Comparative Effectiveness of Alternative Prostate-Specific Antigen–Based Prostate Cancer Screening Strategies: Model Estimates of Potential Benefits and Harms

Fred Hutchinson Cancer Research Center prostate cancer incidence model: underlying PSA growth before and after onset of tumors with a Gleason score of 2 through 7 and 8 through 10.

The dashed and dotted lines are PSA trajectories by age for the 2 categories of Gleason score. Shaded bands around those lines illustrate between-person variability in PSA values based on interquartile ranges. The jagged line illustrates an example PSA trajectory for a man who develops a tumor with a Gleason score of 2 through 7. In this example, PSA exceeds the threshold for biopsy referral on the fifth test of a schematic screening strategy. PSA = prostate-specific antigen.

Fred Hutchinson Cancer Research Center prostate cancer incidence model: healthy, preclinical, clinical, prostate cancer mortality, and other-cause mortality states in the absence of screening.

Effect of screening on prostate cancer survival based on the stage-shift model and a scheme for considering more modest effects in a sensitivity analysis.

The effect of early detection on prostate cancer survival is assumed to follow a stage-shift model; that is, when nonoverdiagnosed patients are shifted from distant to local–regional stage at diagnosis, they receive a corresponding survival improvement. This figure represents a nonoverdiagnosed patient whose prostate cancer would have been detected in distant stage in the absence of PSA screening. His prostate cancer survival in the absence of PSA screening is represented by the curve labeled “no PSA.” In the presence of PSA, his prostate cancer is detected in local–regional stage, and his prostate cancer survival follows a more favorable distribution, namely that for local–regional stage cases. This prostate cancer survival is represented by the curve labeled “PSA.” The new survival begins at his original date of clinical diagnosis because, as a nonoverdiagnosed patient, by definition he cannot die during his lead time. In a sensitivity analysis, we considered a more modest effect of screening on prostate cancer survival. To implement this, the age at prostate cancer death consists of a weighted average of the age at prostate cancer death in the absence of screening and the age at prostate cancer death in the presence of screening: [Death (actual)] = w × [Death (no PSA)] + (1 − w) × [Death (PSA)], where the weight (w = e^−αλ) depends on lead time λ = [Diagnosis (no PSA)] − [Diagnosis (PSA)] and on a variable reflecting the effect of screening on survival α. Consequently, when α is small, the weight is approximately 1, and the actual age at prostate cancer death is the same as that projected in the absence of screening. In contrast, when α is large, the weight is approximately 0, and the actual age at prostate cancer death is the same as that projected in the presence of screening. For values of α between these extremes, weight depends on the lead time, with longer lead times leading to more weight being placed on age at prostate cancer death in the presence of screening. We use 4 values of α to achieve mortality reductions due to screening, ranging from 0 to the full effect expected under the stage-shift model. PSA = prostate-specific antigen.

Observed and projected age-adjusted prostate cancer incidence rates per 100 000 men aged 50 to 84 years, by stage at diagnosis.

In the absence of PSA screening, the model projects constant prostate cancer incidence at approximately the level observed in the core 9 catchment areas of SEER in 1985. In the presence of observed PSA screening (39), the model projects prostate cancer incidence that closely replicates the observed rapid increase and subsequent stabilization in local–regional stage incidence and much of the observed decline in distant stage incidence because of early detection. PSA = prostate-specific antigen; SEER = Surveillance, Epidemiology, and End Results.

Effect of varying individual screening strategy components on projected lifetime harms and benefits relative to the reference strategy.

Only the reference strategy (strategy 8) and strategies that differ from the reference strategy in a single variable are shown. Strategies are identified by the strategy number shown in the Table, and varied screening strategy components are in bold. In strategy 9, the screening interval is biennial if the PSA level is <2.5 µg/L and annual if the PSA level is ≥2.5 µg/L. In strategy 20, the threshold for biopsy referral is a PSA level of 3.5, 4.5, and 6.5 µg/L for ages 50–59, 60–69, and 70–74 y, respectively (21). Strategies are sorted in decreasing order by probability of life saved. NA = not applicable; PSA = prostate-specific antigen.

Tradeoff between lifetime probabilities of life saved by screening and overdiagnosis for selected screening strategies.

Each point represents the tradeoff for 10 of the 35 screening strategies examined in this study: the reference strategy (strategy 8); strategies that differ from the reference by a single screening variable (strategies 3, 5, 6, 9, 18, 20, and 26 [Appendix Figure 3]); and strategies based on recommendations by the National Comprehensive Cancer Network (strategy 1), the American Cancer Society (strategy 9), and Vickers and Lilja (strategy 22) (8) (see the Table for strategy details). The assumed effects of screening on prostate cancer survival correspond to mortality reductions of 29% (the reduction observed in the ERSPC trial after correction for noncompliance), 20%, 10%, and 0% projected in a simulated version of the ERSPC after 11 years of follow-up. Probability of life saved by screening corresponding to a mortality reduction of 29% is based on the assumption that a patient whose stage was shifted from distant to local–regional by screening receives the survival of the earlier stage. Probability of life saved by screening corresponding to mortality reductions of 20%, 10%, and 0% in the simulated version of the ERSPC are based on a generalization of this stage-shift assumption that projects prostate cancer survival on a continuum between no effect for patients with short lead times and the full stage shift for patients with long lead times. Probability of overdiagnosis is based on model-projected competing risks for prostate cancer detection and other-cause mortality. Lines connect projections under the same mortality reduction. The additional NND to prevent 1 prostate cancer death is an established summary measure of the harm–benefit tradeoff in prostate cancer screening compared with no screening, defined as the overdiagnoses divided by lives saved by screening. The NND corresponding to any point in the figure is obtained as the ratio of the probability of overdiagnosis to the probability of life saved. For reference, dashed lines radiating from the origin (representing no screening) illustrate fixed NND values of 5, 10, and 20. For a given probability of overdiagnosis, as the probability of life saved by screening decreases, the corresponding NND increases. For the mortality reduction of 29%, NNDs range from 7.1 (strategy 1) to 3.6 (strategy 26), and for the mortality reduction of 10%, NNDs range from 16.5 (strategy 1) to 9.9 (strategy 26). A strategy that falls between 2 NND lines (for example, “NND = 5” and “NND = 10”) has an NND between those NND values. Different strategies will be preferred depending on relative weighting of the probabilities of life saved and overdiagnosis. Among the strategies considered, strategy 1 maximizes the probability of life saved and will be the preferred strategy if survival is the highest priority. Strategy 26 minimizes the probability of overdiagnosis and will be preferred if the morbidity associated with treatment is the greatest concern. For priorities between these extremes, the preferred strategy will be based on the most favorable balance between probabilities of life saved and overdiagnosis. For example, assuming the mortality reduction of 29%, a target tradeoff of 5 or fewer overdiagnoses per life saved (that is, NND ≤5) would identify strategies above and to the left of the “NND = 5” line. Assuming a mortality reduction of 20%, a target tradeoff of 5 or fewer overdiagnoses per life saved identifies strategy 20 as the only option. No strategy satisfies a target tradeoff of NND ≤3. No strategy satisfies a target tradeoff constraint of NND ≤10 under a mortality reduction assumption of 10%. ERSPC = European Randomized Study of Screening for Prostate Cancer; NND = number needed to detect.