Researchers validate prognostic model for mCRPC

Researchers have developed and validated a clinically useful prognostic model for men with metastatic castration-resistant prostate cancer, which can be used in the front-line setting.

Andrew Armstrong, MD, ScM, of Duke Cancer Institute, Durham, NC, presented on the model at the Genitourinary Cancers Symposium in Orlando, FL.

“Determining prognosis is of utmost importance to patients and their significant others, for life planning and decisions on the appropriateness of therapy and the need for aggressive actions,” Dr. Armstrong told Urology Times.

Prior work by Dr. Armstrong and others have identified a number of prognostic factors in men with advanced prostate cancer, but these new data were needed in order to provide updated data to men that reflect contemporary practice in patients receiving front-line androgen receptor inhibitor therapy and prior to chemotherapy.

Read: Genetic predictors could improve PSA accuracy

Dr. Armstrong and co-authors identified prognostic factors for overall survival in chemotherapy-naive men with metastatic castration-resistant prostate cancer by analyzing the phase III PREVAIL trial (enzalutamide [XTANDI] vs. placebo). His team identified traditional risk factors such as the pattern of spread, the level of PSA, and the presence of pain and other factors, but also identified novel factors such as the neutrophil-to-lymphocyte ratio, each of which independently predicted survival over time. The trial’s dataset is readily applicable to patients in the front-line setting with metastatic castration-resistant prostate cancer, who are considering front-line androgen-receptor directed therapy, he said.

“Importantly, we included enzalutamide treatment in this model, as it is important to communicate prognosis to patients both with and without therapy,” Dr. Armstrong said.

The authors randomly divided 1,159 men into the training set and 550 into the testing set. In the training group, they analyzed 23 prognostic factors, including treatment, and developed a multivariable model predicting overall survival based on the analysis. The final multivariable model, according to the study’s abstract, includes 11 prognostic factors: PSA, treatment, hemoglobin, neutrophil-lymphocyte ratio, liver metastases, time from diagnosis to randomization, lactate dehydrogenase, 10 or greater bone metastases, pain, albumin, and alkaline phosphatase.

Next: What the authors found

The time-dependent area under the curve was 0.74 in the testing set. Median overall survival in the testing set was not yet reached among low-risk patients, but was 34.2 months in the intermediate-risk group and 21.1 months in the high-risk group. The overall survival hazard ratio in the low-risk versus high-risk group was 0.20 and 0.40 in the intermediate- versus high-risk group, according to the abstract.

“Prognosis ranged in our study over 1 to 5 years, and patients and their providers may use this information to help inform on the timing of therapy initiation, the choice, and aggressiveness of therapy choice and for having perspective on decision making,” Dr. Armstrong said. “We did not identify specific predictive factors for enzalutamide benefit, meaning that all variables analyzed including clinical biomarkers, patterns of spread, symptoms, and functional status were unable to distinguish a greater or lesser benefit from enzalutamide. All of these subgroups had a clear survival benefit with enzalutamide over placebo.”

While the authors are planning external validation in a large phase III dataset, Dr. Armstrong said that this is the first large validated prognostic model in chemo-naïve men treated with androgen receptor-directed therapy. Given the widespread use of enzalutamide, the model is clinically applicable to contemporary patients, he said.

Also see - Salvage LND: Complete PSA response achievable

“The inclusion of readily available prognostic markers enables easy incorporation into clinical practice. Future models, however, will need to incorporate genomic alterations such as AR splice variants or other mutations that may be more predictive of enzalutamide response and resistance,” he said.

In addition to clinicians’ using the model for bedside prognoses, the model could be used by clinical trialists designing novel combination therapies with enzalutamide or in the chemo-naïve metastatic castration-resistant prostate cancer space, given the prognostic model’s ability to clearly identify low-, intermediate-, and high-risk subgroups of these men with widely differing survival times.

“In addition, all of these variables have been previously shown to associate with clinical outcomes in metastatic castration-resistant prostate cancer, adding to the clinical validity of the model. The use of this model in chemo-naïve men treated with enzalutamide is novel and of high clinical impact,” he said.

Dr. Armstrong is a consultant/adviser for Medivation and has a financial or other relationship with several pharmaceuticals companies. His institution has received funding from Astellas Pharma, Medivation, and other pharmaceutical companies. Several of his co-authors have a financial or other relationship with Astellas Pharma and/or Medivation and/or other pharmaceutical companies. Several co-authors are employees of Astellas Pharma.

More from Urology Times:

Biochemical recurrence risk high 10 years post-RP

When is a patient cured after prostatectomy?

Ureteroscopy, scribes, coding among hot topics in 2017 (so far)

To get weekly news from the leading news source for urologists, subscribe to the Urology Times eNews.