Archives

  • 2019-10
  • 2019-11
  • 2020-03
  • 2020-07
  • 2020-08
  • br After propensity score matching to balance factors that w

    2020-08-14


    After propensity score matching to balance factors that were associated with the receipt of MRI, standardized differences for all variables were less than 10% between patients who received prostate MRI and their matched counterparts, indicating ade-quate matching based on the selected characteristics (Supple-mentary Table 1). In conditional logistic regression analysis, receipt of prostate MRI surrounding diagnosis was associated with a significantly higher likelihood of receiving observation in the first year (OR = 1.90, 95% CI 1.56-2.32).
    We performed multiple sensitivity analyses to evaluate whether the association between MRI and observation was robust to the time period that imaging was obtained relative to diagnosis and treatment. Analyses performed using a window of 3 months preceding until 3 months following diagnosis, as well as 6 or 9 months revealed similar results (data not shown). Exclusion of 60 patients who received MRI before diagnosis, and the exclusion of 25 patients who received MRI both prior to and following diagnosis also yielded similar findings. To address potential misclassification of MRI obtained for the purpose of treatment planning for radiation therapy or RP, we also explored another cutoff prior to treatment. The use of a shorter interval 
    from MRI to treatment (14 days) did not meaningfully change the association between MRI and observation for low-risk PCa
    DISCUSSION
    Examining the association between prostate MRI sur-rounding the diagnosis of PCa and the use of observation for low-risk PCa in a nationally representative cohort of Medicare beneficiaries, we identified a 3-fold increase in the use of MRI for men with low-risk PCa from 2010-2013, and observed substantial variation in utilization relating to age, race, socioeconomic status, and healthcare factors including urologist density. Compared with matched patients with similar clinical characteristics including diagnosis year, the odds of receiving observation vs definitive treatment were twice as high for patients who received MRI in the period surrounding diagnosis.
    As evidence supporting the safety of AS strengthens with the maturation of numerous longitudinal studies, efforts to decrease primary treatment of low-risk disease have been enthusiastically advocated.3,21-23 Utilizing prostate MRI to improve disease characterization at initial diagnosis has been offered as a means to enhance confi-dence in AS, yet the viability of this approach has not yet been empirically demonstrated. Over the past decade, technical modifications including the 540737-29-9 of mul-tiple imaging parameters such as diffusion-weighted imag-ing have improved the diagnostic accuracy of prostate MRI. These improvements have culminated in the refined prediction of tumor stage, presence of occult high-grade disease, both at initial diagnosis and among men enrolled on AS.24 As a consequence, the use of prostate MRI has been integrated into clinical practice guidelines as a com-ponent of the initial staging evaluation prior to enroll-ment in AS. In addition, a majority of practicing urologists report favorable opinions of the use of prostate MRI in the diagnostic pathway of PCa.25 From this
    Table 1. Characteristics of 8144 patients with low-risk prostate cancer patients by prostate MRI, 2010-2013
    Overall
    Prostate MRI
    Yes
    No
    Race
    Age at Diagnosis (in years)
    Year of Diagnosis
    Marital Status
    Elixhauser Score
    Flu Shot Before Diagnosis
    SEER Region
    Metro/Rural
    Zip Code Median Income
    Stage
    Gleason Score
    Hospital Referral Region Urologist Density (per 100,000)
    SD, standard deviation.
    perspective, our findings are novel and impactful in dem-onstrating a population-level association between MRI imaging and increased use of conservative management. This work can serve to inform efforts to understand the 
    impact of novel technologies on disease management in low-risk PCa.
    We observed increasing frequencies of prostate MRI use from 2010 through 2013, in the backdrop of shifting
    Table 2. Factors associated with receipt of prostate MRI surrounding diagnosis among 8144 patients with low risk prostate cancer, 2010-2013
    Prostate MRI Odds Ratio* 95% Confidence Interval
    Yes No
    Race
    Age at Diagnosis (in years)
    Year of Diagnosis
    Marital Status
    Elixhauser Score
    Flu Shot Before Diagnosis
    SEER Region
    Metro/Rural
    Zip Code Median House Income
    Stage
    Hospital Referral Region Urologist Density (per 100,000)
    * All variables were mutually adjusted in the model.
    detection and treatment patterns. Well-reported declines in prostate-specific antigen screening that occurred fol-lowing recommendations against screening have culmi-nated in lower rates of disease detection across the risk spectrum.26 Moreover, we observed a nearly 20% reduc-tion in the proportion of patients with low-risk PCa receiving treatment, likely reflecting the growing accep-tance of renal tubule practice. We anticipate further changes in practice patterns including decreased rates of screening