Racial and Ethnic Disparity in the Use of Prostate MRI - a UroToday Journal Club - Christopher Wallis & Zachary Klaassen
April 30, 2022
Christopher J.D. Wallis, MD, Ph.D., Assistant Professor in the Division of Urology at the University of Toronto.
Zachary Klaassen, MD, MSc, Urologic Oncologist, Assistant Professor Surgery/Urology at the Medical College of Georgia at Augusta University, Georgia Cancer Center
Christopher Wallis: Hello, and thank you for joining us for this UroToday Journal Club discussion. Today, we're talking about a recent publication entitled, Racial and Ethnic Disparities in the Use of Prostate Magnetic or Resonance Imaging Following Elevated Prostate-Specific Antigen Rest. I'm Chris Wallis, an assistant professor in the Division of Urology at the University of Toronto. With me today is Zach Klaassen, an assistant professor in the Division of Urology at the Medical College of Georgia. You can see here, the citation for this recent work published in JAMA Network Open.
Racial disparities in prostate cancer have been long recognized, and these represent among the greatest disparities in incidence and outcomes of health conditions that are widely tracked. In manifesting how this may come to be, we can examine differences in PSA screening, in the intensity of active surveillance applied for men diagnosed with low-risk prostate cancer, and in the adherence to PSA surveillance following radical prostatectomy. These variations in care may contribute to increased prostate cancer related mortality among Black men.
We know that the use of MRI in the evaluation of men with elevated PSA levels or suspicion of prostate cancer for other reasons has significant benefits. And these data come from the PRECISION trial. And you can see that among men who are a randomized to an initial MRI prior to a biopsy, there is increased diagnosis of clinically significant prostate cancer, a decreased diagnosis of clinically insignificant prostate cancer, and substantial avoidance of biopsy, with up to 30% of men avoiding biopsy altogether. And so these data, combined with other studies, including the PRECISE trial, have formed the basis of guide and recommendations from the AUA and the NCCN to use MRI prior to biopsy in men with suspicion of prostate cancer. However, this approach has not been uniformly adopted and lack of standardization linked to variations of potential disparities in care.
To address this, the author sought to examine relational and ethnic disparities in the use of prostate MRI for men with the elevated PSA using large commercial and Medicare Advantage-based data sets. This is a retrospective cohort study using the Optum dataset, which is health administrative data for both commercially insured and Medicare Advantage health plans. It's inclusive of 50 states and provides laboratory data on approximately 30% of all of beneficiaries. In the present study, the authors examined male enrollees between 2011 and 2017. According to AUA guidelines on PSA screening, they restricted the cohort to manage 40 years and older, and examined those who underwent a single PSA test with no PSA test or prostate MRI in the preceding years and no subsequent PSA test.
Their outcome of interest was the receipt of a prostate MRI. And as most will know, there's no specific CPT code for prostate MRI, so the authors use an established approach using a pelvic MRI code and relevant prostate cancer indication codes. The authors examined the use of prostate MRI within 180 days following PSA testing. They performed bivariate to multivariate logistic progression models to assess the association between PSA testing and receipt of prostate MRI, and they defined elevated PSA in a variety of ways. First using the classical four nanogram per milliliter threshold, then 2.5, and subsequently 10. Multivariable models were fit, accounting for patient age, race, ethnicity, insurance type, health plan, PSA testing year, and state of residence. Patients that had unknown or missing race and ethnicity data were collapsed into an other category. The authors further performed subgroup analyses stratified by patient age and by these PSA thresholds highlighted above.
The authors use Cox proportional hazards models to assess the association between PSA testing results and the time to subsequent prostate MRI founded, of course, at 180 days, as this was a cutoff. They included the same covariates and performed the same stratified subgroup analyses, as alluded to for the logistic regression models.
At this point in time, I'm going to hand it over to Zach to walk us through the results.
Zachary Klaassen: Thanks so much, Chris. This table looks at the annual frequency of PSA testing and subsequent MRI. And you can see here that there were 794,809 patients that received a PSA test. Overall, there was an upward annual trend observed in the percentage of patients with elevated PSA results undergoing prostate MRI, while the mean number of days between PSA tests and prostate MRI fell slightly over this period from 80 days in 2011 to 60 days in 2017.
This table reports the annual age-stratified frequencies and percentages of patients with PSA tests, patients with a subsequent MRI, patients with PSA levels above 4, patients with a PSA result above 4 undergoing prostate MRI, and the mean number of days between PSA tests and subsequent MRI among patients with a PSA greater than 4. Of note, in 2011, a PSA level above 4 was found at 1.1% of patients between ages 40 and 54, 4.4% of patients between 55 and 64 years of age, 9.4% between 65 and 74 years of age, 70.7% between 75 and 84 years of age, and 22% of those 85 years and older, with a similar rate by age group observed over time. In addition, the highest prostate MRI rates were observed among men age 65 to 74 years, and the lowest rates were observed among older populations.
This table looks at the annual frequencies and percentages of patients with PSA tests stratified by race and ethnicity. Patients undergoing prostate MRI, patients with a PSA level above 4, patients with a PSA level above 4, undergoing an MRI, and the mean number of days between PSA tests and subsequent MRI among patients with a PSA greater than 4. While the rates of prostate MRI remain below 1% in all age, race, and ethnicity groups, rates of receiving prostate MRI increased over time for each racial and ethnicity group. Among patients with PSA levels above 4, the rate of receiving prostate MRI within 180 days after the PSA test date was highest among White patients, with the rate also increasing over time for all groups and all ethnicities.
This figure looks at the multivariable regression results for receipt of MRI at different PSA levels. And you can see here that Black patients are in dark gray, Hispanic are an orange, Asian are in blue, and other are in light gray. On the left is a hazard ratio by PSA score, and you can see that there's no difference in time from PSA test to MRI stratified by race. However, on the right side of this figure, this is the odds ratio by PSA score, Black, Hispanic, and Asian men consistently were less likely to receive a prostate MRI versus White men across all PSA values.
This figure looks at the multivariable regression results by receipt of MRI at different age groups and different PSA levels. On the left, you can see this is odds ratio for ages 40 to 54, in particular Black versus White men with a PSA greater than 4, had an odds ratio of receipt of MRI of 0.60 and a 95% confidence interval of 0.38 to 0.95. With regards odds ratios for patients age 55 to 64, several notable findings in this figure. Asian men versus White men with a PSA greater than 2.5 had odds ratio of 0.43 and a 95% confidence interval of 0.21 to 0.86. Again, with the Asian versus White men for PSA greater than 4, odds ratio of 0.37 and a 95% confidence interval of 0.18 to 0.77. And finally, with Hispanic men, as you can see in the orange dot with PSA greater than 10 versus White men, odds ratio of 0.32 and a 95% confidence interval of 0.18 to 0.56.
On the right side of this slide, you can see odds ratio for ages 65 to 74 years age, Black versus White men with a PSA greater than 4, odds ratio of 0.76, 95% confidence interval of 0.64 to 0.91. And for black versus white men with a PSA greater than 10, again, a significant odds ratio of 0.56 and a 95% conference interval of 0.35 to 0.91. Interestingly for men over the age of 75, there was no differences with regards to receipt of MRI based on race or any PSA values.
Several important discussion points for this study. This study found significant racial and ethnic disparities in the use of prostate MRI after PSA testing, which worsens with higher PSA results. As we mentioned, compared with White patients, Black, Hispanic, and Asian men were significantly less likely to undergo MRI at PSA thresholds of either 2.5, 4, or 10. This racial and ethnic disparity was most pronounced for PSA of 4 threshold. These disparities were also observed across all age groups, which was greatest among men 55 to 6 64, and in Medicare-eligible men, 65 to 74 with a PSA above 10. Interestingly, there was no racial or ethnic disparities for men older than 75 years of age, which is a population for which the USPSTF recommends against screening for prostate cancer.
In conclusion, a significant racial and ethnic disparity exists in the use of prostate MRI following elevated PSA test results, and these disparities widen with higher PSA results. These disparities mirror those for other aspects of prostate cancer care and may contribute to known differences among racial and ethnic populations in prostate cancer outcomes. Policy efforts should be directed at eliminating these racial and ethnic disparities, and finally, clear guidelines on the use of prostate MRI and prostate cancer detection may help standardize the evaluation for prostate cancer across all patients.
We thank you very much for your attention. We hope you enjoyed this UroToday Journal Club discussion.