Outcome prediction of prostatic artery embolization: post hoc analysis of a randomized, open-label, non-inferiority trial

Journal Paper/Review - Dec 17, 2018


Abt D, Engeler D, Schmid H, Kessler T, Zumstein V, Markart S, Güsewell S, Mordasini L, Muellhaupt G, Hechelhammer L. Outcome prediction of prostatic artery embolization: post hoc analysis of a randomized, open-label, non-inferiority trial. BJU Int 2018; 124:134-144.
Journal Paper/Review (English)
BJU Int 2018; 124
Publication Date
Dec 17, 2018
Issn Electronic
Brief description/objective

To identify predictors for different treatment outcomes after prostatic artery embolization (PAE) in the treatment of lower urinary tract symptoms secondary to benign prostatic hyperplasia.

A post hoc analysis of data derived from the 48 patients undergoing PAE in a randomized, open-label, non-inferiority trial was performed. Relative changes in the International Prostate Symptoms Score (IPSS), absolute changes in maximum urinary flow rate (Q ), and relative changes in magnetic resonance imaging-assessed prostate volume from baseline to 12 weeks were defined as the outcomes measures of interest. Their association with various baseline characteristics and measures, technical details of PAE, and early postoperative measures were analysed using Spearman rank correlations and Wilcoxon rank-sum tests. The most promising predictors were further evaluated in receiver-operating characteristic (ROC) curve analyses.

Higher total prostate and central gland (i.e. central plus transitional zone) volumes were associated with more pronounced improvements in the IPSS (Spearman rank correlation [rs]: -0.35 and -0.34; P = 0.01 and P = 0.02, respectively) and the Q (rs: 0.31 and 0.39; P = 0.05 and P = 0.01, respectively). ROC curve analyses suggested that volumes of 39 and 38 mL for total prostate and central gland volume, respectively, would be the optimal thresholds with which to predict PAE success as measured by the IPSS. Other anatomical characteristics of the prostate, such as the central gland index, also showed an even more distinct correlation to the improvement in Q (rs: 0.46, P = 0.003). The relative changes in prostate volume were clearly dependent on the technical performance of PAE. Occurrence of postoperative pain and blood levels of prostate-specific antigen and C-reactive protein emerged as potential early-stage outcome predictors after PAE.

Baseline and peri-operative findings might help to guide patient selection and outcome prediction for PAE. Patients with larger prostates have a higher chance of success with PAE. Larger-scale clinical trials including a longer follow-up are warranted to further elucidate the most suitable patients for PAE.