St Gallen PAE Trial: Prostatic Artery Embolisation versus TURP for Benign Prostatic Hyperplasia
The St Gallen PAE trial was a single-center, open-label noninferiority randomized trial of prostatic artery embolisation (PAE) versus monopolar transurethral resection of the prostate (TURP) in 103 men with medically refractory BPH-related lower urinary tract symptoms, with the change in IPSS at 12 weeks as the primary endpoint. Mean IPSS fell by 9.23 points after PAE and 10.77 points after TURP; noninferiority was not shown (difference 1.54 points, 95% CI −1.45 to 4.52, margin 3 points), and TURP gave larger gains in flow rate, residual volume, prostate volume and relief of obstruction.[1]
- Design and population
- Single-center, open-label noninferiority randomized trial. Men aged 40 or older with medically refractory BPH-related LUTS (IPSS 8 or higher, Qmax below 12 mL/s or retention) and a 25 to 80 mL prostate who were candidates for TURP; 99 of 103 randomized were treated.
- Arms
- Prostatic artery embolisation (PAE) with 250–400 µm microspheres versus monopolar transurethral resection of the prostate (TURP)
- Report and follow-up
- 2018 · 12 weeks (primary endpoint); per-protocol analysis of 48 PAE and 51 TURP patients
| Endpoint | Results | Effect | Finding |
|---|---|---|---|
| Mean change in IPSS from baseline to 12 weeks, per protocol (PAE minus TURP; positive favors TURP) | −9.23 points vs −10.77 points | difference 1.54; CI -1.45–4.52; P 0.17 (noninferiority); 0.31 (superiority) | Not met |
Harms: Treatment-related adverse events, total events in first 12 weeks: 36 events vs 70 events; At least one treatment-related adverse event: 62.5% vs 70.6%; Two or more treatment-related adverse events: 16.7% vs 45.1%
Design
The investigator-initiated trial ran in the urology and radiology departments of St Gallen Cantonal Hospital, Switzerland, and was funded by the hospital's research committee without commercial support. Men were randomized between 11 February 2014 and 24 May 2017. Eligibility required age of 40 years or older, an indication for TURP, failure of or unwillingness to continue medical treatment, a prostate of 25 to 80 mL on transabdominal ultrasound, IPSS of at least 8 with a quality-of-life score of at least 3, and a maximum flow rate below 12 mL/s or urinary retention. Exclusions included severe atherosclerosis, aneurysm or severe tortuosity of the aortic bifurcation or internal iliac arteries, acontractile detrusor, neurogenic bladder dysfunction, urethral stenosis, bladder diverticulum or stone, contrast allergy, known prostate cancer and a glomerular filtration rate below 60 mL/min.[1]
Allocation used software-based minimisation, stratified by age (under or over 70 years) and prostate volume (under or over 50 mL). The primary endpoint was the change in IPSS from baseline to 12 weeks. A between-group difference of less than 3 points defined noninferiority for PAE, because a 3-point change is perceived by patients as only a slight improvement; the test was one-sided at an alpha of 0.025. The sample size of 38 patients per group assumed a standard deviation of 4.6 points, and 100 patients were targeted to allow for dropout. Analysis was per protocol. Secondary endpoints, including flow rate, residual volume, MRI prostate volume, pressure-flow studies and adverse events (modified Clavien and CTCAE), were compared with two-sided superiority tests.[1]
Results
Of 144 men assessed, 103 were randomized. Two men allocated to PAE and one to TURP declined surgery, and PAE was not technically possible in one man, so 48 PAE and 51 TURP patients were analyzed; all attended the 12-week visit. IPSS fell by 9.23 points after PAE and 10.77 points after TURP. The difference of 1.54 points was below the 3-point margin, but its upper confidence limit of 4.52 was not, and noninferiority could not be established (P = 0.17; superiority P = 0.31). The standard deviation of the IPSS change was 7.5 points, against 4.6 assumed in the sample size calculation. A baseline-adjusted analysis gave a difference of 2.87 points (95% CI 0.53 to 5.21; noninferiority P = 0.46).[1]
Patient-reported secondary outcomes did not differ at 12 weeks: IPSS quality-of-life score improved by 2.33 versus 2.69 points (P = 0.15), and bladder diary measures did not differ significantly. Functional outcomes favored TURP. Qmax rose by 5.19 mL/s after PAE and 15.34 mL/s after TURP (P < 0.001), postvoid residual fell by 86.36 versus 199.98 mL (P = 0.003), and MRI prostate volume fell by 12.17 versus 30.27 mL (P < 0.001). On pressure-flow studies, detrusor pressure at maximum flow fell by 17.17 versus 41.07 cmH2O (P = 0.002), and 56% of PAE patients versus 93% of TURP patients moved to a less obstructive category (P = 0.003).[1]
At 24 months, 34 of 48 PAE and 47 of 51 TURP patients were evaluable. IPSS fell by 9.21 points after PAE and 12.09 points after TURP (difference 2.88 points, 95% CI 0.04 to 5.72; P = 0.047), a protocol-defined secondary comparison that was not corrected for multiple testing. Qmax rose by 3.9 versus 10.23 mL/s (P < 0.001), postvoid residual fell by 62.1 versus 204.0 mL (P = 0.005) and MRI prostate volume fell by 10.66 versus 30.20 mL (P = 0.005). IPSS quality of life improved by 2.26 versus 3.26 points (P = 0.002), and IIEF-5 fell by 1.59 versus 2.44 points (P = 0.551). Because patients not evaluable at 24 months, including the 10 PAE patients who had TURP, were left out of that analysis, the authors reanalyzed IPSS with the last available value carried forward; the reductions were 7.29 versus 12.53 points (difference 5.24 points, 95% CI 2.23 to 8.19), an exploratory analysis. Urodynamics were declined by 42 of 81 evaluable patients, and among those tested 50% (10/20) of PAE and 82% (14/17) of TURP patients moved to a less obstructive category (P = 0.087).[2]
At 5 years, 18 of 48 PAE and 38 of 51 TURP patients reached the 60-month visit. IPSS fell by 7.78 points after PAE and 11.57 points after TURP (difference 3.79 points, 95% CI −0.66 to 8.24; P = 0.092), Qmax rose by 3.59 versus 9.30 mL/s (P = 0.027) and postvoid residual fell by 27.81 versus 219.97 mL (P = 0.001).[3] The EAU guideline, citing this report, states that 42% of patients initially treated with PAE underwent TURP within five years.[4]
A post hoc analysis of the 48 PAE patients related baseline anatomy to 12-week response. Larger total prostate volume (Spearman rs −0.35 against relative IPSS change, P = 0.01) and larger central gland volume (rs −0.34, P = 0.02) were associated with greater IPSS improvement, and larger volumes were also associated with greater Qmax gains (rs 0.31 and 0.39). Receiver-operating-characteristic analysis suggested volume thresholds of 39 mL for the total prostate and 38 mL for the central gland to predict IPSS success.[5]
Surgery and treatment delivery
A single experienced interventional radiologist performed all embolizations. A 16 F urethral catheter was placed before the procedure, the right common femoral artery was accessed under local anesthesia, and selective internal iliac arteriography with a 5 F catheter mapped the prostatic arteries. Prostatic arteries were catheterized with 1.9 to 2.4 F microcatheters and embolized with 250 to 400 µm Embozene microspheres, bilaterally if possible, with cone beam CT in patients with challenging anatomy. Embolization was considered successful when the normal prostatic blush was lost and angiography showed complete stasis of flow in the prostatic arteries; this was achieved bilaterally in 36 of 48 patients (75.0%) and unilaterally in 12 (25.0%). The catheter was removed the morning after.[1]
Four experienced study physicians performed TURP under spinal anesthesia in 25 patients (49%) or general anesthesia in 26 (51%), using a 24 F monopolar resectoscope with 180 W cutting and 60 W coagulation power, a tungsten loop and mannitol-sorbitol irrigation. A 20 F three-way catheter stayed in for at least 2 days. Resection took 58.25 minutes and removed 25.20 g of tissue.[1]
Procedure time was 122.2 minutes for PAE and 69.5 minutes for TURP (P < 0.001). PAE fluoroscopy time was 50.8 minutes, the dose-area product 176.5 Gy·cm² and the embolic volume 1.0 mL. Hemoglobin fell by 4.3 g/L after PAE and 13.8 g/L after TURP, the catheter stayed in for 1.3 versus 3.3 days, and hospital stay was 2.2 versus 4.2 days (each P = 0.001).[1] Prostatic medications were stopped on the day of TURP and 2 weeks after PAE; all patients received antibiotic prophylaxis until catheter removal or 3 days at the latest, and diclofenac and pantoprazole for one week.[1]
Adverse events
Through 12 weeks there were 36 treatment-related adverse events after PAE and 70 after TURP (P = 0.003). The share of patients with at least one event was similar (62.5% versus 70.6%), but fewer PAE patients had two or more (16.7% versus 45.1%).[1] Severity classes were distributed similarly: Clavien grades 1, 2 and 3 made up 71.4%, 22.9% and 5.7% of events after PAE and 64.5%, 24.2% and 11.3% after TURP.[1] Over 24 months the totals were 43 and 78 events, with only 7 and 8 events between months 3 and 24.[1][2]
Postoperative pain during the hospital stay was reported by more PAE patients (56.3% versus 31.9%; P = 0.03), and severe pain (visual analog scale 6 or higher) by 18.8% versus 4.3%, confined to the first 12 to 24 hours.[1] Procedure-specific complications after PAE were one moderate puncture-site hematoma, one postembolization syndrome with fever lasting under 24 hours, and one area of bladder-wall ischemia that required transurethral removal of necrotic tissue; one patient reported de novo erectile dysfunction.[1] At 12 weeks, among patients with assessable ejaculation, ejaculatory disorders occurred in 56% (14/25) after PAE and 84% (21/25) after TURP, and complete anejaculation in 16% (4/25) versus 52% (13/25).[1][2][7]
Ten PAE patients (21%) had TURP within 2 years for unsatisfying outcomes, five of whom had unilateral embolization; no criteria for further treatment were predefined. Another surgical procedure was needed in 1 PAE patient (2%) and 5 TURP patients (10%), and 2 TURP patients (4.3%) but no PAE patients took medication for LUTS at 24 months.[2]
Limitations
Patients and physicians could not be blinded, and one interventional radiologist performed every embolization at a single center. The sample was too small to establish or exclude noninferiority, partly because the variability of the IPSS change exceeded the assumption used for power. The primary endpoint was measured at 12 weeks, and the analysis was per protocol, excluding four randomized patients. At 24 months, 14 of 48 PAE patients were not evaluable, and the 10 who had TURP were analyzed as dropouts, so the 24-month PAE results may overestimate its effect; the 24-month secondary analyses were exploratory and not corrected for multiple testing, and many patients declined urodynamics. At 5 years, only 18 of 48 PAE and 38 of 51 TURP patients reached the final visit. PAE was considered angiographically successful when normal prostatic blush disappeared and flow in the prostatic arteries ceased; bilateral embolization was performed in 36/48 patients (75%) and unilateral embolization in 12/48 (25%).[1][2][3]
In conversation
WATER, which compared Aquablation with TURP, blinded patients and assessors and set its noninferiority margin at 4.7 IPSS points; this trial was open-label with a margin of 3 points and a 12-week primary endpoint, so the two noninferiority conclusions rest on different margins and follow-up.
Guidelines and approvals
- AUA (2026). Clinicians may offer PAE with particle embolics as an option for patients with prostates of 50 cc or more for LUTS/BPH . The guideline text states that comparative PAE versus TURP studies, including the 2-year report of this trial, showed similar or inferior symptom improvement with PAE. Urologists should assess patients considering PAE beforehand , and PAE should be performed by interventional radiologists specifically trained in it .[6]
- EAU (March 2026). Offer PAE to men with moderate-to-severe LUTS and benign prostatic obstruction who wish to consider minimally invasive options and accept less-optimal outcomes than with TURP ; perform PAE only in units where urologists work with trained interventional radiologists for work-up and follow-up . The guideline marks PAE as still under investigation and rates the evidence that PAE is less effective than TURP and has higher retreatment rates at level 1a.[4]
Report history
| Published | Report | Cutoff | Headline results | Citation |
|---|---|---|---|---|
| 2018-06-19 | Primary report (12-week outcomes) · primary | — | Mean change in IPSS from baseline to 12 weeks, per protocol (PAE minus TURP; positive favors TURP): −9.23 points vs −10.77 points; difference 1.54; CI -1.45–4.52; P 0.17 (noninferiority); 0.31 (superiority) Not met Mean change in IPSS quality-of-life score at 12 weeks (PAE minus TURP): −2.33 points vs −2.69 points; difference 0.35; CI -0.3–1; P 0.15 Mean change in Qmax at 12 weeks (PAE minus TURP): +5.19 mL/s vs +15.34 mL/s; difference -10.15; CI -14.67–-5.63; P <0.001 Mean change in postvoid residual volume at 12 weeks (PAE minus TURP): −86.36 mL vs −199.98 mL; difference 113.62; CI 39.25–187.98; P 0.003 Mean change in MRI prostate volume at 12 weeks (PAE minus TURP): −12.17 mL vs −30.27 mL; difference 18.11; CI 10.11–26.1; P <0.001 Mean change in detrusor pressure at Qmax at 12 weeks (PAE minus TURP): −17.17 cmH2O vs −41.07 cmH2O; difference 23.91; CI 9.09–38.72; P 0.002 Proportion shifting to a less obstructive ICS category on pressure-flow studies at 12 weeks: 56% vs 93%; P 0.003 Mean change in IIEF-5 at 12 weeks (PAE minus TURP): −0.98 points vs −1.84 points; difference 0.87; CI -1.89–3.63; P 0.53 | Abt D, Hechelhammer L, Müllhaupt G, et al. "Comparison of prostatic artery embolisation (PAE) versus transurethral resection of the prostate (TURP) for benign prostatic hyperplasia: randomised, open label, non-inferiority trial." *BMJ.* 2018;361:k2338. doi:10.1136/bmj.k2338 DOI |
| 2018-12-17 | Outcome prediction after PAE (post hoc) · post-hoc | — | Spearman correlation of total prostate volume with relative IPSS change: rs −0.35 vs Not applicable; P 0.01 Spearman correlation of central gland volume with relative IPSS change: rs −0.34 vs Not applicable; P 0.02 Spearman correlation of total prostate volume with Qmax change: rs 0.31 vs Not applicable; P 0.05 Spearman correlation of central gland volume with Qmax change: rs 0.39 vs Not applicable; P 0.01 Optimal volume thresholds for predicting IPSS success on ROC analysis: 39 mL (total prostate); 38 mL (central gland) vs Not applicable; — | Abt D, Müllhaupt G, Mordasini L, et al. "Outcome prediction of prostatic artery embolization: post hoc analysis of a randomized, open-label, non-inferiority trial." *BJU Int.* 2019;124(1):134-144. doi:10.1111/bju.14632 DOI |
| 2021-02-19 | 2-year follow-up · updated | — | Mean IPSS reduction from baseline to 24 months (PAE minus TURP; positive favors TURP): −9.21 points vs −12.09 points; difference 2.88; CI 0.04–5.72; P 0.047 Mean IPSS reduction to 24 months with last observation carried forward (PAE minus TURP): −7.29 points vs −12.53 points; difference 5.24; CI 2.23–8.19; P 0.001 Mean improvement in IPSS quality-of-life score at 24 months (TURP improvement minus PAE improvement): 2.26 points vs 3.26 points; difference 0.99; CI 0.29–1.69; P 0.002 Mean improvement in Qmax at 24 months (PAE minus TURP): +3.9 mL/s vs +10.23 mL/s; difference -6.33; CI -10.12–-2.54; P <0.001 Mean reduction in postvoid residual at 24 months (TURP reduction minus PAE reduction): 62.1 mL vs 204.0 mL; difference 141.91; CI 43.31–240.51; P 0.005 Mean reduction in MRI prostate volume at 24 months (TURP reduction minus PAE reduction): 10.66 mL vs 30.20 mL; difference 19.54; CI 7.7–31.38; P 0.005 Mean reduction in detrusor pressure at Qmax at 24 months (TURP reduction minus PAE reduction): 13.05 cmH2O vs 37.12 cmH2O; difference 24.07; CI 1.93–46.2; P 0.034 Proportion shifting to a less obstructive ICS category from baseline to 24 months: 50% (10/20) vs 82% (14/17); P 0.087 Mean decrease in IIEF-5 at 24 months (TURP decrease minus PAE decrease; positive favors PAE): 1.59 points vs 2.44 points; difference 0.85; CI -1.99–3.69; P 0.551 | Abt D, Müllhaupt G, Hechelhammer L, et al. "Prostatic Artery Embolisation Versus Transurethral Resection of the Prostate for Benign Prostatic Hyperplasia: 2-yr Outcomes of a Randomised, Open-label, Single-centre Trial." *Eur Urol.* 2021;80(1):34-42. doi:10.1016/j.eururo.2021.02.008 DOI |
| 2024-03-25 | 5-year follow-up · updated | — | Mean IPSS reduction from baseline to 5 years (PAE minus TURP; positive favors TURP): −7.78 points vs −11.57 points; difference 3.79; CI -0.66–8.24; P 0.092 Mean improvement in Qmax at 5 years (PAE minus TURP): +3.59 mL/s vs +9.30 mL/s; difference -5.71; CI -10.72–-0.7; P 0.027 Mean reduction in postvoid residual at 5 years (TURP reduction minus PAE reduction): 27.81 mL vs 219.97 mL; difference 192.15; CI 83.79–300.51; P 0.001 | Müllhaupt G, Hechelhammer L, Graf N, et al. "Prostatic Artery Embolisation Versus Transurethral Resection of the Prostate for Benign Prostatic Obstruction: 5-year Outcomes of a Randomised, Open-label, Noninferiority Trial." *Eur Urol Focus.* 2024;10(5):788-795. doi:10.1016/j.euf.2024.03.001 DOI |
References
1. Abt D, Hechelhammer L, Müllhaupt G, et al. "Comparison of prostatic artery embolisation (PAE) versus transurethral resection of the prostate (TURP) for benign prostatic hyperplasia: randomised, open label, non-inferiority trial." BMJ. 2018;361:k2338. doi:10.1136/bmj.k2338
2. Abt D, Müllhaupt G, Hechelhammer L, et al. "Prostatic Artery Embolisation Versus Transurethral Resection of the Prostate for Benign Prostatic Hyperplasia: 2-yr Outcomes of a Randomised, Open-label, Single-centre Trial." Eur Urol. 2021;80(1):34-42. doi:10.1016/j.eururo.2021.02.008
3. Müllhaupt G, Hechelhammer L, Graf N, et al. "Prostatic Artery Embolisation Versus Transurethral Resection of the Prostate for Benign Prostatic Obstruction: 5-year Outcomes of a Randomised, Open-label, Noninferiority Trial." Eur Urol Focus. 2024;10(5):788-795. doi:10.1016/j.euf.2024.03.001
4. Cornu JN, Elterman D, Hashim H, et al. EAU Guidelines on Non-Neurogenic Male Lower Urinary Tract Symptoms (LUTS). Arnhem: EAU Guidelines Office; March 2026. Section 5.3.4.b.
5. Abt D, Müllhaupt G, Mordasini L, et al. "Outcome prediction of prostatic artery embolization: post hoc analysis of a randomized, open-label, non-inferiority trial." BJU Int. 2019;124(1):134-144. doi:10.1111/bju.14632
6. Goueli R, Badlani GH, Welliver C, et al. Management of Lower Urinary Tract Symptoms Attributed to Benign Prostatic Hyperplasia: AUA Guideline. American Urological Association; approved April 2026. Statements 51 to 53.
7. Müllhaupt G, et al. "Ejaculatory disorders after prostatic artery embolization: a reassessment of two prospective clinical trials." World J Urol. 2020;38(10):2595-2599. doi:10.1007/s00345-019-03036-7