Model of Individualized Estimates Shows Risk of Overdiagnosis in Screening-Detected Prostate Cancers Varies Widely by Age, PSA, and Gleason Score
It is estimated that overdiagnosis occurs in approximately 20% to 40% of prostate cancers detected at screening in U.S. men aged 50 to 84 years, with the likelihood of overdiagnosis varying widely according to patient age and tumor characteristics. In a study reported in the Journal of the National Cancer Institute, Gulati et al used a computer microsimulation model of prostate cancer natural history including information on age, Gleason score, and prostate-specific antigen (PSA) level to predict likelihood of overdiagnosis. They found that risk of overdiagnosis ranges from 2.9% to 88.1%, varying considerably according to age and PSA level.
Study Details
The computer model was used to generate virtual life histories for 10,000 patients with nonmetastatic prostate cancer in the presence and absence of PSA screening, including an indicator of whether screening-detected cancers are overdiagnosed. A logistic regression model was fit to patients diagnosed by screening with PSA < 10 ng/mL and a nomogram was constructed to predict individualized risk of overdiagnosis on the basis of age, Gleason score, and PSA level at diagnosis.
The calibrated microsimulation model closely reproduced incidence trends from the Surveillance, Epidemiology, and End Results registries by age, stage, and Gleason score. In the model, 33% of men would have disease onset in their lifetime and 38% of these men would be diagnosed without screening, with an average interval from onset to diagnosis of 14 years.
Among the 10,000 virtual patents diagnosed by PSA screening, age distributions were 2.2% for 50 to 54 years, 10.8% for 55 to 59 years, 16.7% for 60 to 64 years, 17.1% for 65 to 69 years, 20.6% for 70 to 74 years, 19.3% for 75 to 79 years, and 13.3% for 80 to 84 years. Gleason scores were ≤ 6 for 64.0% and ≥ 7 for 36.0%, and PSA distributions were 39.6% with 4.0 to 4.9, 21.1% with 5.0 to 5.9, 14.0% with 6.0 to 6.9, 8.1% with 8.0 to 8.9, and 5.9% with 9.0 to 9.9 ng/mL.
Effects of Age, Gleason Score, and PSA
The fitted logistic regression model showed a significant 12.9% increase in risk of overdiagnosis with each 1-year increase in age between 50 and 84 years (odds ratio [OR] = 1.129, P < .001), a significant 19.5% decrease in risk for Gleason score ≥ 7 vs ≤ 6 (OR = 0.805, P < .001), and a significant 16.6% decrease in risk for each 1-ng/mL increase in PSA at diagnosis between 4 and 10 ng/mL (OR = 0.834, P < .001).
Predictions of Overdiagnosis Risk
Predictions based on the developed nomogram were found to be reasonably accurate, as indicated by area under the receiver operating curve of 0.75. The predictions show a marked effect of age on risk of overdiagnosis. For example, among men with Gleason score ≤ 6 and PSA level of 4.0 to 4.9 ng/mL, risk increased from 11.6% for men aged 50 to 54 years to 59.9% for those aged 70 to 74 years and 83.4% for those aged 80 to 84 years. As the investigators noted, this is “a range that may have important clinical implications for decisions about pursuing aggressive treatment.”
Similarly, lower PSA levels were associated with greater risk of overdiagnosis, with risk approximately doubling between levels of 9.0 to 9.9 ng/mL and 4.0 to 4.9 ng/mL, in many cases within the same age group and Gleason score category. For example, among men aged 50 to 54 years with Gleason score ≤ 6, risk was 5.0% among those with PSA of 9.0 to 9.9 ng/mL and 11.6% among those with a level of 4.0 to 4.9 ng/mL.
Overall, among patients with Gleason score ≤ 6, risk by age group ranged from 11.6% to 83.4% at PSA of 4.0 to 4.9 ng/mL, from 9.9% to 80.8% at 5.0 to 5.9 ng/mL, from 8.4% to 77.8% at 6.0 to 6.9 ng/mL, from 7.1% to 74.5% at 7.0 to 7.9 ng/mL, from 6.0% to 70.9% at 8.0 to 8.9 ng/mL, and from 5.0% to 67.0% at 9.0 to 9.9 ng/mL. Among patients with Gleason score ≥ 7, risk by age group ranged from 9.6% to 80.2% at PSA of 4.0 to 4.9 ng/mL, from 8.1% to 77.2% at 5.0 to 5.9 ng/mL, from 6.9% to 73.8% at 6.0 to 6.9 ng/mL, from 5.8% to 70.1% at 7.0 to 7.9 ng/mL, from 4.9% to 66.2% at 8.0 to 8.9 ng/mL, and from 4.1% to 62.0%, at 9.0 to 9.9 ng/mL.
The investigators concluded, “The chances of overdiagnosis vary considerably by age, Gleason score, and PSA at diagnosis. The overdiagnosis nomogram presents tailored estimates of these risks based on patient and tumor information known at diagnosis and can be used to inform decisions about treating PSA-detected prostate cancers.”
Roman Gulati, MS, of Fred Hutchinson Cancer Research Center, is the corresponding author for the Journal of the National Cancer Institute article.
The study was supported by the National Cancer Institute.
The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.
