Contemporary prostate cancer radiation therapy trials may underestimate the risk of biochemical recurrence

Background: Radiotherapy is often given with androgen deprivation therapy for prostate cancer which causes a reduction in testosterone levels, which when below castrate levels, can cause the prostate specific antigen (PSA) levels to be artificially low.

Aim: To determine if high-level radiotherapy clinical trials are underestimating biochemical recurrence (BCR) rates due to inadequate measurement of testosterone levels.

Methods: The study plans for clinical trials performed by the Radiation Therapy Oncology Group (RTOG [now NRG]) on clinicaltrials.gov were reviewed for details on testosterone measurement in trials from 1994 to 2023, namely if the testosterone levels were indexed to PSA levels.

Outcomes: PSA being indexed to testosterone levels and other metrics of testosterone measurement, including time of day of measurement, assay used, and mean testosterone measurement.

Results: Five of 21 (24%) trials stipulated that testosterone levels should be indexed to PSA levels. Eleven of 21 (52%) trials made no mention of testosterone. No trial reported testosterone assay or time of day of measurement. Thirteen of 21 (62%) trials did not require regular follow-up testosterone measurements.

Clinical implications: The number of clinical trials indexing or regularly measuring testosterone was surprisingly low, which could cause an underestimation of BCR rates.

Strengths and limitations: Strengths include being the first study, to our knowledge, to analyze the details of testosterone measurement in high-level radiotherapy trials. Limitations include only analyzing RTOG/NRG trials, analyzing unpublished data, and using clinicaltrials.gov rather than official trial protocols to determine details of testosterone measurement.

Conclusion: Indexing of testosterone levels to PSA levels in high-level radiotherapy trials using androgen deprivation therapy was uncommon, possibly rendering data on BCR unreliable, potentially underestimating BCR rates.