Increase in Prostate Cancer Distant Metastases at Diagnosis in the United States

Following the introduction of prostate-specific antigen (PSA) screening in the early 1990s, there has been a 50% decline in prostate cancer–specific mortality and more than a 70% decline in the incidence of metastases at diagnosis.¹ Given the recent declines in PSA screening and prostate cancer incidence, we sought to assess the effect of these changes on prostate cancer presentation.

Using the most recent Surveillance, Epidemiology, and End Results (SEER) release, we identified 1 107 111 men 40 years or older diagnosed with pathologically confirmed prostate cancer and 545 399 from 2004 to 2013. Temporal variations in cancer characteristics and PSA (available for 2010-2013) were assessed using the Cochrane-Armitage test for categorical variables and Jonckheere-Terpstra test for PSA. Analysis was stratified by age (<75 vs ≥75 years). Incidence of distant metastasis was derived quarterly with SEER Collaborative staging. After standardization to the 2000 US census, restricted cubic spline model with 4 knots was used to examine temporal variation in incidence. Detailed statistical methods and SAS data log may be found at http://mdepinet.org/cornell-claims-based-research-initiative-cbri/.

This study took place from January 1, 2004, to December 31, 2013. It was approved by the Weill Cornell Medicine Institutional Review Board.

As the Table shows, in men younger than 75 years, there was an increase in the proportion of men presenting with distant metastases from 2.7% (95% CI, 2.5%-2.9%) to 4.0% (95% CI, 3.8%-4.2%) and in the proportion presenting with intermediate- and high-grade prostate cancer from 46.3% (95% CI, 45.9%-46.9%) to 56.4% (95% CI, 55.9%-56.9%) (P < .01, respectively). Similarly, in men 75 years or older, there was an increase in the proportion of distant metastases from 6.6% (95% CI, 6.2%-7.0%) to 12.0% (95% CI, 11.2%-12.7%) and intermediate- and high-grade prostate cancer from 58.1% (95% CI, 57.2%-59.0%) to 72% (95% CI, 71%-73%) (P < .01, respectively). Between 2010 and 2013, median PSA at diagnosis increased from 6.0 ng/mL (interquartile range [IQR], 4.6-9.1 ng/mL) to 6.4 ng/mL (IQR, 4.8-10.0 ng/mL) among men younger than 75 years and increased from 9.0 ng/mL (IQR, 5.8-17.3 ng/mL) to 9.7 ng/mL (IQR, 6.2-21.0 ng/mL) among men 75 years or older (both P < .01). (To convert PSA to micrograms per liter, multiply by 1.0.) In adjusted analysis (Figure), the incidence of distant metastases at diagnosis in men 75 years or older decreased from 2004 to 2011 and increased afterward. No change was observed in men younger than 75 years.

This study demonstrates a significant increase in distant metastases at diagnosis in men 75 years or older. The absence of an increase in men younger than 75 years may be due to the US Preventive Services Task Force’s recommendation in 2008 against prostate cancer screening in men 75 years or older, followed by its recommendation in 2012 against screening regardless of age. Although Schröeder et al² demonstrated that PSA screening decreases prostate cancer–specific mortality, the contrasting evidence from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial strongly influenced the US Preventive Services Task Force’s recommendation against PSA screening.³ Recently, Shoag et al⁴ found that the true contamination of PSA screening in the cancer screening trial’s control arm approached 90%, rendering the study uninterpretable. Whether the benefits associated with population-based screening outweigh overdiagnosis and overtreatment remains controversial. Nevertheless, to our knowledge, this is the first national, population-based study to demonstrate that the decline in PSA screening has significantly altered prostate cancer presentation. In men 75 years or older, the incidence of metastasis at diagnosis is increasing following nadir in 2011. Moreover, the proportion of men diagnosed with clinically significant cancer (Gleason grade of 7-10) increased significantly. These features are associated with a lower likelihood of cure and higher risk for prostate cancer–specific mortality.

Our findings differ from those of Jemal et al,⁵ who annually assessed metastases with SEER Summary Stage, the most basic staging system used in cancer registries (it was last updated in 2000). However, clinicians are more likely to understand and use the American Joint Committee on Cancer’s TNM Staging System. We used SEER Collaborative Stage, which standardizes across staging systems to improve data quality.⁶ We also more granularly assessed quarterly incidence. Moreover, yearly incidence ratio analyses from Jemal et al demonstrate that incidences of distant metastases in 2012 and 2013 were significantly higher than in 2011 among men 75 years or older (incidence ratio approximately equal to 1.13; P < .05), which is consistent with our findings.

Our study must be interpreted in the context of the study design. Although we demonstrate an increase in the incidence of distant metastases at diagnosis in men 75 years or older, we did not examine prostate cancer–specific mortality, which requires longer observation before meaningful differences can be discerned. Finally, the categorical age cutoffs presented may not correspond with actual demographic change in screening practices.

Corresponding Author: Jim C. Hu, MD, MPH, Department of Urology, Weill Cornell Medicine–New York Presbyterian Hospital, 525 E 68th St, Starr Pavilion, Ninth Floor, New York, NY 10065 (jch9011@med.cornell.edu).

Correction: The first sentence of the Methods section was corrected online on February 2, 2017.

Published Online: December 29, 2016. doi:10.1001/jamaoncol.2016.5465

Author Contributions: Drs Hu and Sedrakyan had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Hu, Shoag, Sedrakyan.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Hu, Mao, Shoag.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Hu, Mao, Wright, Sedrakyan.

Administrative, technical, or material support: Hu, Mao.

Study supervision: Hu, Nguyen, Shoag, Sedrakyan.

Conflict of Interest Disclosures: Dr Hu reported being on the Speakers’ Bureau for Intuitive Surgical and Genomic Health. Dr Nguyen reported consulting for Ferring and Nanobiotix. No other disclosures were reported.

Funding/Support: This study is supported by Weill Cornell Medicine Patient-Centered Comparative Effectiveness Research Program.

Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: Samprit Banerjee, PhD, Department of Healthcare Policy and Research, Weill Cornell Medicine, provided statistical analysis. He was not compensated for his contribution.