PSA Response to Antiandrogen Withdrawal To Identify the Response on Type of Antiandrogens - Editorial

In 2021-2022, many more men globally will receive a potent AR inhibitor in either the metastatic hormone sensitive prostate cancer (mHSPC), non-metastatic castration resistant prostate cancer (nmCRPC), or mCRPC settings, with improved long term outcomes based on multiple positive phase 3 trials of abiraterone, enzalutamide, apalutamide, and darolutamide. This new form of maximal or combined androgen blockade in addition to ADT has extended survival, delaying symptomatic and metastatic progression, improving durable remissions based on PSA and imaging response criteria, major successes for our patients.

With first generation AR inhibitors such as flutamide and bicalutamide, prolonged responses to these therapies often led to anti-androgen withdrawal responses which were observed at progression after the patient stopped these agents. These withdrawal responses were due to point mutations in the ligand binding domains of the AR (such as T877A), turning these older agents into agonists that stimulated AR activity and PSA production. Withdrawal responses typically would last 3-6 months, but occasionally patients would experience prolonged withdrawal responses lasting over a year. However, the prevalence of withdrawal responses to stopping these novel AR inhibitors has not been well described.


Soo and colleagues conducted a systematic review of all studies that described withdrawal responses to abiraterone plus prednisone or enzalutamide through November of 2020. They found 23 studies encompassing 1474 men, typically with mCRPC, who were evaluable for analysis, most of whom received 1st generation AR inhibitors (1212) but 108 were included who received a second generation AR inhibitor and 154 received abiraterone plus prednisone. Withdrawal responses were relatively common with first generation inhibitors (26% and 11% with any vs. PSA50 declines, respectively) but less common with abiraterone (26% and 4%), and rare with enzalutamide (7% and 1%). In most of these reports, durability of responses were unfortunately not described, but in those that did, these responses tended to last between a few weeks and a few months and thus were transient.

Collectively, this data suggests that clinical trials may not need to observe patients for these withdrawal responses to enzalutamide as they are very rare and linked to rare F876L agonistic mutations potentially, and deferring potentially effective therapy during this washout for this reason seems unnecessary. For abiraterone plus prednisone, much of the withdrawal response could be due to prednisone agonism and certain AR ligand binding domain mutations such as H874Y or L702H, and glucocorticoid withdrawal syndrome. Knowledge of such mutations from ctDNA panels could be helpful in identifying such patients in advance. Unfortunately, the data clearly also shows that most patients cannot rely on a withdrawal response to abi/enza in this mCRPC setting and should be prepared to rapidly move to an alternative more effective agent upon progression, ideally one that does not share cross-resistance, such as a taxane, sipuleucel-T, radium-223, olaparib or rucaparib, and soon Lu177-PSMA-617, as well as clinical trials of novel approaches.

Written by: Andrew J. Armstrong, MD, Professor of Medicine, Associate Professor in Pharmacology and Cancer Biology, Professor in Surgery, Member of the Duke Cancer Institute, Durham, NC

Read the Full-Text Article: PSA Response to Antiandrogen Withdrawal: A Systematic Review and Meta-Analysis