The State of Evidence in Prostate Artery Embolization

Abstract

Prostate artery embolization (PAE) has emerged over the past two decades as a minimally invasive, nonsurgical treatment for benign prostatic hypertrophy (BPH). While the majority of evidence for PAE stems from retrospective cohort studies, several seminal randomized controlled trials have been performed comparing short-term outcomes of PAE to transurethral resection of prostate (TURP) and against a sham procedure. Across clinical trials, PAE demonstrates consistent improvement in urological symptoms and quality of life in patients with BPH with low complication rates. When compared to TURP, the results are comparable, but there is a trend for better outcomes in certain clinical parameters with TURP. PAE is a suitable option for patients who are not surgical candidates, prefer nonsurgical treatment with an earlier return to routine activities, and wish to better preserve sexual function.

Benign prostate hypertrophy (BPH) is a common urological disorder affecting aging men. The estimated incidence of BPH is 50% in men older than 50 years increasing to 80% by the 8th decade of life. ¹ Symptomatic BPH results in well-recognized lower urinary tract symptoms (LUTS) including frequency, incontinence, nocturia, and urinary bladder outlet obstruction. The constellation of symptoms can impose a significant impact on men's physical and psychosocial health, resulting in a lower quality of life (QoL). Several treatments are established in the management of BPH-related symptoms. ² Alpha blockers and 5-alpha reductase inhibitors are often utilized as first-line pharmacological treatment, although symptoms can remain refractory to medical therapy alone. Surgical intervention encompassing endoscopic and open urological techniques are available, with transurethral resection of prostate (TURP) remaining the reference surgical treatment for BPH. While proven to be successful in treating urological symptoms, the complication rate following TURP can be as high as 20%, with postprocedure sexual dysfunction from retrograde ejaculation being a major problem. ³ Furthermore, a proportion of patients with troublesome LUTS are not suitable surgical candidates or have preference for less invasive nonsurgical options.

Prostate artery embolization (PAE) has emerged over the past two decades as a minimally invasive day-case treatment for patients with symptomatic BPH. Via a percutaneous arterial approach, PAE delivers embolic material to the prostatic vascular bed incrementally inducing gland shrinkage through an inflammatory cascade, improving clinical symptoms. ⁴ For a treatment directed at an aging male population, PAE has inherent advantages over surgical options such as TURP. PAE can be conducted as a day-case procedure without hospitalization, performed under local rather than general or spinal anesthesia, is associated with earlier return to routine daily activities, and does not routinely require the placement of a catheter postprocedure. Furthermore, evidence suggests that PAE is cost-effective in the treatment of BPH-related LUTS compared to TURP. ^{5 6}

PAE was first described by DeMeritt et al in the year 2000 for the treatment of a patient with severe hematuria of prostatic origin. ⁷ Inadvertently, PAE also leads to improvement in BPH-related urinary outlet obstruction in this patient. A plethora of studies have since then investigated the role of PAE in the management of BPH with reproducible findings. ^{8 9 10 11 12 13 14 15 16} In a retrospective study of 630 patients, Pisco et al demonstrated high technical and clinical success for PAE. ¹⁷ Technical success was 98.1%, with bilateral and unilateral PAE performed in 92.6 and 7.4% of patients, respectively. Clinical success rate was defined as improved symptoms (International Prostate Symptom Score [IPSS] ≤15 points and a decrease of at least 25% from the baseline score), improved QOL (QOL score ≤3 points or a decrease of at least 1 point from baseline), and no need of any medical or other therapy after PAE. Clinical success was reported as 85.1% at short-term follow-up (12 months), 81.9% at medium-term follow-up (1-3 years), and 76.3% at long-term follow-up (3–6.5 years). A recent retrospective study has published outcomes in 1,072 patients who underwent successful PAE with a mean follow-up of 4 years. Clinical success rate was 80% in the first 2 years of follow-up, 70% between 3 and 5 years of follow-up, and 60% between 6 and 10 years of follow-up. Despite these figures, reintervention rates were 20% in the first 5 years and 30 to 60% in more than 5 years post-PAE. ¹⁸

While cohort studies in PAE have demonstrated promising outcomes, the robustness of PAE against more established surgical options such as TURP has remained in question meriting validation through randomized clinical trials. While the current evidence for PAE predominately stems from retrospective cohort studies, a number of randomized control trials (RCTs) have now been conducted to better establish the efficacy of PAE in the treatment of BPH. This article illustrates the major outcomes of the level 1 evidence for PAE ( Table 1 ), with a focus on technical success, clinical efficacy, and safety profile.

Table 1. Summary of randomized controlled trials investigating prostatic artery embolization.

Study	Methods	Participants	PAE technique	Follow-up
Gao et al 19 China	PAE vs. TURP	PAE: 57 TURP: 57	Bilateral/unilateral 355–500 μm PVA	24 mo
Carnevale et al 20 Brazil	PAE vs. TURP	PAE: 15 P-PAE: 15 TURP: 15	Bilateral 300–500 μm embospheres	12 mo
Radwan et al 21 Egypt	PAE vs. TURP	P-PAE: 20 M-TURP: 20 B-TURP: 20	Bilateral 300–500 μm microspheres	6 mo
Insausti et al 22 Spain	PAE vs. TURP	PAE: 23 TURP: 22	Bilateral 300–500 μm PVA	12 mo
Abt et al 23 Switzerland	PAE vs. TURP	PAE: 48 TURP: 51	Bilateral/unilateral 250–400 μm Embozene	24 mo
Zhu et al 24 China	PAE vs. TURP	PAE: 20 TURP: 20	Bilateral 100–300 or 310–500 μm microspheres	12 mo
Pisco et al 25 Portugal	PAE vs. Sham	PAE: 40 Sham: 40	Bilateral 300–500 μm PVA	6–12 mo

Abbreviations: PAE, prostate artery embolization; P-PAE, PErFecTED (proximal embolization first then embolize distally) technique; PVA, poly-vinyl alcohol; TURP, transurethral resection of prostate.

Evidence from Randomized Controlled Trials

Most randomized controlled trials to date have assessed outcomes of PAE compared to TURP as the reference standard but have also been conducted against a sham procedure to negate the placebo effect. Gao et al first performed a randomized controlled trial comparing the outcomes of 57 patients undergoing PAE and TURP, respectively, with follow-up up to 24 months. ¹⁹ Carnevale et al subsequently performed a RCT of 30 PAE patients compared to 15 TURP patients with 12-month follow-up, including 15 patients treated by the PErFecTED PAE technique (proximal embolization first then embolize distally). ²⁰ Radwan et al further assessed the role of the PErFecTED technique in 20 PAE patients against 40 TURP patients, with follow-up to 6 months. ²¹ Insausti et al conducted a noninferiority randomized trial of 23 patients treated with PAE compared to 22 undertaking TURP with 12-month follow-up. ²² Abt et al conducted a randomized controlled trial of 48 patients undergoing PAE compared to 51 patients undergoing TURP with longer follow-up up to 24 months. ²³ Zhu et al conducted a randomized controlled trial of 20 patients undergoing PAE and TURP, respectively, with follow-up to 12 months. ²⁴ This trial has not been considered further in our article as it has been published only in the Chinese language. A single sham-controlled study has been conducted to date, with Pisco et al comparing 6-month outcomes of 40 patients undergoing PAE and a sham procedure ²⁵ ( Table 1 ).

Technical Success

Clinical trials have demonstrated high procedural success in performing PAE. When defined as successful bilateral prostate artery embolisation, technical success across studies has ranged between 75 and 100%. Insausti et al and Radwan et al both demonstrated 100% success rate in performing bilateral embolization in 23 and 20 patients, respectively. ^{21 22} Gao et al reported somewhat lower technical success. Bilateral and unilateral embolisation was performed in 84.2 and 10.5% of patients, respectively. Embolization was not possible in three patients (5.3%) due to tortuosity and atherosclerosis of the iliac arteries. ¹⁹ Similarly, Carnevale et al achieved a technical success rate of 86.7% with the reminder of patients undergoing unilateral embolization due to severe atherosclerosis or occlusion of the inferior vesical artery on one side. ²⁰ Abt et al demonstrated a success rate of 75%, with 25% patients undergoing unilateral rather than bilateral embolization. ²³ Pisco et al performed bilateral embolization in all but three patients (7.5%). ²⁵ The reason for unilateral embolization in these two studies is not clearly stated but presumed due to vascular disorders. Interestingly, both studies assessing PErFecTED PAE demonstrated 100% procedural success in this technique. ^{20 21}

Clinical Outcomes

Urinary Symptoms

The IPSS, a validated patient questionnaire used to assess BPH-related symptoms, is the major outcome measure utilized by clinical trials in the assessment of PAE success. All clinical trials have consistently demonstrated statistically significant improvement in IPSS in PAE patients when compared to baseline. When compared to TURP, however, the results are less conclusive. Abt et al demonstrated a mean IPSS improvement of 9.2 points in PAE patients compared to 12.1 points in TURP patients at 12-month follow-up, in favor of the latter. ²³ Interestingly, no significant difference was found with regard to improvement in the Chronic Prostatitis Index (CPSI) score, an alternative measure of pain and urinary symptoms relating to BPH. Carnevale et al showed greater improvement in IPSS in TURP patients with a mean score reduction of 21.5 compared to 12.5 in PAE patients at 12-month follow-up. ²⁰ Radwan et al also demonstrated significantly better improvement in IPSS in the TURP cohort compared to PAE at 6 months of follow-up, with 18 and 14 points mean score reduction, respectively. ²¹ Other trials, on the other hand, have demonstrated comparable IPSS outcomes between PAE and TURP. Insausti et al demonstrated similar reduction in IPSS between PAE and TURP patients, 21.0 and 18.2 points, respectively, which did not reach statistical significance. ²² Gao et al also showed no significant difference in IPSS improvement between PAE and TURP patients at 24 months of follow-up, although mean postprocedure IPSS scores were significantly lower with TURP in the first 3-month follow-up. ¹⁹ With regard to the PErFecTED PAE technique, similar improvement in IPSS was reported in one study, whereas inferiority in another when compared to TURP. ^{20 21} Against a sham procedure, Pisco et al demonstrated significant improvement of 13.5 points in IPSS with PAE compared to 5.0 in the sham cohort at 6-month follow-up. ²⁵ As expected, sham patients who later crossed over to the treatment arm demonstrated significant improvement in IPSS scores 6 months after PAE treatment.

Functional urological parameters have also been utilized in randomized controlled trials to assess the efficacy of PAE. Urinary flow rate, in terms of maximum (Qmax) or average (Q-average) flow rate, and post-void urine residual (PVR) are the markers most consistently reported. In all studies, uroflowmetry and PVR were improved in PAE patients compared to baseline. As with IPSS, however, the comparative improvement in PAE against TURP has been less conclusive, although somewhat in favor of TURP. Abt et al demonstrated greater improvement in Qmax (10.2 vs. 3.9 mL/s) and reduction in PVR (62 vs. 204 mL) in TURP patients compared to PAE. ²³ Similarly, Carnevale et al showed statistically greater improvement in Qmax scores in TURP patients than both standard and PErFecTED PAE, with mean scores of 27.1, 10.1, and 16.7 mL/s at 12 months of follow-up, respectively. ²⁰ Notably, in this study, Qmax scores at baseline were significantly lower in PAE patients than in the TURP arm. On the other hand, Insausti et al demonstrated similar improvement in Qmax and PVR between PAE and TURP patients. ²² Gao et al also showed similar improvements in peak urinary flow and PVR between the two cohorts at 24-month follow-up. ¹⁹ As with IPSS results, Qmax and PVR were initially significantly better in TURP patients in the first 3 months of follow-up . Radwan et al showed better improvement in Q-average with TURP but similar improvement in PVR between PAE and TURP patients. ²¹ As expected, Pisco et al demonstrated significantly better improvement in Qmax and PVR with PAE when compared to a sham procedure. ²⁵

Prostate Volume

Prostate volume, assessed by either ultrasonography or magnetic resonance imaging, has been assessed in PAE clinical trials. Predictably, surgical resection resulted in lower prostate volumes across studies. Abt et al demonstrated reduction in prostate volume in TURP patients of 30.2 mL compared to 10.6 mL in PAE patients than PAE at 24 months of follow-up. ²³ Carnevale et al also found significantly improved reduction in prostate volumes in TURP (24.6 mL) when compared to either standard (12.1 mL) or PErFecTED PAE (16.2 mL). ²⁰ Insausti et al showed greater improvement with TURP than with PAE, with 44.7 and 20.5 mL reduction in prostate volume, respectively. ²² Gao et al also found a greater improvement reduction with TURP, with a mean prostate volume reduction of 36.9 mL compared to 29.8 mL in PAE at 24 months of follow-up. ¹⁹ Radwan et al also showed significantly smaller prostate volumes post-TURP compared to PAE. ²¹ Pisco et al found significant reduction in prostate volumes in PAE, whereas in sham patients it was not significantly altered. ²⁵

Quality of Life

The potential role of PAE in improving QoL in patients with BPH is clearly an important surrogate marker for clinical success and has been assessed consistently across clinical trials. Most trials have assessed QoL using a patient scale questionnaire ranging from 0 (delighted) to 6 (terrible) in assessing outcomes. Carnevale et al demonstrated similar improvement in QoL between TURP, standard PAE, and PErFecTED PAE, with mean scores of 0.9, 2.2, and 1.6 at 12-month follow-up. ²⁰ Gao et al showed similar QoL outcome between PAE and TURP patients at latest 24-month follow-up, although the results in the first 3 months significantly favored TURP. ¹⁹ Insausti et al demonstrated better improvement in QoL with PAE compared to TURP, with a mean reduction in QoL scores of 3.78 and 3.09 points, respectively. ²² Abt et al, on the other hand, demonstrated better improvement in QoL with TURP patients compared to PAE, but this was assessed differently using the QoL portion of the IPSS questionnaire (3.26 vs. 2.26 points). ²³ When compared to sham, Pisco et al demonstrated significant improvement in QoL following PAE, refuting the placebo effect. ²⁵

Adverse Events

Clinical trials suggest a trend for lower adverse events with PAE compared to TURP, particularly with regard to sexual dysfunction. Abt et al demonstrated less adverse events in PAE patients when compared to TURP, which occurred most commonly during the first 3 months postprocedure. ²³ The most common adverse event in both cohorts was irritation, pain, or discomfort. Ejaculatory disorders occurred less with PAE and predominately related to diminished ejaculation, whereas complete anejaculation was more common with TURP. In the study of Carnevale et al, TURP patients suffered more severe adverse events. ²⁰ In PAE patients, adverse events were scarce predominately relating to transient pain, hematospermia, or nontarget embolization treated conservatively. Moreover, retrograde ejaculation occurred in all patients treated with TURP, whereas reduction in ejaculatory volume occurred only in three PAE patients. Similarly, Insausti et al showed fewer adverse events with PAE, with the most common being urinary retention in five patients. ²² Erectile and ejaculatory disorders were more common with TURP, with retrograde ejaculation being the most common occurring in eight patients. On the other hand, Gao et al demonstrated more frequent adverse events in the PAE group than with TURP. Acute urinary retention (25.9%) and postembolization syndrome (11.1%) were the most common adverse events following PAE. ¹⁹ Notably, however, this study has been criticized for including both technical and clinical failures as complications in the PAE cohort and not considering hemorrhage requiring transfusion as a major complication in TURP. In the study of Radwan et al, postembolization syndrome (20%) and urinary retention (10%) were the most common complications of PAE, although the frequency of all complications between the two cohorts was not specifically provided. ²¹ Interestingly, Pisco et al demonstrated similar overall frequency of adverse events between PAE and sham patients, with the most common complication in all patients undergoing PAE being hematospermia (9.0%) or hematuria (6.4%). ²⁵

Need for Retreatment

Retreatment rates following PAE require careful consideration. Abt et al reported 10 PAE patients (21%) who required retreatment with TURP within 2 years due to unsatisfactory relief of symptoms. ²³ A further PAE patient required treatment with a nonspecified surgical intervention. In contrast, five TURP patients (10%) required an alternative surgical intervention. Two PAE patients (10%) who developed acute urinary retention in the study of Radwan et al required further unspecified surgical intervention. ²¹ Two patients in the study of Carnevale et al treated with standard PAE, representing 6.6% of the total PAE cohort, had recurrence of LUTS and underwent successful reintervention with TURP. ²⁰ Gao et al reported five patients (9.4%) in the PAE cohort required reoperation compared to two patients with TURP. ¹⁹ While not a RCT, the recent retrospective study by Bilhim et al reports that largest single cohort follow-up, with 1,072 patients treated with PAE over 10 years. ¹⁸ In this cohort, 20% required reintervention in the first 5 years, and 30 to 60% up to 10 years post-PAE. ¹⁸

Discussion

Several randomized controlled trials have now been published assessing the role of PAE in the treatment of BPH. The outcomes of these trials validate the findings of retrospective cohort studies. PAE consistently improves clinical symptoms, functional parameters, and QoL when compared to baseline. While there is a trend favoring TURP in certain clinical parameters, results from the current clinical trials remain inconclusive. Furthermore, there is a trend for lower adverse events following PAE, specifically with regard to sexual dysfunction.

A few systematic reviews and meta-analyses have been conducted to elucidate the overall efficacy of PAE on BPH treatment. A systematic review of three RCTs in addition to two prospective nonrandomized studies suggested that improvement in functional outcomes does favor surgical intervention, although the actual differences are quantitatively small. ²⁶ Significant advantages in terms of safety profile and postprocedure erectile function are, on other hand, seen with PAE. Similarly, a further review of 860 patients from both RCTs and retrospective single cohort studies suggested that TURP is associated with better improvement in several clinical parameters including IPSS, QoL, and Qmax. ²⁷ The improvement in PVR and rate of complications appear comparable between the two cohorts. PAE, on the other hand, was again associated with significantly better sexual function. However, these reviews are hindered by inherent limitations in the amalgamation of randomized clinical trials with lower quality studies. More robust systematic reviews have recently been conducted. In a systematic review and meta-analysis exclusively of randomized controlled trials, Xu et al demonstrated that while prostate volume and Qmax improvement were greater following TURP, there were no significant differences in terms of IPSS, QoL, and PVR between PAE and TURP. ²⁸ The complication rate was also comparable between PAE and TURP, but the former resulted in a lower rate of sexual dysfunction. Similarly, a recently published Cochrane review suggests that PAE may provide similar improvement in urological symptoms and QoL compared to TURP, but PAE is associated with less ejaculatory disorders. ²⁹ It is clear that for patients who consider the preservation of sexual function a paramount factor in their treatment, PAE may offer a better option than TURP, albeit at the expense of a potentially but not conclusively lower clinical benefit.

There are several drawbacks from current clinical trials which need to be considered. Only a few randomized clinical trials have been conducted thus far with a relatively small number of patients. The majority of trials report on short-term outcomes with only two RCTs publishing data beyond 12 months. As a result, retreatment rates are unlikely to be representative in these clinical trials. In their retrospective study of more than 1,000 patients, Bilhim et al demonstrated reintervention rates for PAE as 3.4% at 1 year but increasing to 21.1% at 5 years and to 58.1% at 10 years of follow-up which is an important consideration in PAE. ¹⁸ Blinding in the aforementioned clinical trials has been a further issue. While the study of Pisco et al blinded participants, two clinical trials were open label and the remainder did not offer information on blinding and thus unlikely to have been conducted. The retention rate in most clinical trials is generally high at follow-up. However, in the clinical trial by Abt et al, where clinical outcomes predominately favored TURP, 29.2% of patients were lost to follow-up and over half refused to undergo further urodynamic evaluation at last follow-up. ²³ The minimal clinically important difference (MCID) was not consistently reported on in all studies, an important consideration in the effectiveness of each intervention. The cessation or continuation of medical therapy during the periprocedural or postprocedural period, an important confounding factor, was also not consistently specified. The certainty in the evidence as established by Cochrane has been deemed low to very low as a result of the small sample size, short-term follow-up, and blinding concerns in current clinical trials.

PAE has not been widely investigated against other urological interventions besides TURP. One retrospective study has compared outcomes of PAE to that of open prostatectomy, demonstrating better improvement in IPSS and Qmax with the latter but a lower rate of adverse events with PAE. ³⁰ However, the utilization of open prostatectomy has faded in favor of TURP, making these results less applicable to current clinical practice. Several modern urological techniques have been developed in recent years for the treatment of BPH including Holmium laser enucleation of the prostate (HoLEP) and transurethral prostatic lift (UroLift). ^{31 32} Future clinical trials are warranted to investigate the efficacy of PAE against these modern techniques and are fortunately seemingly underway (NCT05155891, NCT04236687).

The current major evidence confirms that PAE forms an important part of the treatment armamentarium of patients with BPH-related LUTS. Patients who are not suitable for surgical intervention or prefer a nonsurgical treatment with earlier return to daily activities and better preservation of sexual function are likely to be the most suitable candidates for PAE. Patient counseling and selection is as such paramount to obtaining a promising outcome with PAE. Future robust multicenter clinical trials are warranted to assess long-term outcomes of PAE and its efficacy against other urological techniques besides TURP. Future studies must also prioritize techniques to minimize reintervention rates following PAE.