Long-Term PAE Results: What Do We Know

Abstract

Prostatic artery embolization (PAE) is a minimally invasive technique with proven safety and efficacy to treat lower urinary tract symptoms (LUTS) due to benign prostatic obstruction (BPO) or benign prostatic hyperplasia (BPH). In this review, we discuss the required level of evidence to implement and adopt treatment options for patients with LUTS due to BPO/BPH. Focus is given on the long-term (>3 years) data after PAE with reported outcomes including cohort sizes, follow-up times, reintervention rates (repeat PAE and prostatectomy), need for LUTS/BPO medical therapy, and improvements in International Prostate Symptom Score/quality of life score, peak flow rate (Qmax), postvoid residual, prostate volume, and prostate-specific antigen. The durability of treatment effects after PAE and need for prostatic reinterventions need to be taken into consideration when discussing treatment options with patients and referring colleagues from other medical specialties. Developments in medical devices used for PAE have allowed for a continuous drop in unilateral PAE rates over the last 12 years and will probably play a role in optimizing technical and thus clinical outcomes for patients with LUTS due to BPH/BPO.

Many minimally invasive techniques (MITs) have been tested as alternatives to prostatic surgery for patients with lower urinary tract symptoms (LUTS) due to benign prostatic obstruction (BPO) or benign prostatic hyperplasia (BPH) over the last 40 years. These MITs included dilation of the urethra with balloons or stents, thermal and laser interstitial ablation therapies, transurethral needle ablation (TUNA), transurethral microwave therapy (TUMT), and high-intensity focused ultrasound. Most of these MITs proved to be safe and effective in restricted short- and mid-term cohort studies, but failed to provide long-term sustained relief of LUTS requiring high reintervention rates in the first 5 years after treatment or provided inconsistent LUTS relief across different studies and centers. There are defined requirements for widespread implementation of MITs for patients with LUTS due to BPO. ¹ For prostatic artery embolization (PAE), most of these requirements have been met. They include a proof of concept study, ^{2 3} a placebo/sham comparison study, ⁴ randomized controlled trials against accepted alternative treatments, ^{5 6 7} different cohort studies to understand the generalizability, consistency of treatment effects and potential harms, and systematic reviews and meta-analysis of high-quality primary studies. ^{8 9 10 11} The quality of these data is not only dependent on the type of study but is greatly influenced by the number of included patients, length of follow-up, inclusion and exclusion criteria need to allow good generalizability, and relevant outcomes need to be clearly outlined from the start and reported. ¹ Most of these PAE studies ^{4 5 6 7 8 9 10 11} have reported large cohorts with data from over 1,000 patients from single-arm prospective studies ⁹ and 600 patients from randomized trials. ^{10 11} These studies have provided and reported the most relevant outcome measures including subjective symptom (LUTS) assessment with the International Prostate Symptom Score/quality of life score (IPSS/QoL), patient satisfaction, objective measurements with peak flow rate (Qmax), postvoid residual (PVR), pressure flow studies and prostate volume measurements, erectile and ejaculatory function assessment, length of hospital stay, complications, and speed of return to normal activities. ^{4 5 6 7 8 9 10 11} Cost-effectiveness studies also exist showing that PAE is a cost-effective treatment option. ^{12 13} If all of these data are there, what are we missing with PAE? Well, as with other MITs, PAE needs to stand the test of time. For LUTS/BPO studies, short term is defined as less than 12 months, medium term is defined as 12 to 36 months, and long term is defined as more than 36 months. ¹ In the past, there has been a paucity of MITs studies reporting data in the long term. Most MITs studies report outcomes up to 1 year with less than five studies reporting outcomes over 3 years with small cohort sizes. ¹ Most of these PAE studies have focused on data up to 3 years. ^{4 5 6 7 8 9 10 11 12 13} Long-term PAE studies need to report the need for additional medical treatment, reoperation rates, and urodynamic parameters such as maximum urinary flow rate and PVR volumes along with the remainder outcome measures specified earlier. In this review, we will look at the PAE studies that exist to date focusing on the 2-year data from the Swiss randomized trial, ¹⁴ the data from the UK-ROPE study, ¹⁵ and the long-term PAE studies from the French, ¹⁶ Brazilian, ¹⁷ Portuguese, ¹⁸ and Chinese groups ¹⁹ .

Two-Year Data from the Swiss Randomized Trial

This was a single-center randomized trial comparing PAE with transurethral resection of the prostate (TURP) for prostate sizes of 25 to 80 mL. PAE cohort included 48 patients who were compared with 51 patients receiving TURP. Reductions at 2 years in IPSS (−12.09 vs. −9.21 points, p  = 0.047) and QoL (−3.26 vs. −2.26 points, p  = 0.002) were superior with TURP. No significant differences were noted between PAE and TURP regarding erectile function changes. Objective outcomes were significantly more improved after TURP than after PAE at 2 years, including prostate volume reduction (−30.20 vs. −10.66 mL, p  = 0.005), prostate-specific antigen (PSA) reduction (−2.94 vs. −1.76 mg/L, p  = 0.048), Qmax improvement (10.23 vs. 3.9 mL/s, p  < 0.001), and PVR reduction (204.0 vs. 62.1 mL, p  = 0.005). Adverse events and preservation of ejaculatory function favored PAE over TURP. Reintervention rates after PAE was 21% with 10 patients receiving TURP within 2 years after PAE due to unsatisfying results. Need for LUTS/BPO medical therapy after PAE was reported to be 0% at 2 years. No repeat PAE was done in this cohort. No data was provided over 2 years. ¹⁴

Data from the UK-ROPE Study

This was a multicenter prospective national registry comparing PAE with prostatic surgery performed in the United Kingdom (UK Register of Prostate Embolization: UK-ROPE). Most data were obtained up to 1-year post-PAE with 216 patients in the PAE cohort and 89 patients in the TURP cohort. Reintervention rates were also provided above 1 year (up to 2 years). There was no upper size limit for prostate volume of included patients. Mean baseline prostate volume of the PAE cohort was 101.2 cm ³ (range: 59.5–125 cm ³ ), which was significantly larger than the TURP group (mean: 65.6 cm ³ , range: 50.0–80.0 cm ³ ). Propensity score was used to match groups for all between-group comparisons generating 65 matched pairs. However, prostate volume was not used for matching. Reintervention rates over 1 year were 19.9% ( n  = 43) in the PAE cohort (all received TURP) and 5.6% ( n  = 5) in the TURP cohort. As with the Swiss trial, ¹⁴ TURP provided significant better IPSS/QoL outcomes and greater prostate volume reduction and Qmax improvements when compared with PAE. PAE had a better safety profile with less adverse events than TURP with a faster discharge from hospital. Ejaculation outcomes after intervention also favored PAE (reduced ejaculation in 24.1% of PAE patients vs. 47.5% of TURP patients) as well as return to daily activities (5 days for PAE vs. 14 days for TURP, p  = 0.0001). ¹⁵

French Long-Term PAE Data

This was a single-center retrospective study from the Paris group comparing voiding and storage subscores of the IPSS after PAE. The focus of the study was to assess the differences in IPSS subscore improvements after PAE. However, long-term data above 3 years with a prostate surgery rate of 22.5% ( n  = 11/49) 4 years after PAE were also reported. IPSS/QoL changes were compared from baseline to last follow-up at 4 years. However, prostate volume, Qmax, and PVR changes were assessed only up to 3 months after PAE. ¹⁶

Brazilian Long-Term PAE Data

This was a single-center retrospective study from the São Paulo group that included 317 patients across all prostate volumes, with LUTS, and also patients experiencing acute urinary retention (AUR) over a 10-year period (2008–2018). The study reported outcomes up to 2 years in 317 patients (100%), up to 4 years in 75 patients (24%), up to 5 years in 34 patients (11%), with only 2 patients reported above 5 years (1%). Mean follow-up was approximately 2 years confirming that most of these patients had follow-up below 3 years. For the 37 patients experiencing AUR, the bladder catheter was successfully removed 2 weeks after PAE in 34 patients (92%). Recurrence of LUTS was reported to be 11% at 1 year, 20% at 2.5 years, and 65% at 6.5 years. Prostatic reintervention rates over time were not provided, with an overall rate of 12.0% ( n  = 38), with repeat PAE in 3.8% ( n  = 12), and prostatic surgery in 8.2% ( n  = 26) of the reported patients. ¹⁷

Portuguese Long-Term PAE Data

We have recently published our long-term PAE data from 2009 to 2019, including 1,072 patients across all prostate volumes, with LUTS, and also patients experiencing AUR. The study reported outcomes up to 2 years in 1,072 patients (100%), up to 4 years in 352 patients (33%), up to 5 years in 277 patients (26%), up to 8 years in 109 patients (10%), and up to 10 years in 23 patients (2%). Mean follow-up was 4.39 ± 2.37 years confirming that the majority of these patients had follow-up above 3 years. For the 64 patients experiencing AUR, the bladder catheter was successfully removed 2 weeks after PAE in 57 patients (89%). Recurrence of LUTS was reported to be 20% at 1 year, 22% at 2.5 years, and 30% at 6.5 years. Prostatic reintervention rates over time were provided: 3% at 1 year; 16% at 2 years; 18% at 4 years; 21% at 5 years; 40% at 8 years; and 58% at 10 years. Freedom from prostatectomy rates were as follows: 99% at 1 year; 97% at 2 years; 91% at 4 years; 88% at 5 years; 73% at 8 years; and 65% at 10 years. Repeat PAE rates over time were 2% at 1 year; 6% at 2 years; 8% at 4 years; 10% at 5 years; 13% at 8 years; and 23% at 10 years. Repeat PAE ( n  = 97) allowed for 74% ( n  = 72) of treated patients to remain prostatectomy free during follow-up. Need for LUTS/BPO medical therapy in this cohort was 31% ( n  = 335). Cumulative clinical success rates over time, including all patients receiving more than one PAE but also patients receiving prostate medication during follow-up who remained prostatectomy free and with IPSS/QoL scores below 15/3 points, were 81% at 1 year; 78% at 2 years; 73% at 4 years; 72% at 5 years; 66% at 8 years; and 60% at 10 years. The improvements in subjective outcome measures (IPSS/QoL) after PAE remained significant throughout the 10 years; however, most objective parameters as prostate volume, PSA, and PVR changes from baseline lost significance after 5 years. Qmax remained statistically significantly improved over the 10-year follow-up period. This study also showed that the developments in medical devices over the 10-year period with better microcatheters and wires in last 5 years allowed for a reduction in unilateral PAE rates from 13% during 2009 to 2014 to 6.5% during 2015 to 2019. We also compared clinical outcomes between four different embolic agents including polyvinyl alcohol (PVA) particles, spherical PVA (Bead Block, Boston Scientific Corporation), trisacryl gelatin microspheres (Embosphere, Merit Medical Systems, Inc.), and polyzene-coated hydrogel microspheres (Embozene, Varian Medical Systems) and found no significant differences in safety profiles and clinical outcomes. We have recently looked at our technical results in all PAE-treated patients from 2019 to 2022 and the unilateral PAE rate is currently at 2% (unpublished data Bilhim, T. 2022), proving that developments in medical devices used for PAE are enabling more efficient procedures. ¹⁸

Chinese Long-Term PAE Data

Another recently published long-term PAE single-center and single-arm cohort study from China reported outcomes in 125 patients from February 2014 to February 2020. Mean follow-up time was 2.88 ± 1.28 years with 110 patients with follow-up at 2 years, 40 patients with follow-up at 4 years, and 14 patients with follow-up at 5 years. The cumulative clinical success rates at 2, 3, 4, and 5 years were 82.4, 65.5, 52.4, and 37.4%, respectively. The cumulative LUTS recurrence rates at 1, 2, 3, 4, and 5 years were 6.8, 12.7, 30.6, 44.5, and 60.4%, respectively. The cumulative and reintervention rates at 1, 2, 3, 4, and 5 years were 5.9, 10.2, 26.3, 37.4, and 50.8%, respectively. IPSS/QoL and PV improvements after PAE remained statistically significant during the follow-up evaluations. ¹⁹

Conclusion

PAE has come a long way since the proof-of-concept studies in 2008. ^{2 3} A large volume of quality data exists from the past 12 years (since 2010) to prove the value of PAE in all subtypes of BPH and across all prostate volumes. ^{4 5 6 7 8 9 10 11 12 13 14 15} Long-term data (>3 years) are still scarce, ^{16 17 18} but are already out there with large cohort sizes and reporting the most relevant outcomes for patients with LUTS due to BPO/BPH. Table 1 provides a short summary of the most relevant PAE studies that were outlined during this review. Only four studies have reported long-term outcomes more than 3 years after PAE. The French cohort reported outcomes up to 4 years in 49 patients, ¹⁶ the Brazilian cohort reported outcomes up to 5 years in 34 patients (only 2 reported patients above 5 years), ¹⁷ the Portuguese cohort reported outcomes up to 5 years in 277 patients and 8 to 10 years in 150 patients, ¹⁸ and the Chinese cohort reported outcomes up to 4 years in 40 patients and up to 5 years in 14 patients. ¹⁹ More long-term data after PAE are sure to come. There are many MITs proposed nowadays for patients with LUTS/BPO. ²⁰ Arguably, PAE is the one with the largest and most robust level of evidence, including the long-term data. The durability of treatment effects after PAE and need for prostatic reinterventions need to be taken into consideration when discussing treatment options with patients and referring colleagues from other medical specialties. Developments in medical devices used for PAE have allowed for a continuous drop in unilateral PAE rates over the last 12 years and will probably play a role in optimizing technical and thus clinical outcomes for patients with LUTS due to BPH/BPO.

Table 1. Summary of mid- and long-term PAE studies included in this review.

Study	No. pts <2 y	No. pts 4 y	No. pts 5 y	No. pts 8 y	No. pts 10 y	Reintervention rate at last follow-up	Repeat PAE rate	Prostate surgery rate	Prostate medication rate after PAE	Mean ± SD FU (y)	IPSS/QoL change a	IIEF change a	PV change a	PSA change a	Qmax change a	PVR change a	Comments
Swiss RCT 14	48	–	–	–	–	21% ( n  = 10)	0%	21% ( n  = 10)	0%	2 y	−9.21 b /− 2.26 b	−1.59 c	−10.7 b	−1.76 b	3.9 b	−62.1 b	Randomized trial vs. TURP restricted to PV <80 mL
UK-ROPE study 15	216	–	–	–	–	19.9% ( n  = 43)	0%	19.9% ( n  = 43)	N/A	1 y	−10.9 b /− 2.6 b	1.0 c	−28.6 b	N/A	4.4 b	−40.4 b	National prospective registry comparing PAE with surgery
French study 16	49	30	–	–	–	22.5% ( n  = 11)	0%	22.5% ( n  = 11)	N/A	4.3 ± 0.9	−8.2 b /− 2.2 b	N/A	−23.8 b	N/A	8.9 b	−34.6 b	Single-arm study
Brazilian study 17	317	75	34	2	–	12.0% ( n  = 38)	3.8% ( n  = 12)	8.2% ( n  = 26)	10.7% ( n  = 34)	2.25 (1–8)	−16 b /− 4 b	N/A	−39 b	−2.3 b	6 b	−70 b	Single-arm study
Portuguese study 18	1,072	352	277	109	23	58.1% ( n  = 233)	23.1% ( n  = 97)	35% ( n  = 136)	31.3% ( n  = 335)	4.39 ± 2.37	−10.1 b /− 1.9 b	−1.7 c	−6.8 c	−1.12 c	2.7 b	−8.35 c	Single-arm study
Chinese study 19	110	40	14	–	–	50.8% ( n  = 3)	4.8% ( n  = 6)	21.6% ( n  = 27)	6.4% ( n  = 8)	2.88 ± 1.28	−14 b /− 2 b	N/A	−28.7 b	N/A	N/A	N/A	Single-arm study

Abbreviations: FU, follow-up; IIEF, International Index of Erectile Function; IPSS/QoL, International Prostate Symptom Score/quality of life; no. pts, number of patients reported; N/A, not available; PAE, prostatic artery embolization; PSA, prostate-specific antigen (ng/mL); PV, prostate volume (cm ³ ); PVR, postvoid residual (mL); Qmax, peak flow rate (mL/s); RCT, randomized controlled trial; SD, standard deviation; TURP, transurethral resection of the prostate; UK-ROPE, United Kingdom Register of Prostate Embolization.

^aChanges from baseline to last follow-up.

^bStatistically significant.

^cStatistically nonsignificant.