Controversies in Prostate Artery Embolization: Future Best Practice

Abstract

Prevalence of lower urinary tract symptoms secondary to benign prostatic hyperplasia is correlated with age. Men seeking treatment options with a low side effect profile often turn to prostate artery embolization (PAE). PAE continues to be refined with advanced tools and optimized techniques. Nonetheless, there exist controversies in terms of best practices for the management of lower urinary track symptoms (LUTS) with PAE. These controversies are essential for medical progress. Herein we suggest best practices moving forward based on currently available data. Given extensive safety data, we recommend PAE be considered alongside medical management and as a precursor to surgery. Given demonstrated efficacy across gland sizes, PAE can be performed in a single session, ideally in a hybrid angio-CT suite, without preoperative cross-sectional imaging. PAE should be initially performed with 300- to 500-μm size particles, and instead consider exploring other particles and sizes for repeat PAE. Finally, PAE can also be considered as first-line option for recurrent disease given the efficacy and excellent safety profile. This article is not meant to purport a dogma, but rather to serve as a guide to the experienced practitioner in challenging his or her own biases when performing PAE.

Prostate Artery Embolization for Treatment of Lower Urinary Tract Symptoms

Benign prostatic hyperplasia (BPH) is an androgen-driven growth of prostatic transition zone tissues that lead to benign prostatic enlargement (BPE). ¹ Prevalence of BPH, and associated lower urinary tract symptoms (LUTS), is correlated with age. Of note, this pathophysiology mechanism is not completely elucidated as it remains unclear why two men of the same age cohort can have significantly different gland sizes, and why two men with similar gland sizes can have significantly different symptom severity. ² BPE is the most common cause of LUTS, operating through two components: a static component (physical obstruction of the urinary tract named “bladder outlet obstruction”) and a dynamic component (dysfunctional autonomic stimulation of urethral smooth muscle tone).

Diagnostic imaging is not part of standard urologic workup for BPH. Even post void residual (PVR) measurements are considered optional at initial stages of diagnosis. After a medical history and urinalysis, baseline International Prostate Symptom Score (IPSS) and Quality of Life (QoL) questionnaires are obtained and used to track any treatment response. IPSS with QoL is a validated eight-question subjective assessment for LUTS, and arguably a surrogate for overall patient satisfaction. ³ Gland size assessments can be made via formulas using transrectal ultrasound (TRUS) imaging, or contouring from magnetic resonance imaging (MRI) or computed tomographic (CT) images, although it remains unclear whether variability in estimated sizes among these modalities impacts patient treatment choice. ⁴ Since its initial deployment for the treatment of refractory hematuria, prostate artery embolization (PAE) has transformed into a minimally invasive treatment option for LUTS due to BPH. ⁵ PAE appears to alleviate both static and dynamic components of BPE, although the exact mechanism and extent for each is still under investigation. ^{6 7}

Several prospective studies have demonstrated short to mid-term safety and efficacy of PAE across gland sizes. ^{8 9} Recently, two of the most experienced groups have published their 10-year experience with PAE, demonstrating long-term safety and efficacy. ^{10 11} Carnevale et al presented data on 317 men with a mean follow-up of 27 months with IPSS improvement of 16 points, and reported LUTS recurrence after PAE to be less than 25% and median time to recurrence to be 72 months. ¹¹ No patients reported urinary incontinence or erectile dysfunction. In addition, Bilhim et al published the largest cohort of 1,072 patients, with the longest follow-up data to date. ¹⁰ With a mean follow-up of 52 months, patients demonstrated significant improvement in LUTS symptoms and quality of life at last follow-up despite prostatic gland regrowth to its preintervention volume. Compared to Carnevale et al, IPSS reduction of 10 points was reported by Bilhim et al. Likewise, LUTS recurrence at last follow-up was less than 30% with a median time to recurrence of approximately 36 months. Approximately 5% of patients suffered grade ≥3 adverse events, of which the majority (>80%) were penile skin lesions due to non-target embolization—something that can potentially be avoided with higher resolution imaging of nonprostatic collaterals. Furthermore, three grade 4 events occurred in the context of less than 300 μm embolization particle sizes—another controversial topic that will be discussed later. The long-term LUTS recurrence rates after PAE likely reflect the hormonally driven progressive nature of BPH, and are in line with other transurethral therapies where complete gland removal is not performed.

Transurethral resection of the prostate (TURP) is often considered the gold standard surgical approach for the management of BPH, particularly for glands less than 80 mL. Above greater than 80 mL, open prostatectomy (OP) is the gold standard. However, systemic reviews and meta-analyses comparing PAE to TURP have shown equivalent subjective symptomatic improvement. ¹² Furthermore, PAE is associated with a better adverse event profile and shorter hospital stay compared to TURP or OP, and, as of recently, also compared to the newer minimally invasive surgical therapies (MISTs). ^{12 13 14} The most recent randomized trial comparing TURP to PAE reported almost half as many adverse events with PAE compared to TURP, particularly related to sexual function. Of note TURP did result in superior improvements in objective LUTS assessments, and both PAE and TURP resulted in improved subjective LUTS assessments. Although the evidence is mounting, PAE remains an “investigational therapy” per AUA guidelines. ¹⁵ However, the UK NICE guidelines now support the use of PAE and the EAU guidelines lie in the middle. ¹⁶ Retrospective, prospective, and randomized safety and efficacy data to date clearly demonstrate the role of PAE in the treatment of LUTS secondary to BPH; any further debate about its role in guidelines is purely academic or political.

Over the past decade, there have been numerous advancements in PAE with regard to imaging, patient selection, embolization techniques, and tools. Despite attempts to standardize the procedural workflow, many components like catheter selection and particle size are based on patient anatomy and physician preference. ¹⁷ The goal of these standardization efforts has been to make PAE treatment and outcomes reproducible across centers, and in doing so spur high-quality comparative effectiveness research for the management of BPH. This will bring PAE in line with other urologic therapies which have become standardized and are less operator dependent. ¹⁸ However, the multiple changeable components of PAE also allow adaptability specific to each patient's anatomy. The purpose of this article is to highlight recent evidence on such components and suggest “controversial” best practices based on available evidence. Review of additional debates such as initial access choice (radial vs. femoral) and cost-effectiveness has been covered elsewhere. ^{19 20}

Controversy No. 1: PAE Should Be Considered Alongside Medical Management

First-line management for LUTS due to BPH includes lifestyle modifications and medical management with alpha-blockers, phosphodiesterase inhibitors, and/or 5-alpha reductase inhibitors (5ARI). However, these medications are not without side effects, particularly in men of advanced age; and rates of continuation at 1 year are reported to be less than 50%. ²¹ Intolerable side effects, such as orthostatic hypotension and sexual dysfunction, are often the basis for early discontinuation. Previously unknown side effects of long-term medication use, such as cardiac failure and depression, are motivating many patients to forgo trial of 5ARIs and seek “early surgical intervention.” ^{22 23} In fact, recent studies have indicated that prolonged medication therapy has lower efficacy and higher costs than alternative options. ²⁴ Given the safety and efficacy profile of PAE, the authors' controversial viewpoint is that PAE should be considered alongside medical management, and as a precursor to more invasive modalities when early surgical intervention is preferred. Several studies have demonstrated that treatment with PAE does not preclude future medical or surgical management should they be warranted. ⁸

Most recently, 2-year follow-up of a randomized, albeit open-labeled non-inferiority study showed that difference in reduction of IPSS from baseline between PAE and TURP at 2 years was less than 3 points (2.73). ⁸ However, both the initial 12-week results and the 2-year follow-up did not achieve high enough sample size to show definitive non-inferiority of PAE compared to TURP. But, again, patients who underwent PAE experienced almost half as many adverse events. Indeed, safety of PAE has been well documented in comparison to TURP and MISTs. ^{14 25} Additionally, PAE appears to mildly improve erectile function and maintain ejaculatory function, whereas surgical therapies significantly diminish such functions. ³ Given the side effects of medical management and surgical treatment, along with age-related erectile dysfunction, patients should be given the option to individually weigh symptomatic management of BPH alongside maintenance of sexual function. In that regard, PAE should be considered alongside other first-line treatment options.

This paradigm of comparing PAE to medical management and as a precursor to invasive treatment will be a clearer pathway for patients, and lead to increased collaboration with the authors' urologic colleagues. Presenting PAE as a direct alternative to TURP or MISTs will likely confuse patients and create unwanted competition. Fig. 1 depicts an overview of current management of LUTS due to BPH, alongside a proposed one that considers PAE as a precursor to invasive treatment. Of note, for glands greater than 80 cc, the recommended treatment per urologic guidelines is either OP or laser enucleation. While PAE can be utilized as is in these patients, there may also be a role for its use as a downstaging tool to facilitate other MISTs, but formal reports on this are sparse. ²⁶

Fig. 1.

Overview of current benign prostatic hypertrophy (BPH) management per American Urologic Association (AUA) guidelines (black), and the authors' vision for prostatic artery embolization (PAE) in said treatment algorithm (red). *Based on specific minimally invasive surgical treatment (MISTs) gland size cutoff. rPAE, repeat PAE; LUTS, lower urinary tract symptoms; IPSS, International Prostate Symptom Score; QoL, quality of life; IIEF, International Index of Erectile Function; BPH, Benign Prostatic Hypertrophy; PVR, Post-Void Residual; HoLEP, Holmium Laser Enucleation of the Prostate; ThuLEP, Thulium Laser Enucleation of the Prostate; TURP, transurethral prostatectomy.

Controversy No. 2: PAE Can Be Performed without Preprocedural Cross-Sectional Imaging

PAE is a challenging procedure, requiring experience and extensive training to develop catheterization dexterity and understanding of male pelvic arterial anatomy. There is considerable variability among men in prostatic artery origins, and even more so in collateral blood supply both to the prostate and to neighboring organs. ²⁷ However, efficacy and safety depend on thorough identification of pelvic arterial anatomy such that all blood supply to the prostate is occluded while preventing any nontarget embolization. Using digital subtraction angiography (DSA) only at predefined angulations can lead to incomplete identification of prostatic supply, and high levels of radiation exposure. ²⁷ Thus, to better identify soft tissues, target vessels, and anastomoses, both preprocedural MRI or CTA and intraoperative cone-beam computed tomography (CBCT) are utilized. ²⁷

However, as mentioned previously, preoperative imaging is not part of routine BPH workup. And practically, for the consideration of PAE alongside medical management, or other therapies, to take fruition, PAE needs to be performed without prior imaging while still keeping adverse event rates due to non-target embolization near zero. In addition, recent studies have shown that intraprocedural CBCT imaging can better identify prostatic supply than conventional CTA despite the associated scatter and artifact, and in fact reduce air kerma. ^{28 29} The downside to solely relying on intraprocedural imaging is that complete anatomical evaluation needs to be done while the patient is still on the procedural table. So to further aid and expedite the process, novel planning and guidance software have been used. ³⁰ McClure et al have reported initial experience of using guidance software across three centers, and found that it resulted in lower radiation use while increasing understanding of anatomical variants. However, they performed CBCT from each internal iliac artery.

Therefore, the authors' next controversial viewpoint is that PAE should be performed with advanced imaging and image analysis tools such as intraprocedural CBCT or angio-CT and vessel segmentation software, respectively. With enough experience and high-resolution intraprocedural imaging, we believe PAE can be performed in a single session. A hybrid angio-CT suite allows us to obtain soft-tissue imaging void of motion artifacts and scatter, similar to conventional CTA, while obtaining a high-resolution render of pelvic arterial anatomy, similar to a distal aortic or iliac CBCT. Workflow optimization in the setting of trauma and superior diagnostic imaging during hepatic radioembolization have been achieved with performing these IR procedures in a hybrid angio-CT suite. ^{31 32} The authors described how satellite lesions and perfused liver volumes could be more accurately assessed during mapping angiography. Similarly, automated feeder detection software has improved identification and tracking of arteries intended for embolization. ³³ We propose that PAE should take advantage of these advancements and be performed in a single session combining both three-dimensional arterial roadmap planning and postembolization organ perfusion assessment. By minimizing the imaging burden on patients, PAE is more likely to be co-opted prior to invasive therapies.

We have previously presented the authors' experience doing PAE in a single session in a hybrid angio-CT suite, and reported the authors' rates of protective coil embolization to be approximately three times that of published ( N  = 43 patients; 65 vs. 20–26%), ³⁴ without major or minor adverse events. Additionally, improved soft-tissue imaging and contrast resolution has led us to identify subtle variations in origins and collaterals with higher frequencies than prior reports. ³⁴

Figs. 2 and 3 depict the authors' current PAE workflow from initial patient evaluation after referral to IR through postprocedural follow-up, and representative imaging acquired during the procedure. Finally, although IPSS reduction appears to be correlated with gland size, PAE has been shown to be clinically effective regardless of size. ³⁵ As such with appropriate tools and technical proficiency, PAE can be pursued without prior cross-sectional imaging documenting gland size, similar to medical therapy. ¹

Fig. 2.

The authors' PAE workflow from preprocedural evaluation to postprocedural monitoring. PAE, prostatic artery embolization; CBC, complete blood count; CMP, complete metabolic profile; PT, prothrombin time; INR, international normalized ratio; UA, urinalysis; IPSS, International Prostate Symptom Score; QoL, quality of life; IIEF, International Index of Erectile Function; PVR, Post-Void Residual; CT, Computed Tomography; PSI, Pounds per Square Inch; DSA, Digital Subtraction Angiography; PA, Prostatic Artery.

Fig. 3.

Representative single-session angio-CT (top) and cone-beam CT (bottom) images. ( i–ii ) Nexaris angio-CT and embolization guidance (Siemens Healthineers, Germany) selecting prostatic arteries on right originating from internal pudendal, and on left originating as a trifurcation off internal pudendal and obturator. ( iii ) 3D planning with automatic arterial segmentation to determine optimal DSA angulation, ( iv–v ) corresponding oblique and AP DSA showing left prostatic supply (arrows). ( vi–vii ) ArtisQ CBCT and embolization guidance (Siemens Healthineers) selecting prostate artery on right originating from anterior division of iliac artery. ( viii ) 3D planning with automatic arterial segmentation to determine optimal DSA angulation ( ix ) corresponding oblique DSA showing prostatic artery (arrow).

Controversy No. 3: 300- to 500-μm Size Embolic Particles Should Be the Initial Embolization Choice

A variety of embolization particles exist, and have most extensively been compared and contrasted in cases of uterine fibroid embolization. ³⁶ While such comparative studies exist for PAE, there is no consensus in terms of best practice. Commonly used particles in PAE are typically polyvinyl alcohol (PVA) polymers (BeadBlock, ¹⁰ Boston Scientific, MN), trisacryl gelatin (TAG) polymers (Embosphere, ³⁷ Merit Medical, UT), and polyzene-coated hydrogel microspheres (Embozene, ⁸ Varian, CA). Other agents that are gaining momentum for PAE include polyethylene glycol (PEG) (HydroPearl, ³⁸ Terumo Interventional, NJ) and N-butyl cyanoacrylate (n-BCA) (Glubran, ³⁹ GEM, Italy). The material properties of these particles have been elegantly described elsewhere, and key points are summarized in Fig. 4 . ^{40 41}

Fig. 4.

Key features of common PAE embolic agents (left) and anatomical depiction of PAE vasculature sizes (right). PAE, prostatic artery embolization; IIA, internal iliac artery; PA, prostatic artery.

Still studies have tried to compare particle sizes, and results to date have been mixed. ^{42 43} More recently, Torres et al randomized patients to three embolization groups with TAG microspheres: 100 to 300 μm, 300 to 500 μm, or combination of the two. ⁴⁴ While no significant differences in IPSS/QoL improvements were noted, there was a significant increase in frequency of minor adverse events in the 100- to 300-μm only group. Studies specifically comparing sexual function in relation to particle sizes are lacking. It appears that erectile function, as measured by IIEF (International Index of Erectile Function), is maintained in PAE regardless of embolic agent used. However, in one study, rates of decreased ejaculate volume in the short term were much higher in patients treated with 100- to 300-μm particles (6/15) compared to patients treated with 300- to 500-μm particles (1/15). ⁴⁵ Furthermore, there have been rare reports of central gland detachment where large amounts of particles less than 300 μm were used. ⁴⁶ This is typical for therapies like Holmium Laser Enucleation of the Prostate (HoLEP) and Thulium Laser Enucleation of the Prostate (ThuLEP) where enucleation of the prostate is desired to relieve physical obstruction. However, the trade-off of this enucleation is a higher incidence of retrograde ejaculation and sexual dysfunction. ⁴⁷ As such it is unclear whether this level of necrosis is a desired outcome of PAE, transforming the safety profile of the procedure to more of an endovascular resection. ⁴⁸

Thus, the authors' next controversial view is that, at least for initial PAE, 300- to 500-μm-size particles should be utilized. Recent randomized trials evaluating PAE have also utilized 300- to 500-μm microspheres demonstrating safety and efficacy without significant adverse events. ^{5 9} The diameter of the main prostatic artery is approximately 2,000 μm, and is correlated with gland volume. ⁴⁹ Intraprostatic branches off the prostatic trunk are approximately 50 to 300 μm in diameter and appear to not correlate with gland size. Therefore, if complete distal occlusion is desired, an increasing amount of small particles will need to be deployed at branching points for large glands. This increases the risk of non-target embolization-related adverse events. As previously discussed, numerous studies utilizing only 100- to 300-μm particles have reported similar improvements in IPSS and QoL at the expense of increased adverse events. Some groups have explored initially utilizing very limited amounts of 100- to 300-μm particles followed by 300 to 500-μm particles when no visible collaterals are present; however, this requires thorough knowledge of prostatic vascular anatomy and meticulous angiographic technique. ¹¹ To maximize the safety profile of PAE, and maintain an advantage over other more invasive options, nontarget adverse events must be avoided.

There are minimal data on PEG microspheres and n-BCA glue so far, and it remains to be seen if glue can distally penetrate and occlude what anatomical studies have shown to be necessary to induce tissue effects. ^{38 50} Particularly, regarding glue, the benefits of a quicker, lower irradiating procedure need to be weighed against risk of nontarget ischemic necrosis and incomplete permanent occlusion if the glue mixture polymerizes prior to intended distribution. Glue has traditionally been used for endovascular embolization of arteriovenous malformations and internal bleeding, such as those secondary to peptic ulcers. ⁵¹ Clear rationale exists for these diseases where rapid occlusion is preferred, and a flow differential exists. However, PAE is a pressure-dependent embolization as opposed to a flow-directed one. Thus, slow controlled injections are used to prevent nontarget reflux. ¹⁷ One would hypothesize that such a technique with glue would lead to premature polymerization and embolization, and conversely rapid injection increases the risk of nontarget occlusion. However, in experienced hands, the initial data for glue do look promising with significant radiation exposure reduction and significantly faster procedure. ³⁹ It will be interesting to see long-term data on glue efficacy, retreatment success, and adverse events in the context of PAE in comparison to the 10+ year data already available for microspheres.

However, based on current data we recommend using 300- to 500-μm embolization particles for the patient's first PAE, which likely occludes the more proximal internodal vessels but maintains safety. ⁴⁹ PAE is considered a very technically challenging procedure and adding on the substantial experience required to safely handle n-BCA certainly adds complexity to the procedure. In settings of repeat PAE, smaller particles or newer materials can be utilized, if repeat PAE is desired by the patient in lieu of surgical resection.

Controversy No. 4: Retreatment Rates Depict a False Narrative of PAE Utility

This segues into the authors' final controversy: the false narrative that PAE retreatment rates and LUTS recurrence are a major drawback of the procedure compared to other treatment options. As mentioned previously, BPH is a hormonal process that can recur even after TURP, with rates up to 30% depending on technique. ⁴⁷ Therefore, zero recurrence in a benign progressive hormonal disease is not feasible, short of complete gland removal. It is true that as of now, LUTS recurrence rates at 1 year are higher with PAE (10–25%) and MISTs (10–30%) compared to TURP (<10%). ^{52 53} But true PAE recurrence rates are likely lower if cases of unilateral embolization are excluded. ⁸ Nonetheless, PAE is the only procedure that is utilized across gland sizes and the only alternative to surgery in large and very large glands. ⁵⁴ Compared to prostatectomy or laser enucleation, PAE does not require general anesthesia. PAE, at worst, has no impact on erectile function, and in fact may improve function. ⁵² Additionally, PAE does not require an intraprocedural urethral instrumentation or postprocedural Foley, hence minimizing discomfort and risk of urethral strictures. Unlike prostate cancer, for many BPH patients the worst outcome after any treatment may not be recurrence, but rather an irreversible decline in sexual and urinary function.

PAE after TURP or MISTs and TURP after PAE are both feasible for refractory disease. ^{10 55} Given the few studies so far, it is unclear whether repeat TURP or MISTs could in theory significantly alter prostatic vasculature to the point of making PAE untenable. In contrast, repeat PAE has been shown to be feasible, safe, and efficacious, particularly in patients who respond to initial PAE. ^{56 57} It remains to be explored why a small selection of patients do not respond to initial PAE. And although repeat PAE is still shown to be safe in these patients, they may fare better from proceeding with more definitive treatments. ⁵⁸

As an extension to this argument, do PAE retreatment rates matter to the patient? Cost notwithstanding, for some it may, but we suspect that for many it does not. If we accept that glandular regrowth is inevitable, then isn't PAE still the safest best minimally invasive option to explore first? Take, for example, a middle-aged patient who is just starting to experience moderate-to-severe LUTS due to BPH. Although he may wish to improve his LUTS symptoms, he may also wish to conserve his genitourinary function for the time being even more. After age-related erectile dysfunction or urinary incontinence sets in, he may then weigh his priorities differently. Just using the aforementioned best practices, in this scenario, he may opt to undergo two to three PAEs over the course of his life before moving forward with physical removal of his prostate. But as discussed in previous sections, efforts are underway to improve PAE technical success, such that rates and risks of reintervention are minimized.

Furthermore, it may be prudent to preserve access to the main prostatic artery during initial PAE. Revascularization after microsphere embolization is known to happen. ⁵⁹ A handful of studies have attempted to utilize different particles or coil after embolization to make prostatic occlusion more permanent, and longer-term follow-up is pending. ⁶⁰ In this study, Galla et al did note that distal catheterization to densely packed detachable coils was feasible in one patient who required repeat PAE because the coil was nonocclusive. However, short-term urodynamic, IPSS, and QoL improvement were in line with reports where no adjunctive coil embolization is used. One patient still suffered a severe adverse event due to nontarget embolization. Nonetheless, we still posit that for initial PAE, it may be advantageous to maintain full access to all prostatic vasculature, and for patients who respond to initial PAE but require a repeat PAE, more permanent occlusion may be opted for. On a final note, PAE, even when repeated, appears to be more cost-effective than TURP. ^{61 62} However, it remains to be discussed, whether a sequential therapy of PAE followed by TURP in refractory cases will still be deemed cost-effective. The immediate goals of future technical refinements of PAE should therefore not be 0% recurrence rates, but rather improving safety and efficacy such that the risks of initial intervention and any repeat interventions are minimized.

Summary

PAE continues to be optimized with newer particles, tools, and techniques. ^{63 64 65} It is unclear how many studies will be sufficient to move PAE from “investigational” use into “first-line” use in guidelines, but data to date are more than sufficient to describe the role of PAE in the management of LUTS secondary to BPH. As data matures and evolves, it is clear these debates are more academic than evidence based and are not in the best interest of patients.

Conclusions

In summary, the authors' recommendations are as follows:

1. PAE should be considered alongside medical management, and prior to more invasive treatments like TURP.

2. Intraprocedural CBCT/angio-CT and vessel segmentation software should be utilized during PAE to fully delineate male pelvic arterial anatomy.

3. 300- to 500-μm embolic particles should be considered before smaller size particles or liquid embolics for initial PAE.

4. For patients who wish to prioritize a low side-effect profile for their BPH treatment, repeat PAE may be a feasible option if they respond to initial PAE.

These recommendations should not be accepted de facto . Challenging the status quo is necessary to improve standard of care, as is abiding to the principle of primum non nocere .