A comprehensive analysis of prostate cancer treatments has revealed that randomized controlled trials (RCTs) accurately predict—and may even slightly underestimate—the benefits of these therapies in real-world clinical settings. The study, published in JNCI Cancer Spectrum, compared outcomes of drugs approved between 2010 and 2019 for metastatic castration-resistant prostate cancer (mCRPC).
"What we actually saw is that clinical trials tend to underestimate the benefit of these drugs in the real world," explained Dr. Stephen J. Freedland, professor of urology at Cedars-Sinai in Los Angeles, who led the investigation. "The key take-home is that the benefits we see in clinical trials, we actually can replicate in the real world."
Bridging the Gap Between Trials and Clinical Practice
The research team examined several prostate cancer therapies with different mechanisms of action, including androgen receptor-targeted therapies like abiraterone acetate (Zytiga) and enzalutamide (Xtandi), the immunotherapy sipuleucel-T (Provenge), and the radiopharmaceutical radium-223 (Xofigo).
A common concern in oncology is whether the carefully selected patient populations in clinical trials—who tend to be younger and healthier than the average patient—truly represent how treatments will perform in everyday practice. This study provides compelling evidence that the efficacy demonstrated in controlled research settings translates effectively to diverse patient populations.
"I think it does give us confidence that clinical trials can predict how these drugs will work in the real world," Dr. Freedland noted. "It might have been a healthy, predominantly White patient population that was studied, but it's going to apply to my more representative, my more diverse group of patients that I'm seeing."
Methodology and Key Findings
The researchers employed a methodologically rigorous approach to account for selection bias that naturally occurs after drug approval. They compared outcomes in patients who would have qualified for these medications before they became available to those who received the treatments after approval.
"If you don't account for that [selection bias], the real-world benefits seem to be a little bit less than what we saw in the clinical trials," Dr. Freedland explained. "But if you take patients who would have had access [and] would have been eligible for those drugs, but before they became available—so basically, there is no selection bias of who got the drug and who didn't, because nobody got the drug because it wasn't approved, but the patient would have qualified for the drug—what we actually saw is that clinical trials tend to underestimate the benefit."
The consistency across different drug classes was particularly noteworthy. "What was reassuring is we looked at drugs with 3 different mechanisms of action... All of them showed similar or better effectiveness in the real world than we saw in the clinical trials," Dr. Freedland emphasized.
Implications for Future Therapies
While these findings are encouraging for currently approved treatments, Dr. Freedland acknowledged the limitations in predicting how newer therapies might perform. "The challenge is, from a statistical study design point of view, you need enough years after that drug has been approved to see how it's going to work in the real world... We need to follow patients, so it's always going to lag by a few years."
Nevertheless, the broad consistency across different drug classes suggests that the pattern may hold for newer therapeutic approaches as well. "Is that a guarantee that'll apply to every new drug, every new mechanism action? Of course not. But I think it does give us more confidence than just 1 drug and 1 mechanism of action. It seems to be pretty broad," Dr. Freedland said.
Clinical Significance
For clinicians treating prostate cancer patients, these findings provide reassurance that the evidence base supporting treatment decisions is robust. The modest differences between trial and real-world outcomes suggest that physicians can confidently apply clinical trial data to their practice, even when treating patients who might not have met the strict eligibility criteria of the original studies.
This research addresses a critical question in oncology about the external validity of clinical trials and offers a positive answer: the benefits demonstrated in controlled research settings do translate to everyday clinical practice, potentially with even greater efficacy than initially reported.
