged 0-100 (range 1.3-6.3 extra deaths at a glandular dose of 1-5 mGy per view), against 1121 avoided deaths in this population. Advancing the lower age limit for screening to include women aged 40-74 was predicted to induce 3.7 breast cancer deaths per 100,000 women aged 0-100 (range 2.9-14.4) at biennial screening, but would also prevent 1302 deaths. CONCLUSION: The benefits of mammography screening between age 40 and 74 were predicted to outweigh the radiation risks.

Overdetection, overtreatment and costs in prostate-specific antigen screening for prostate cancer.

BACKGROUND: Prostate cancer screening with prostate-specific antigen (PSA) has shown to reduce prostate cancer mortality in the European Randomised study of Screening for Prostate Cancer (ERSPC) trial. Overdetection and overtreatment are substantial unfavourable side effects with consequent healthcare costs. In this study the effects of introducing widespread PSA screening is evaluated. METHODS: The MISCAN model was used to simulate prostate cancer growth and detection in a simulated cohort of 100,000 men (European standard population) over 25 years. PSA screening from age 55 to 70 or 75, with 1, 2 and 4-year-intervals is simulated. Number of diagnoses, PSA tests, biopsies, treatments, deaths and corresponding costs for 100,000 men and for United Kingdom and United States are compared. RESULTS: Without screening 2378 men per 100,000 were predicted to be diagnosed with prostate cancer compared with 4956 men after screening at 4-year intervals. By introducing screening, the costs would increase with 100% to 60,695,000 euro. Overdetection is related to 39% of total costs (23,669,000 euro). Screening until age 75 is relatively most expensive because of the costs of overtreatment. CONCLUSION: Introduction of PSA screening will increase total healthcare costs for prostate cancer substantially, of which the actual screening costs will be a small part.

Cost-effectiveness of different reading and referral strategies in mammography screening in the Netherlands.

In mammography screening with double reading, different strategies can be used when the readers give discordant recommendations for referral. We investigated whether the results of the Dutch breast cancer screening programme can be optimised by replacing the standard referral strategy by consensus. Twenty-six screening radiologists independently and blinded to outcome read a test set consisting of previous screening mammograms of 250 cases (screen-detected and interval cancers) and 250 controls. Their referral recommendations were paired and, in case of discrepancy, re-read according to three referral strategies: (1) decision by one of the readers; (2) arbitration by a third reader; (3) referral if both readers agree (consensus). Data allowed studying other referral strategies, including referral if any reader suggests, as well. Double reading with referral if any reader suggests resulted in a 1.03 times higher sensitivity (76.6%) and a 1.31 times higher referral rate (1.26%) than double reading with consensus. To estimate the cost-effectiveness, the outcomes were used in a microsimulation model. Even if double reading with referral if any reader suggests results in four times as high referral rates and an accompanying increase of biopsies or other invasive procedures, the cost-effectiveness of 4,190 Euros per life-year gained may well be in the range of acceptable cost-effectiveness for Dutch health care programmes.

Trends in the usage of adjuvant systemic therapy for breast cancer in the Netherlands and its effect on mortality.

Adjuvant systemic therapy was introduced in the Netherlands as a breast cancer treatment in the early 1980s. In this paper, we describe the trends in the usage of adjuvant systemic treatment in the period 1975-1997 in the Netherlands. The main aim of our study was to assess the effects of adjuvant tamoxifen and polychemotherapy on breast cancer mortality, compared to the effects of the mammography screening programme. The computer simulation model MIcrosimulation SCreening ANalysis, which simulates demography, natural history of breast cancer and screening effects, was used to estimate the effects. Use of adjuvant therapy increased over time, but since 1990 it remained rather stable. Nowadays, adjuvant therapy is given to 88% of node-positive patients aged 50-69 years, while less than 10% of node-negative patients receive any kind of adjuvant treatment. Adjuvant treatment is given independent of the mode of detection (adjusted by nodal status and size). We predict that the reduction in breast cancer mortality due to adjuvant therapy is 7% in women aged 55-74 years, while the reduction due to screening, which was first implemented in women aged 50-69 years in 1990-97, will be 28-30% in 2007. In conclusion, although adjuvant systemic therapy can reduce breast cancer mortality rates, it is anticipated to be less than the mortality reduction caused by mammography screening.

MISCAN: estimating lead-time and over-detection by simulation.

OBJECTIVE: To estimate the mean lead-time and rate of over-detection associated with screening for prostate cancer with prostate-specific antigen. METHODS: Simulation models, fitted to the results of the Rotterdam section of the European Randomized Study of Screening for Prostate Cancer, were used to predict the mean lead-time and over-detection rate in population-based screening programmes. RESULTS: The mean lead-time is estimated to be 11-12 years and over-detection to occur in half the cases found by population screening. The estimates are compared with published estimates. CONCLUSIONS: The effect of lead-time and over-detection on the balance of positive and negative consequences of screening cannot be neglected.

Breeding biology of muscovy duck (Cairina moschata) under natural incubation: the use of the weibull function and a beta-binomial model to predict nest hatchability

In this paper we tested the Weibull function and beta-binomial distribution to analyse and predict nest hatchability, using empirical data on hatchability in Muscovy duck (Cairina moschata) eggs under natural incubation (932 successfully incubated nests and 11 822 eggs). The estimated parameters of the Weibull function and beta-binomial model were compared with the logistic regression analysis. The maximum likelihood estimation of the parameters was used to quantify simultaneously the influence of the nesting behaviour and the duration of the reproduction cycle on hatchability. The estimated parameters showed that the hatchability was not affected in natural dump nests, but in artificial dump nests and in nests with non-term eggs the hatchability was reduced by 10 and 25%, respectively. Similar results were obtained using logistic regression. Both models provided a satisfactory description of the observed data set, but the beta-binomial model proved to have more parameters with practical and biological meaningful interpretations, because this model is able to quantify and incorporate the unexplained variation in a single parameter theta (which is a variance measure). Copyright 1999 Academic Press.