Disparities in PET Imaging for Prostate Cancer at a Tertiary Academic Medical Center

DISCUSSION

This study found differential use patterns for molecular imaging modalities by race/ethnicity. Although both ¹⁸F-fluciclovine and ⁶⁸Ga-PSMA-11 have dramatically improved the ability to evaluate overall disease burden in patients with prostate cancer, ⁶⁸Ga-PSMA-11 has emerged as the superior tracer, with higher sensitivity (7). The current study demonstrates that when available through clinical trials, access to ⁶⁸Ga-PSMA-11 PET was significantly limited for Black patients, compared with non-Hispanic White patients, by a factor of nearly 4.

The ways in which crucial intermediary steps in health-care delivery between patient presentation and health-care outcomes, such as advanced diagnostic imaging, contribute to disparities in health outcomes are poorly understood across many disease types and imaging modalities. Disparities in health-care imaging have been previously described, somewhat extensively in the field of mammography, and more generally with regard to the frequency of missed care opportunities, as well as differences in wait times to receive advanced imaging (12–18). A more recent study of the follow-up of incidental findings on abdominal imaging demonstrated that elderly patients and emergency department patients were less likely to complete follow-up imaging and that insurance status and race may also contribute to differences in follow-up rates (19). However, the literature examining disparities in radiology remains sparse. The current study demonstrates the essential need for more studies of this kind in radiology as a critical precondition for developing policies and procedures that can identify and eliminate structural barriers to equitable care delivery.

The exact factors that contributed to disparities in use of imaging in this study are unclear. As with disparities in prostate cancer health outcomes, different aspects of social determinants of health may play a role. Our nSES index included a range of variables, including multiple measures of education, income, employment, and housing. The study population was more affluent than average, with 66.6% of patients belonging to the highest statewide nSES tertile, and there was also a higher percentage of non-Hispanic White patients (78.8%) than found in regional census data (52.5%) (20). nSES was not a significant covariate in our analysis, suggesting that additional factors contributed to the disparities that were observed. However, fewer Black patients (40.0%) than non-Hispanic White patients (68.6%) belonged to the highest nSES tertile, and a larger percentage of Black patients belonged to a lower nSES tertile than any other group (27.5%). The combined effect of these demographic patterns may have contributed to our results in ways that could not be directly demonstrated through our specific regression analysis. Additionally, of note, a recent study on the use of ¹⁸F-fluciclovine in a separate tertiary academic medical center suggested possible under-referrals of Black patients for molecular imaging, and similar trends may have occurred at our medical center. More work is needed to better understand the complex social and regulatory factors, including unconscious bias, that may influence imaging access (21).

Although, historically, both imaging agents have been used for similar prostate cancer populations at the study institution, the most glaring difference between the routine use of each of these agents is that ⁶⁸Ga-PSMA-11 has been available only through research trials whereas ¹⁸F-fluciclovine is covered by both government and commercial payors, secondary to its FDA approval in 2016. Remarkably, despite the requirement for study participation and the possibility of self-pay, nearly 6 times as many patients in this study were imaged with ⁶⁸Ga-PSMA-11 as with ¹⁸F-fluciclovine. Thus, the current study’s findings highlight potential disparities in access to imaging research trials for Black patients. This problem is common to many clinical trials but often can be difficult to detect, particularly in radiology trials, as many enroll relatively fewer patients (22). As we improve our understanding of how bias and racism may limit health-care access, more work is needed to develop enrollment strategies designed to promote equitable recruitment, including novel digital marketing interventions, revised screening processes, and sliding-scale financial reimbursement.

Although meaningful conclusions can be drawn from this study, there are several limitations worth noting. For one, this was a single-site retrospective study; therefore, the study sample may not be representative of all patients for whom molecular imaging may be considered. The regression analysis did not adjust for disease attributes such as Gleason score or prostate-specific antigen laboratory value. However, the current institution-specific practice is that both of these modalities are applied similarly, and the use of exhaustive and relatively large study cohorts helps to mitigate this potential confounder. Additionally, differences in individual drivers of nSES, such as income or wealth, may have contributed to the differences that we identified in our study if obscured by the larger nSES composite.

Despite these limitations, this study had several strengths. It had a large sample size, which increased the ability to detect differences between the characteristics of patients who received each type of imaging study. The clear discrepancy highlights the need for more innovative and equitable recruitment strategies, as well as the important role that government regulatory agencies can potentially play in facilitating health equity. Data from ⁶⁸Ga-PSMA-11 PET trials is currently under FDA review, and potential approval in the near future could increase the availability of this drug to more patients.

Finally, at many institutions, ⁶⁸Ga-PSMA-11 PET has been available only in clinical trials through a cost-recovery mechanism, by which the FDA allows centers to pay for the cost of the radiotracer if no corporate entity is developing the drug (23). As ⁶⁸Ga-PSMA-11 was not patented, academia took the initiative in development and most trials were performed under cost recovery. Through this mechanism, patients often would be financially liable for the direct cost of the radiotracer and possibly the cost of the technical component of the PET imaging, which could pose a significant barrier to low-income groups. Additionally, there may be variability in how individual institutions approved cost recovery across institutions.