Predictors of clinical Outcome after Prostate artery

This copy is for personal use only. To order printed copies, contact [email protected] Original Research n Vascular

Tiago Bilhim, MD, PhD João Pisco, MD, PhD José A. Pereira, MD Nuno Vasco Costa, MD Lúcia Fernandes, MD Luís Campos Pinheiro, MD, PhD Marisa Duarte, MD António G. Oliveira, MD, PhD

Purpose:

To assess predictors of outcome after prostate artery embolization (PAE) for benign prostatic hyperplasia with spherical particle polyvinyl alcohol (sPVA) and compare outcomes with the use of nonspherical particle polyvinyl alcohol (nsPVA).

Materials and Methods:

This was a single-center retrospective institutional review board–approved study conducted from 2009 to 2015 in patients undergoing PAE with sPVA (n = 186; mean age 6 standard deviation, 65.5 years 6 7.7) and nsPVA (n = 300; mean age, 65.3 years 6 7.6). The two cohorts were compared and analyzed for predictors of outcome with a Cox proportional hazards model and linear regression. PostPAE prostate ischemia was measured with contrast material–enhanced magnetic resonance (MR) imaging in 23 patients with nsPVA and 25 patients with sPVA. The 24hour post-PAE prostate-specific antigen (PSA) level was registered in 133 patients with sPVA. Prognostic values of MR imaging and PSA levels 24 hours after PAE were assessed with Cox and random-effects regressions.

Results:

Predictors of clinical failure were older age (age over 65 years, P = .002), unilateral procedure (P = .002), and higher baseline International Prostate Symptom Score (IPSS, P = .033). Adjusted hazard ratio for clinical failure of sPVA was 1.273 (P = .16). Acute urinary retention was a predictor of lower IPSS after PAE (P = .002). The mean proportion of prostate ischemia was 11% with sPVA and 10% with nsPVA (P = .65). Lower IPSS after PAE was associated with a higher proportion of prostate ischemia (P = .009). Patients with a PSA level of at least 75 ng/mL (75 mg/L) 24 hours after PAE had a greater decrease in IPSS (P = .01). Prostate ischemic volume and PSA level 24 hours after PAE were correlated (Pearson r = 0.64, P = .014).

Conclusion:

Clinical outcome was similar after PAE with sPVA and nsPVA. Younger age (up to 65 years), bilateral PAE, lower baseline IPSS, and acute urinary retention were predictors of better clinical outcome. The PSA level 24 hours after PAE correlated with prostate ischemia, and both correlated with clinical outcome.

1

From the Department of Interventional Radiology, Hospital Saint Louis, Rua Luz Soriano n°182, 1200-249, Lisbon, Portugal (T.B., J.P., J.A.P., N.V.C., L.F., M.D.); Departments of Anatomy (T.B.), Radiology (T.B., J.A.P., N.V.C., L.F.), and Urology (L.C.P.), NOVA Medical School and Faculdade de Ciências Médicas, Universidade Nova de Lisboa, Lisbon, Portugal; and Departamento de Farmácia, Universidade Federal do Rio Grande do Norte, Natal, Brazil (A.G.O.). Received October 21, 2015; revision requested December 22; revision received February 15, 2016; accepted March 9; final version accepted March 10. Address correspondence to T.B. (e-mail: [email protected]). RSNA, 2016

q

Radiology: Volume 000: Number 0—   2016 n radiology.rsna.org

and Interventional Radiology

Predictors of Clinical Outcome after Prostate Artery Embolization with Spherical and Nonspherical Polyvinyl Alcohol Particles in Patients with Benign Prostatic Hyperplasia1

RSNA, 2016

q

1

VASCULAR AND INTERVENTIONAL RADIOLOGY: Predictors of Outcome after Prostate Artery Embolization

B

enign prostatic hyperplasia is associated with lower urinary tract symptoms (LUTS) in aging men (1–7). Prostate artery embolization (PAE) is a minimally invasive therapy that has been shown to be safe and effective for relief from LUTS associated with benign prostatic hyperplasia (8–22). Nonspherical particle polyvinyl alcohol (nsPVA) has been the most widely used embolic agent for PAE (9–16,18–20). However, other embolic agents that have been used are tris-acryl gelatin microspheres (Embosphere Microspheres; Biosphere Medical, Roissy, France) (8,21) and polyzene-coated hydrogel microspheres (Embozene; CeloNova, San Antonio, Tex) (17,22).

Advances in Knowledge nn No significant differences were found in success rates 12 months after prostate artery embolization (PAE) between spherical particle polyvinyl alcohol (sPVA) and nonspherical particle polyvinyl alcohol (nsPVA, P = .08). nn The independent predictors of better clinical outcome were younger age (up to 65 years, P = .002), lower baseline International Prostate Symptom Score (IPSS, P = .033), acute urinary retention (P = .02), and bilateral PAE (P = .002). nn The proportion of prostate ischemia measured with contrastenhanced MR imaging in the first month after PAE was 0.11 6 0.11 with sPVA and 0.10 6 0.13 with nsPVA (P = .65). nn Lower IPSS during follow-up was associated with a higher proportion of prostate ischemia (P = .009), a greater volume of ischemia (P = .08), and a prostatespecific antigen (PSA) level of at least 75 ng/mL (75 mg/L) 24 hours after PAE (P = .01). nn There was a positive correlation between PSA level 24 hours after PAE and ischemic volume (r = 0.64, P = .014). 2

The clinical failure rate of PAE varies from 3% to 40% and depends largely on the definitions used, the embolic agent choice, and the population studied (8,11,16–19). Furthermore, most clinical failures occur immediately after PAE in patients that never improve (ie, nonresponders) (9–16). However, no studies were found that were focused on the identification of potential baseline patient characteristics or the use of different embolic agents that could be associated with poor clinical outcome. Likewise, the potential prognostic roles of prostate-specific antigen (PSA) values 24 hours after PAE and prostate ischemia measured with magnetic resonance (MR) imaging in the first month after PAE have been scarcely reported (23– 25) and were thus tested in the present study. We aimed to assess predictors of outcome after PAE for benign prostatic hyperplasia with spherical particle polyvinyl alcohol (sPVA) and compare outcomes with the use of nsPVA.

Bilhim et al

bias this work. No funding sources were involved. All authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study Population We conducted a single-center retrospective cohort study of a prospectively collected database from 2009 to 2015 in patients undergoing PAE with sPVA and nsPVA at our institution. The study was approved by the institutional review board, and an informed consent form was signed by every patient. Before and after PAE, the patients were evaluated by a urologist with 20 years of experience (L.C.P.) by using the International Prostate Symptom Score (IPSS), the Quality of Life questionnaire (QoL), and the International Index of Erectile Function (IIEF) and by using the following parameters: prostate volume measured with transrectal ultrasonography, peak urinary flow rate (Qmax), post-void residual volume (PVR), and PSA level. All

Materials and Methods The authors confirm that they had full access to all of the data in this study and had final responsibility for the decision to submit the article for publication. There were no conflicts of interest for any authors or institutions, and there were no financial or personal relationships with other people or organizations that could inappropriately influence or

Implications for Patient Care nn The ideal candidates for PAE are younger patients (up to 65 years of age) with acute urinary retention and baseline IPSS of less than 23 points. nn Bilateral PAE should always be attempted because it is associated with a better clinical outcome. nn The volume of prostate ischemia detected with contrast-enhanced MR imaging 2–4 weeks after PAE and PSA level of at least 75 ng/ mL (75 mg/L) 24 hours after PAE are useful prognostic indicators of clinical outcome after PAE.

Published online before print 10.1148/radiol.2016152292 Content codes: Radiology 2016; 000:1–12 Abbreviations: CI = confidence interval IIEF = International Index of Erectile Function IPSS = International Prostate Symptom Score LUTS = lower urinary tract symptoms nsPVA = nonspherical particle PVA PAE = prostate artery embolization PSA = prostate-specific antigen PVA = polyvinyl alcohol PVR = post-void residual volume Qmax = peak urinary flow rate QoL = Quality of Life questionnaire sPVA = spherical particle PVA Author contributions: Guarantors of integrity of entire study, T.B., J.P., J.A.P., N.V.C., L.C.P., A.G.O.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; agrees to ensure any questions related to the work are appropriately resolved, all authors; literature research, T.B., J.P., N.V.C., L.F., L.C.P., M.D.; clinical studies, T.B., J.P., J.A.P., N.V.C., L.F., L.C.P.; experimental studies, J.P., L.C.P.; statistical analysis, T.B., J.P., L.F., L.C.P., A.G.O.; and manuscript editing, all authors Conflicts of interest are listed at the end of this article.

radiology.rsna.org n Radiology: Volume 000: Number 0—   2016


patients underwent preprocedural computed tomographic (CT) angiography to study the pelvic arteries, and images were reviewed by a radiologist with 5 years of experience (T.B.) (9–15). The inclusion criteria were men over the age of 55 years with a diagnosis of benign prostatic hyperplasia, moderate to severe LUTS (IPSS  18), and QoL score of at least 3, who were refractory to medical treatment for at least 6 months or who refused to undergo medical therapy; those with Qmax of up to 12 mL/ sec or with acute urinary retention and prostate volume larger than 30 mL; and those with sexual dysfunction or those who accepted the risk of developing sexual dysfunction after treatment. Exclusion criteria were malignancy, advanced atherosclerosis and tortuosity of the iliac arteries and/or prostate arteries on preprocedural CT angiographic images, secondary renal insufficiency (due to prostatic obstruction), large bladder diverticula or stones, neurogenic bladder, detrusor muscle failure, active urinary tract infection, and unregulated coagulation parameters (9–15). Patients who underwent a previous prostate surgery or PAE were not excluded.

Embolization Procedure Pre- and periprocedural patient medication was described previously (9–15). All patients presented to the hospital 2 hours before the procedure and were discharged on the same day, 3–6 hours after embolization. Embolization was performed with local anesthesia and a unilateral femoral approach whenever feasible by interventional radiologists with 6 years (T.B., J.P., M.D.) and 3 years (J.A.P., N.V.C., L.F.) of experience with PAE. A Roberts uterine catheter (Cook, Bloomington, Ind) and a 0.035-inch hydrophilic guidewire (Terumo, Tokyo, Japan) were used to catheterize both internal iliac arteries and their anterior divisions. Angiography of the anterior division of the internal iliac arteries was performed in ipsilateral anterior oblique projection (35°) with caudal-cranial angulation (210°). For selective catheterization of the prostate arteries, coaxially placed microcatheters (2.0–2.7-F Progreat;

Terumo, Tokyo, Japan) and 0.016-inch hydrophylic guidewires (GT; Terumo) were used. Before embolization, selective angiography of the prostate arteries was also performed with the same ipsilateral oblique projections. For PAE, 300–500µm particles (Bead Block; Biocompatibles BTG, London, United Kingdom) for sPVA or 100–300-µm particles (Cook) for nsPVA were injected through the microcatheter with the tip in the middle third of the prostate artery trunk. To prepare the sPVA, 4 mL of contrast medium (Ioversol, 350 milligrams of iodine per milliliter; Covidien, Dublin, Ireland) was added to the particles in the vial and slowly stirred until a homogeneous solution was obtained. Particle polyvinyl alcohol (PVA) was prepared as reported previously (9–11,15). Each vial of particle PVA (1 mL) was diluted in a 60-mL solution of contrast medium and saline solution at a 1:1 ratio. Post-PAE control angiography was performed immediately after the particle PVA injection and 3 minutes after the sPVA injection, per manufacturer instructions. The end point of PAE was considered when all arterial branches that supplied the prostate were completely occluded with reflux toward the prostate artery origin (9–15) (Fig 1). Procedural time (from femoral puncture access to catheter removal after PAE), fluoroscopy time, and radiation dose were registered.

Study Design and Outcome Measures Technical success was defined as PAE on at least one pelvic side. Pain was assessed during the procedure, at the time of discharge, and 1 week later with a 10-cm visual analogic scale (9– 15). Adverse events were registered according to the Society of Interventional Radiology reporting criteria (26,27). Patients were prospectively followed up with serial measurements of IPSS, QoL score, prostate volume, PVR, Qmax, PSA level, and IEEF during patient evaluations performed at 1, 6, 12, 18, and 24 months. Clinical success was assessed at each evaluation after PAE and defined as the presence of all of the following: (a) at least a 25% decrease in IPSS from baseline, (b) an IPSS up to 15 points, (c) at least a 1-point


Bilhim et al

decrease in the QoL score from baseline and a QoL score of up to 3 points, and (d) no need for any additional medical or surgical therapy for LUTS. Patient outcomes in the sPVA group were analyzed separately and then compared with those of a historical cohort that consisted of all patients (n = 300) that had been treated with PAE by using 100–300-µm particles for nsPVA at the same institution from 2009 to 2013 according to the inclusion and exclusion criteria and the technique described herein. The short-term and midterm results of 250 of the 300 patients have been reported previously (9–11,15,18). In these prior articles, investigators presented the overall clinical outcomes after PAE (9–11), compared unilateral PAE to bilateral PAE (15), and reported the results of various PVA particle sizes for PAE (18). None of the prior articles dealt with the identification of predictiors of clinical outcome after PAE presented herein. Both cohorts were analyzed for identification of potential predictors of poor clinical outcome after PAE (T.B., A.G.O.). Prostate ischemic volume after PAE was measured with contrast material–enhanced MR imaging within the first month after treatment by using a 1.5-T body imaging unit (Achieva; Philips, Best, the Netherlands) with a four-channel Sense pelvic coil (Philips). Measurements of ischemic volume, total prostate volume, and proportion of ischemic volume to total prostate volume were compared between the sPVA and nsPVA groups by using axial T1-weighted fat-suppressed turbo spin-echo sequences with a repetition time (msec)/echo time (msec) of 500/14, field of view of 300 cm2, matrix of 400 3 300, 5-mm-thick sections, gap of 0.5 mm, and acquisition time of 3 minutes 53 seconds. Intravenous injection of gadovist (Magnevist; Bayer, Berlin, Germany) was performed with 0.2 mmol per kilogram of body weight at 2.5 mL/sec. All measurements were performed by the same radiologist (T.B.) with 5 years of experience in MR imaging of the prostate by using segmented manual measurements on each axial section. In 133 patients from the sPVA cohort, the PSA values were 3


Bilhim et al

Figure 1

Figure 1: Digital subtraction angiograms of a PAE procedure performed with sPVA in a 62-year-old man. (a) Image acquired after selective catheterization of the left prostate artery (white arrow) depicts central gland opacification and the tortuous capsular branches with a “corkscrew” pattern (black arrow). (b) Image was acquired after sPVA embolization of the left prostate artery (arrow). Complete occlusion of the prostate branches with reflux into the main artery origin is seen. (c) Image was acquired 12 months after PAE, with relapsing LUTS after selective catheterization of the left internal pudendal artery. The previously embolized prostate artery disappeared, and collateral blood flow into the central gland of the prostate (white arrow) developed from the penile artery (black arrow). (d) It was not possible to catheterize the prostate artery arising from the penile artery, so another collateral prostate artery arising from the superior vesical artery was catheterized (white arrows). The large collateral vessels with anastomoses to the penile artery are evident (black arrows) and represent accessory pudendal arteries. (e) To avoid untargeted embolization to the penis, the microcatheter was advanced distally to the accessory pudendal arteries (black arrows) into the prostate artery (white arrows). Image acquired after selective embolization shows occlusion of the central gland prostate branches with preservation of the accessory pudendal arteries (black arrows).

registered in the first 24 hours after embolization, compared between unilateral and bilateral PAE, and analyzed as a potential prognostic marker.

Statistical Analysis Confidence intervals (CIs) for proportions were exact (binomial) CIs. Baseline 4

characteristics of the sPVA and nsPVA cohorts were compared with t tests and x2 tests. Rates of clinical success over time were estimated with the KaplanMeier method, and the difference between the two groups was tested with the log-rank test. The Cox proportional hazards model was used to estimate

the hazard rate associated with the use of sPVA by controlling for baseline patient variables (age, presence of acute urinary retention, unilateral procedure, prostate volume, and PSA level) and for evaluation of potential prognostic factors of clinical outcome. We also looked for patient variables associated with



Bilhim et al

Table 1 Summary Statistics of Outcome Measures before and after PAE with sPVA in 186 Patients with Moderate to Severe LUTS That Was Refractory to Medical Therapy Time Point and Parameter Baseline No. of Observations Mean 1 Mo No. of Observations Mean 6 Mo No. of Observations Mean 12 Mo No. of Observations Mean 18 Mo No. of Observations Mean 24 Mo No. of Observations Mean

IPSS

QoL Score

IIEF

Prostate Volume (cm3)

PSA Level (ng/mL)

PVR (mL)

Qmax (mL/sec)

176 22.4 6 6.1

176 4.2 6 0.7

155 16.6 6 7.8

185 88.6 6 51.3

185 4.5 6 4.4

159 91.9 6 91.4

161 10.3 6 5.5

185 11.8 6 7.0

185 2.5 6 1.2

173 16.8 6 8.0

175 66.6 6 39.7

176 3.3 6 3.1

140 66.0 6 73.8

156 13.3 6 7.3

143 10.9 6 7.0

142 2.4 6 1.2

134 18.1 6 7.9

134 66.9 6 38.0

131 2.6 6 2.1

108 54.4 6 57.3

122 13.1 6 6.7

49 10.2 6 6.5

49 2.2 6 1.1

44 18.4 6 7.0

40 71.1 6 46.2

45 2.7 6 2.2

33 76.6 6 92.8

36 13.0 6 6.3

10 9.1 6 5.7

10 2.2 6 08

8 18.6 6 6.0

9 83.6 6 82.3

8 1.9 6 1.9

8 49.3 6 49.6

8 13.8 6 8.1

8 8.1 6 2.7

8 2.4 6 0.7

6 18.8 6 7.0

8 90 6 85.5

7 2.2 6 1.9

7 61.2 6 46.0

8 13.1 6 8.4

Note.—Mean values are presented as means 6 standard deviations. Of the 186 included patients, 10 patients had acute urinary retention before PAE; thus, IPSS, QoL score, IIEF, PVR, and Qmax were not available at baseline. The remaining missing observations at baseline and follow-up are due to missing data and/or follow-up time period not yet reached. To convert nanograms per milliliter to micrograms per liter, multiply by 1.0.

IPSS over time (age, presence of acute urinary retention, unilateral procedure, PVA type, and pre-PAE value of all effectiveness parameters) by using randomeffects linear regression with an autoregressive (1) correlation structure. By using the same model, we tested the effect of PVA type on each outcome measure over time after adjustment for patient variables and the baseline value of the outcome variable. The prognostic value of prostate ischemic volumes detected with MR imaging and PSA values 24 hours after PAE in terms of clinical success was assessed with Cox regression, and random-effects regression was used to assess the prognostic value of IPSS over time. For analysis of the correlation of prostate ischemic volume with PSA level 24 hours after PAE, we used Pearson correlation on logtransformed values of both variables. All tests were two sided, and the significance limit was set at P less than .05. No correction for multiple comparisons was performed. Statistical analysis was conducted (A.G.O., T.B.) with Stata

12.0 software (Stata, College Station, Tex).

Results PAE with sPVA Between October 2012 and April 2015, 232 patients underwent PAE with sPVA; however, 46 patients were lost to follow-up. Thus, 186 patients were included for analysis. Mean patient age 6 standard deviation was 65.5 years 6 7.7 (range, 43–87 years). Table 1 shows the baseline data of the patient population and the summary statistics of the follow-up cohort. Ten patients (5.4%) had acute urinary retention with bladder catheters for 1–6 months before PAE. Twenty-seven patients (14.5%) refused to undergo medical therapy, 87 patients (46.8%) underwent medical therapy with a1-adrenergic receptor antagonist (alfuzosin, 10 mg per day; doxazosin, 4 mg per day; or tamsulosin, 0.4 mg per day), 15 patients (8.1%) underwent medical


therapy with 5-a-reductase inhibitors (finasteride, 5 mg per day; or dutasteride, 0.5 mg per day), and 57 patients (30.6%) were treated with combination therapy (a1-adrenergic receptor antagonist plus 5-a-reductase inhibitors). Six patients (3.2%) had undergone prostate surgery to relieve LUTS 5–20 years before the study, and 13 patients (7.0%) had undergone PAE 6 months to 3 years before the study. For PAE, sPVA was administered with the use of local anesthesia by using a unilateral femoral approach in 172 patients (92.5%) and by using a bilateral femoral approach in 14 patients (7.5%). PAE was bilateral in 166 patients (89.2%) and unilateral in 20 patients (10.8%). Only one vial of 300–500-µm sPVA was used per patient. A mean volume of 2.8 mL (range, 2.5–3.8 mL) of sPVA was slowly injected into each prostate artery by avoiding reflux to other arteries. The post-PAE angiographic images showed complete occlusion of the prostatic vascularization. The median procedure time was 75 minutes (range, 30–255 5


Table 2 Mean Reduction in Outcome Measures at Different Time Points after PAE with sPVA Variable and Time Point IPSS 1 mo 6 mo 12 mo 18 mo 24 mo QoL score 1 mo 6 mo 12 mo 18 mo 24 mo Qmax (mL/sec) 1 mo 6 mo 12 mo 18 mo 24 mo Prostate volume (cm3) 1 mo 6 mo 12 mo 18 mo 24 mo PSA level (ng/mL) 1 mo 6 mo 12 mo 18 mo 24 mo PVR (mL) 1 mo 6 mo 12 mo 18 mo 24 mo IIEF 1 mo 6 mo 12 mo 18 mo 24 mo

No. of Observations

Mean Difference from Baseline

95% CI

176 136 47 10 8

210.7 6 7.87 211.3 6 7.64 212.1 6 6.66 29.5 6 3.66 28.6 6 3.42

211.8, 29.5 212.6, 210.0 214.0, 210.1 212.1, 26.9 211.5, 25.8

176 135 47 10 8

21.70 6 1.26 21.72 6 1.26 21.83 6 1.22 21.60 6 1.07 21.50 6 1.20

21.89, 21.52 21.93, 21.50 22.19, 21.47 22.37, 20.83 22.50, 20.50

140 111 36 8 8

2.60 6 6.46 2.85 6 6.38 2.33 6 7.05 4.68 6 9.84 4.05 6 9.81

175 134 40 9 8

221.5 6 25.4 221.9 6 29.0 215.2 6 23.3 225.3 6 35.0 228.4 6 36.1

225.3, 217.8 226.8, 216.9 222.7, 27.8 252.2, 1.7 258.6, 1.7

175 130 45 8 7

21.17 6 3.32 21.78 6 2.98 21.77 6 2.78 21.21 6 2.56 21.34 6 2.73

21.67, 20.68 22.30, 21.26 22.60, 20.93 23.35, 0.92 23.86, 1.19

126 97 32 8 7

229.1 6 105.0 245.3 6 106.3 237.7 6 129.4 240.0 6 68.5 240.8 6 72.0

247.6, 210.6 266.7, 223.9 284.3, 9.0 297.2, 17.3 2107.4, 25.8

151 115 34 6 5

0.40 6 6.37 1.76 6 6.06 1.44 6 6.22 0.83 6 3.43 20.20 6 2.59

20.62, 1.43 0.64, 2.88 20.73, 3.61 22.77, 4.43 23.41, 3.01

1.52, 3.68 1.65, 4.05 20.05, 4.72 23.55, 12.91 24.15, 12.25

Note.—Mean values are presented as means 6 standard deviations. To convert nanograms per milliliter to micrograms per liter, multiply by 1.0.

minutes). The median fluoroscopy time was 18.2 minutes (range, 4.9–88.3 minutes). The median radiation dose was 2401 dGy · cm2 (range, 655–9202 dGy · cm2). All patients recovered uneventfully and were discharged the same day, 3–6 6

hours after PAE, and were thus treated as outpatients. There were no major adverse events and there was no urinary incontinence after PAE with sPVA. The mean pain score (on a visual analog scale of 0–10)

Bilhim et al

during and after PAE before discharge was 0.7 (range, 0–3). In the week after PAE, the mean pain score was 0.7 (range, 0–6), with 22 patients (11.8%) describing moderate (3–6) pain scores, while 44 patients (23.6%) felt slight (1,2) pain; the remainder (120 patients; 64.5%) experienced no pain after PAE. Minor adverse events were urinary frequency more severe than before PAE in 95 patients (51.1%), dysuria in 79 patients (42.5%), hematospermia in 15 patients (8.1%), hematuria in 14 patients (7.5%), rectal bleeding in 10 patients (5.4%), and groin hematoma in six patients (3.2%). These symptoms spontaneously subsided 1–2 weeks after PAE. One patient (0.5%) had a urinary tract infection that was treated with antibiotics. Nine patients (4.8%) had small skin lesions (ie, blue dots) in the glans of the penis that appeared within the first week after PAE and spontaneously subsided approximately 1 month later. The angiographic images were reviewed retrospectively, and all patients had accessory pudendal arteries with large anastomoses between the prostate and penile arteries (Fig 1). None of these patients or those in the remaining cohort experienced impotence after PAE. Overall, six patients (3.2%) felt a temporary decrease in erectile function during the first month after PAE that spontaneously subsided 6 months after PAE, including two of the nine patients that had small skin lesions in the glans of the penis. Nine of the 10 patients with acute urinary retention were able to remove the bladder catheter and void spontaneously 2–4 weeks after PAE. One patient was not able to void spontaneously and underwent open prostatectomy. Overall, 42 patients (22.6%) still had moderate to severe LUTS after PAE: 32 patients (17.2%) continued medical therapy to relieve LUTS, seven patients underwent prostate surgery (3.8%), and three patients (1.6%) underwent repeat PAE 12 months after the first procedure. The prostate arteries of these three patients were occluded 12 months after the initial PAE procedure, but collateral circulation developed (Fig 1). Of



these 42 patients without improvement after PAE, nine relapsed and had initial improvement but relapsing LUTS (21.4%), and 33 were nonresponders (78.6%) (ie, patients in whom LUTS never improved after PAE). The median follow-up time was 6 months: 118 subjects were evaluated at 6 months, and 39 were evaluated at 12 months. There was a statistically significant improvement over time in IPSS (P , .001), QoL score (P , .001), Qmax (P = .01), prostate volume (P , .001), PVR (P = .029), and PSA level (P , .001). There was also an improvement in IIEF that did not reach statistical significance (P = .065, Table 2).

Predictors of Clinical Outcome after PAE The control group consisted of 300 patients treated with PAE from 2009 to 2013 with 100–300-µm nsPVA in whom the procedure was technically successful. Median follow-up time was 24 months: 238 subjects were evaluated at 6 months, and 221 were evaluated at 12 months. Comparative baseline characteristics of the two groups are shown in Table 3. Patients in the control group were more likely to have acute urinary retention, a lower Qmax, and a better IIEF score. After PAE, when compared with the nsPVA control group, the group treated with sPVA had a greater decrease in prostate volume (P , .001), PVR (P = .005), PSA level (P = .001), and IIEF score (P = .013, Fig 2). No significant differences were found in IPSS, QoL score, and Qmax improvements. Kaplan-Meier estimates of success rates 12 months after PAE were 63.7% (95% CI: 54.9%, 71.2%) for sPVA and 70.6% (95% CI: 65.1%, 75.4%) for nsPVA (P = .08 with the log-rank test; hazard ratio, 1.31 [95% CI: 0.95, 1.82]). Multivariate analysis with Cox regression showed that the independent predictors of clinical failure were older age (hazard ratio, 1.034; 95% CI: 1.013, 1.058; P = .002), unilateral procedure (hazard ratio, 2.019; 95% CI: 1.300, 3.134; P = .002), and higher baseline IPSS (hazard ratio, 1.030; 95% CI: 1.002, 1.059; P = .033). In a Cox regression model that controlled for these predictors of failure, the adjusted

Bilhim et al

Table 3 Baseline Characteristics of Patients Who Underwent PAE with sPVA and nsPVA Variable Age (y) No. of patients with acute urinary retention* No. of patients who underwent unilateral PAE IPSS QoL score Qmax (mL/sec)* Prostate volume (cm3) PVR (mL) PSA level (ng/mL) IIEF*

sPVA (n = 186) 65.5 6 7.79 10 (5.4)

nsPVA (n = 300) 65.3 6 7.6 42 (14)

P Value .73 .005

20 (10.8)

279 (9)

.39

22.4 6 6.1 4.2 6 0.7 10.3 6 5.5 88.6 6 51.3 91.9 6 91.4 4.5 6 4.4 16.6 6 7.8

22.8 6 5.8 4.3 6 0.9 9.1 6 4.3 82.5 6 38.0 111.2 6 93.4 5.5 6 6. 6 19.7 6 7.9

.54 .21 .01 .13 .09 .08 .0001

Note.—Mean values are presented as means 6 standard deviations. Numbers in parentheses are percentages. To convert nanograms per milliliter to micrograms per liter, multiply by 1.0. * Statistically significant differences.

hazard ratio for sPVA was 1.273 (95% CI: 0.708, 1.785; P = .16; Fig 3; Table 4). Multivariate analysis with multiple regression showed that unilateral procedure was an independent predictor of lower IPSS reduction (P = .002). Multivariate analysis with multiple regression showed that the independent predictors of lower IPSS at follow-up were bilateral procedure (P = .002), acute urinary retention (P = .02), and lower baseline IPSS (P , .001). Contrast-enhanced MR images obtained after bilateral PAE were available in 23 patients in the nsPVA group and 25 patients in the sPVA group. Mean prostate volume ischemia assessed after PAE with sPVA was 7.1 cm3 6 9.5 (range, 0.2–36.3 cm3; Fig 4) and 11.6 cm3 6 19.2 (range, 0–73.6 cm3) after PAE with nsPVA (P = .16). Seven patients (30.4%) from the nsPVA group did not show any signs of ischemia, while all patients from the sPVA group had ischemic lesions in the central gland of the prostate (P , .001 with the Fisher exact test). The ratio of ischemic volume to total prostate volume was 0.11 6 0.11 (range, 0.01–0.53) and 0.10 6 0.13 (range, 0–0.4) after PAE with sPVA and nsPVA, respectively (P = .65). Lower IPSS during follow-up was associated with a higher proportion of prostate ischemia (P = .009) and also


with a greater volume of ischemia (P = .08, Fig 5). The PSA level 24 hours after PAE was available in 133 patients of the sPVA cohort. The mean 24-hour postPAE PSA level in patients who underwent bilateral PAE (n = 118) was 127.4 ng/mL 6 160.5 (127.4 mg/L 6 160.5), while it was 58.6 ng/mL 6 11 (58.6 mg/L 6 11) in those with unilateral PAE (n = 15, P = .1). Fourteen patients who underwent bilateral embolization with sPVA had both PSA measurements 24 hours after PAE and contrast-enhanced MR images after PAE. There was a positive correlation between 24-hour post-PAE PSA level and ischemic volume (r = 0.64, P = .014, Fig 6). A statistically significant association was found between higher levels of 24-hour post-PAE PSA level and lower IPSS over time (P = .01). Patients with a 24-hour post-PAE PSA level of at least 75 ng/mL (75 mg/L) had a greater decrease in IPSS than those with a 24-hour post-PAE PSA level less than 75 ng/mL (75 mg/L, P = .01). Patients with 24-hour postPAE PSA level of at least 75 ng/mL (75 mg/L) had a marginally statistically insignificant higher probability of clinical success after PAE than patients with a 24-hour post-PAE PSA level less than 75 ng/mL (75 mg/L, P = .059, Fig 7). 7


Discussion In our study, we aimed to answer two main questions: Is sPVA a safe and effective embolic choice for PAE, and which patients respond poorly to PAE? Regarding the first question, there were no major adverse events in the patients who underwent embolization with sPVA; most patients felt no pain or only slight pain during or after PAE, and all patients were treated as outpatients without any readmission due to adverse events. The minor adverse event rate and type were similar to those previously reported with nsPVA

(18), with 40%–50% of patients reporting severe urinary frequency and dysuria in the first 2 days after PAE. Hematospermia, hematuria, and rectal bleeding were reported in fewer than 10% of patients and lasted up to 2 weeks, similar to previous reports with other embolic agents (8,11,18). Up to 5% of patients had small skin lesions in the glans of the penis after PAE due to the presence of accessory pudendal arteries. These lesions were self-limited, and no cases of impotence after PAE were reported. In the presence of large anastomoses between the prostate arteries and the penile

Bilhim et al

arteries (accessory pudendal arteries), the embolization technique should be adjusted to avoid untargeted embolization of the penis, as discussed previously (14,28). Clinical outcome was not significantly different between patients who underwent PAE with sPVA and nsPVA. After PAE, 20%–36% of patients still had moderate to severe LUTS and were thus considered clinical failures. Up to 80% of patients with poor clinical outcome were nonresponders (ie, patients that never substantially improved after PAE), while the remaining 20% of patients initially responded

Figure 2

Figure 2: Graphs show the change in clinical outcomes from baseline in patients who underwent PAE with sPVA (solid lines) and nsPVA (dashed lines). Vertical lines correspond to standard errors of the differences.

8



to PAE but had LUTS relapse in the first 2 years after PAE. Improvements in outcome measures occurred mainly within the first month after PAE. Afterward, most outcome measures remained stable up to 24 months, proving the long-term potential of PAE and highlighting the fact that most clinical failures are immediate (within the first month) in nonresponders. The second goal of our study was therefore to analyze potential predictors of outcome that could help identify patients at higher risk of poor clinical outcome after PAE. Patients with acute urinary retention responded better to PAE than those without. One possible explanation is the prostatic inflammation that usually occurs with bladder catheterization, leading to arterial vasodilatation and rendering the prostate gland more susceptible to ischemia. This may be a good indication for PAE, as patients with acute urinary retention are considered worse surgical candidates (29). Older patients (ie, those older than 65 years of age) were at higher risk of having poor clinical outcome, and patients with unilateral PAE had almost twice the odds of poor clinical outcome when compared to those with bilateral PAE, as described previously (15). Older patients have more atheroscleroctic changes in the pelvic arteries and are thus more likely to undergo unilateral

PAE (15). Even with bilateral PAE, intraprostatic atherosclerosis may lead to more “hypovascular” prostates in older patients, thus rendering them less susceptible to ischemia (30). Patients with more severe LUTS were at higher risk of clinical failure, probably because the mean IPSS improvement of 9–12 points after PAE is not

Bilhim et al

enough to reduce the IPSS below 15 points, the threshold used in the present study for the definition of clinical failure. Prostate volume failed to correlate with clinical outcome, with conflicting previous reports (31,32). Patients with large prostate volumes, considered unfeasible for transurethral resection of the prostate, may

Figure 3

Figure 3: Kaplan-Meier estimates of the cumulative probability of clinical success in patients treated with sPVA (solid lines) and nsPVA (dashed lines) and in subgroups defined by prognostic factors.

Table 4 Cox Regression Analysis of Predictive Factors of Clinical Failure after PAE Univariate Analysis Variable Age Acute urinary retention Unilateral PAE sPVA Pre-PAE IPSS Pre-PAE QoL score Pre-PAE IIEF Pre-PAE prostate volume Pre-PAE PSA level Pre-PAE PVR Pre-PAE Qmax

Multivariate Analysis

Regression Coefficient

Hazard Ratio

95% CI

P Value

Regression Coefficient

Adjusted Hazard Ratio

95% CI

P Value

0.034 20.722 0.760 0.272 0.023 0.151 20.017 20.001 20.029 20.001 20.016

1.035 0.486 2.138 1.313 1.023 1.163 0.983 0.999 0+972 0.999 0.984

1.014, 1.056 0.256, 0.922 1.404, 3.256 0.945, 1.823 0.996, 1.052 0.966, 1.401 0.964, 1.003 0.995, 1.002 0.939, 1.006 0.998, 1.001 0.950, 1.020

.001 .027 ,.001 .104 .096 .11 .09 .46 .10 .59 .38

0.034 … 0.702 … 0.030 … … … … … …

1.035 … 2.019 … 1.030 … … … … … …

1.013, 1.058 … 1.300, 3.134 … 1.002, 1.059 … … … … … …

.002 … .002 … .033 … … … … … …


9


Figure 4

Figure 4: T1-weighted fat-suppressed contrast-enhanced MR images of the prostate acquired before and after PAE with sPVA. (a) Axial image acquired before embolization depicts central gland enlargement mainly due to the right lobe hypertrophy (arrow). (b) Coronal image acquired before embolization depicts central gland enlargement mainly due to right lobe hypertrophy (arrow). (c) Axial image obtained 12 days after PAE shows large ischemic zones in the right lobe (white arrow) with smaller ischemic areas in the left lobe (black arrows), probably due to the smaller enlargement of the left prostate lobe. (d) Coronal image obtained 12 days after PAE shows large ischemic zones in the right lobe (white arrow) with smaller ischemic areas in the left lobe (black arrows), probably due to the smaller enlargement of the left prostate lobe.

undergo PAE instead of open prostatectomy or use PAE as a “downsizing” technique to allow endoscopic treatment (33). The potential prognostic value of PSA level 24 hours after PAE has been shown previously (23). The 24hour post-PAE PSA level is an indirect marker of prostate necrosis due to ischemia, and patients who underwent bilateral PAE had mean values higher than those with unilateral PAE, reinforcing the importance of bilateral PAE and leading to greater prostate 10

necrosis than unilateral PAE. There was a significant positive correlation between ischemic prostate tissue on contrast-enhanced MR images and PSA level 24 hours after PAE, proving the link between ischemia, prostate necrosis, and PSA release into the bloodstream. Despite bilateral PAE, only a mean 10% of prostate volume ischemia was detected with contrastenhanced MR imaging in the first month after embolization, with a wide range of variability between patients. There were no significant differences

Bilhim et al

in prostate ischemic volumes after PAE with sPVA and nsPVA. In our study, higher volumes of prostate ischemia were used to predict greater decreases in IPSS after PAE, contrary to previously reported findings (24). Prostate ischemia detected with MR imaging tends to disappear after the first month after PAE (24,25), and in our study, it was assessed 2–4 weeks after PAE. In future studies conducted to evaluate the potential role of prostate ischemia, investigators should perform MR imaging in the first 24 hours after PAE and 1–2 weeks after PAE to assess ischemia involution. Our study has several limitations. It was not a prospective randomized study conducted to compare different embolic agents or particle sizes for PAE; the two cohorts analyzed herein were not similar at baseline, and follow-up data were not the same for the two cohorts. Adjusted analyses were performed to compensate for potential baseline confounders. Patients who underwent embolization with nsPVA received embolic particles of different sizes that ranged between 100 and 300 µm, while all patients in the sPVA group received 300–500-µm particles, which could induce bias in the outcome measurements after PAE due to the differences in embolic particle sizes. Future prospective trials conducted to compare different embolic agents and particle sizes should be performed to assess the best embolic choice for PAE by using prostate ischemia detected with MR imaging. The median lobe was not evaluated as a potential risk factor of poor outcome after PAE. MR images of the prostate and 24-hour post-PAE PSA levels were not available in all patients, which may lead to selection bias. In conclusion, PAE with sPVA and nsPVA induces similar clinical relief that is sustained up to 2 years after PAE for patients with benign prostatic hyperplasia. Younger patients (up to 65 years of age), lower baseline IPSS, acute urinary retention, and bilateral PAE are predictors of better clinical outcome. PAE induces 10% of prostate gland ischemia measured with MR



Figure 5

Bilhim et al

4. Michel MC, Mehlburger L, Bressel HU, Schumacher H, Schäfers RF, Goepel M. Tamsulosin treatment of 19,365 patients with lower urinary tract symptoms: does comorbidity alter tolerability? J Urol 1998;160 (3 Pt 1):784–791. 5. McConnell JD, Bruskewitz R, Walsh P, et al. The effect of finasteride on the risk of acute urinary retention and the need for surgical treatment among men with benign prostatic hyperplasia. Finasteride Long-Term Efficacy and Safety Study Group. N Engl J Med 1998; 338(9):557–563. 6. Varkarakis J, Bartsch G, Horninger W. Long-term morbidity and mortality of transurethral prostatectomy: a 10-year follow-up. Prostate 2004;58(3):248–251.

Figure 5: Graph shows the IPSS change from baseline after PAE between subgroups defined by the median proportion of prostate ischemia: at least 7% (solid line) and less than 7% (dashed line) (P = .009). Vertical lines correspond to standard errors of the differences.

Figure 6

Figure 7

7. Baazeem A, Elhilali MM. Surgical management of benign prostatic hyperplasia: current evidence. Nat Clin Pract Urol 2008;5(10): 540–549. 8. Carnevale FC, da Motta-Leal-Filho JM, Antunes AA, et al. Quality of life and clinical symptom improvement support prostatic artery embolization for patients with acute urinary retention caused by benign prostatic hyperplasia. J Vasc Interv Radiol 2013;24(4): 535–542. 9. Pisco JM, Pinheiro LC, Bilhim T, Duarte M, Mendes JR, Oliveira AG. Prostatic arterial embolization to treat benign prostatic hyperplasia. J Vasc Interv Radiol 2011;22(1): 11–19; quiz 20.

Figure 6: Correlation analysis between logtransformed PSA level 24 hours after PAE and prostate ischemic volume detected with contrastenhanced MR imaging (r = 0.64, P = .014).

imaging in the first month after embolization. The 24-hour post-PAE PSA level correlates with prostate ischemia on MR images, and both correlate with clinical outcome. Disclosures of Conflicts of Interest: T.B. disclosed no relevant relationships. J.P. disclosed no relevant relationships. J.A.P. disclosed no relevant relationships. N.V.C. disclosed no relevant relationships. L.F. disclosed no relevant relationships. L.C.P. disclosed no relevant relationships. M.D. disclosed no relevant relationships. A.G.O. disclosed no relevant relationships.

Figure 7: Kaplan-Meier estimates of the probability of clinical success after PAE between subgroups defined by median PSA level 24 hours after PAE (hazard ratio, 0.967 [95% CI: 0.993, 1.001]; P = .059).

References 1. Levy A, Samraj GP. Benign prostatic hyperplasia: when to ‘watch and wait,’ when and how to treat. Cleve Clin J Med 2007;74(Suppl 3):S15–S20. 2. Garraway WM, Collins GN, Lee RJ. High prevalence of benign prostatic hypertrophy in the community. Lancet 1991;338(8765): 469–471. 3. Burnett AL, Wein AJ. Benign prostatic hyperplasia in primary care: what you need to know. J Urol 2006;175(3 Pt 2):S19– S24.


10. Pisco J, Campos Pinheiro L, Bilhim T, et al. Prostatic arterial embolization for benign prostatic hyperplasia: short- and intermediate-term results. Radiology 2013;266(2): 668–677. 11. Pisco JM, Rio Tinto H, Campos Pinheiro L, et al. Embolisation of prostatic arteries as treatment of moderate to severe lower urinary symptoms (LUTS) secondary to benign hyperplasia: results of short- and mid-term follow-up. Eur Radiol 2013;23(9):2561– 2572. 12. Bilhim T, Pisco JM, Furtado A, et al. Prostatic arterial supply: demonstration by multirow detector angio CT and catheter angiography. Eur Radiol 2011;21(5):1119–1126. 13. Bilhim T, Casal D, Furtado A, Pais D, O’Neill JE, Pisco JM. Branching patterns of the male internal iliac artery: imaging findings. Surg Radiol Anat 2011;33(2):151–159. 14. Bilhim T, Pisco JM, Rio Tinto H, et al. Prostatic arterial supply: anatomic and imaging findings relevant for selective arterial embolization. J Vasc Interv Radiol 2012; 23(11):1403–1415.

11


15. Bilhim T, Pisco J, Rio Tinto H, et al. Unilateral versus bilateral prostatic arterial embolization for lower urinary tract symptoms in patients with prostate enlargement. Cardiovasc Intervent Radiol 2013;36(2):403–411.

21. Kurbatov D, Russo GI, Lepetukhin A, et al. Prostatic artery embolization for prostate volume greater than 80 cm3: results from a single-center prospective study. Urology 2014;84(2):400–404.

16. Li Q, Duan F, Wang MQ, Zhang GD, Yuan K. Prostatic arterial embolization with small sized particles for the treatment of lower urinary tract symptoms due to large benign prostatic hyperplasia: preliminary results. Chin Med J (Engl) 2015;128(15):2072–2077.

22. Grosso M, Balderi A, Arnò M, et al. Prostatic artery embolization in benign prostatic hyperplasia: preliminary results in 13 patients. Radiol Med (Torino) 2015;120(4):361–368.

17. Bagla S, Martin CP, van Breda A, et al. Early results from a United States trial of prostatic artery embolization in the treatment of benign prostatic hyperplasia. J Vasc Interv Radiol 2014;25(1):47–52. 18. Bilhim T, Pisco J, Campos Pinheiro L, et al. Does polyvinyl alcohol particle size change the outcome of prostatic arterial embolization for benign prostatic hyperplasia? Results from a single-center randomized prospective study. J Vasc Interv Radiol 2013;24(11):1595–1602.e1. 19. Gao YA, Huang Y, Zhang R, et al. Benign prostatic hyperplasia: prostatic arterial embolization versus transurethral resection of the prostate—a prospective, randomized, and controlled clinical trial. Radiology 2014;270(3):920–928. 20. Somani BK, Hacking N, Bryant T, et al. Prostate artery embolization (PAE) for benign prostatic hyperplasia (BPH). BJU Int 2014;114(5):639–640.

12

23. de Assis AM, Moreira AM, de Paula Ro drigues VC, et al. Prostatic artery embolization for treatment of benign prostatic hyperplasia in patients with prostates . 90 g: a prospective single-center study. J Vasc Interv Radiol 2015;26(1):87–93. 24. Frenk NE, Baroni RH, Carnevale FC, et al. MRI findings after prostatic artery embolization for treatment of benign hyperplasia. AJR Am J Roentgenol 2014;203(4):813–821. 25. Brook OR, Faintuch S, Brook A, Goldberg SN, Rofsky NM, Lenkinski RE. Embolization therapy for benign prostatic hyperplasia: influence of embolization particle size on gland perfusion. J Magn Reson Imaging 2013;38(2):380–387. 26. Angle JF, Siddiqi NH, Wallace MJ, et al. Quality improvement guidelines for percutaneous transcatheter embolization: Society of Interventional Radiology Standards of Practice Committee. J Vasc Interv Radiol 2010;21(10):1479–1486. 27. Sacks D, McClenny TE, Cardella JF, Lewis CA. Society of Interventional Radiology clin-

Bilhim et al

ical practice guidelines. J Vasc Interv Radiol 2003;14(9 Pt 2):S199–S202. 28. Bilhim T, Pereira JA, Fernandes L, Rio Tinto H, Pisco JM. Angiographic anatomy of the male pelvic arteries. AJR Am J Roentgenol 2014;203(4):W373–W382. 29. Yoon PD, Chalasani V, Woo HH. Systematic review and meta-analysis on management of acute urinary retention. Prostate Cancer Prostatic Dis 2015;18(4):297–302. 30. Kozlowski R, Kershen RT, Siroky MB, Krane RJ, Azadzoi KM. Chronic ischemia alters prostate structure and reactivity in rabbits. J Urol 2001;165(3):1019–1026. 31. Wang M, Guo L, Duan F, et al. Prostatic arterial embolization for the treatment of lower urinary tract symptoms caused by benign prostatic hyperplasia: a comparative study of medium- and large-volume prostates. BJU Int 2016;117(1):155–164. 32. Bagla S, Smirniotopoulos JB, Orlando JC, van Breda A, Vadlamudi V. Comparative analysis of prostate volume as a predictor of outcome in prostate artery embolization. J Vasc Interv Radiol 2015;26(12):1832–1838. 33. Nejmark AI, Nejmark BA, Tachalov MA, Arzamascev DD, Torbik DV. Superselective prostatic artery embolization as a preparatory step before TURP in the treatment of benign prostatic hyperplasia in patients with large prostates [in Russian]. Urologiia 2015;(2):60–62, 64.
