IMPRESS: Collagenase Clostridium Histolyticum for Peyronie's Disease in Two Phase 3 Trials

Bottom line Positive

In two placebo-controlled trials of stable Peyronie's disease with 30° to 90° curvature, intralesional collagenase clostridium histolyticum with plaque modeling reduced curvature and bother more than placebo with modeling at 52 weeks (pooled curvature 34% versus 18.2%); three corporal ruptures needed surgical repair.

Both co-primary endpoints were met in each trial; the pooled net curvature difference was 7.7°.

IMPRESS I and II (Investigation for Maximal Peyronie's Reduction Efficacy and Safety Studies) were identical double-blind, placebo-controlled phase 3 trials of intralesional collagenase clostridium histolyticum (CCH) with plaque modeling versus placebo injections with modeling in 832 men with stable Peyronie's disease and 30° to 90° of curvature, with percent change in curvature and the Peyronie's Disease Questionnaire (PDQ) bother score at 52 weeks as co-primary endpoints. Each trial met both endpoints; in the pooled post hoc analysis, curvature fell by a mean 34% (17.0°) with CCH and 18.2% (9.3°) with placebo.^[1][2]

Design and population

Two identical double-blind, placebo-controlled randomized phase 3 trials. 832 men aged 18 or older with stable Peyronie's disease of at least 12 months and dorsal or lateral curvature of 30° to 90°.

Arms

Intralesional collagenase clostridium histolyticum (0.58 mg per injection) with plaque modeling versus placebo injections with plaque modeling

Report and follow-up

2013 · 52 weeks: up to 4 treatment cycles at weeks 0, 6, 12 and 18, then non-treatment follow-up to week 52

Endpoint	Results	Effect	Finding
Mean percent change in curvature from baseline to week 52 (adjusted, last observation carried forward, mITT 199 versus 104); percentage-point difference (CCH minus placebo)	−35.0% vs −17.8%	difference -17.2; CI -26.7–-7.6; P <0.01	Met
Mean percent change in curvature from baseline to week 52 (adjusted, last observation carried forward, mITT 202 versus 107); percentage-point difference (CCH minus placebo)	−33.2% vs −21.8%	difference -11.4; CI -19.5–-3.3; P <0.01	Met
Mean change in Peyronie's Disease Questionnaire bother domain score (0 to 16, higher is worse) from baseline to week 52 (mITT 199 versus 104); difference (CCH minus placebo)	−2.8 vs −1.6	difference -1.2; CI -2.4–-0.03; P <0.05	Met
Mean change in Peyronie's Disease Questionnaire bother domain score (0 to 16, higher is worse) from baseline to week 52 (mITT 202 versus 107); difference (CCH minus placebo)	−2.6 vs −1.5	difference -1.1; CI -2.1–-0.002; P <0.05	Met

Harms: Penile hematoma: 65.5% vs 19.2%; Penile swelling: 55.0% vs 3.2%; Penile pain: 45.4% vs 9.3%

NCT01221597 · Report DOI

Design

Men were 18 or older, in a stable relationship and willing to have vaginal intercourse, with Peyronie's symptoms for at least 12 months, stable disease by investigator assessment, and dorsal, lateral or dorsal/lateral curvature of 30° to 90° in a single measurable plane. Exclusions included ventral curvature, an isolated hourglass deformity, a calcified plaque that would prevent injection (non-contiguous stippling was allowed), a plaque near the base of the penis, prior Peyronie's surgery, erectile dysfunction unresponsive to phosphodiesterase type 5 inhibitors, clinically significant abnormal penile hemodynamics on duplex ultrasound, and anticoagulant use in the 7 days before each dose (aspirin up to 165 mg daily was allowed). Intralesional, oral or topical Peyronie's therapy within 3 months, shock wave therapy within 6 months and mechanical devices within 2 weeks were also exclusions. Curvature was measured after an erection induced with prostaglandin E1 or trimix.^[3]

Men were stratified by baseline curvature (30° to 60° versus 61° to 90°) and randomized 2:1 to CCH or placebo: 277 versus 140 in IMPRESS I and 274 versus 141 in IMPRESS II. Mean age was 57.3 to 58.2 years across the four groups; mean baseline curvature was 48.8° and 49.0° (IMPRESS I, CCH and placebo) and 51.3° and 49.6° (IMPRESS II), and a history of erectile dysfunction was recorded in 46.2% and 48.9% of CCH recipients and 53.6% and 53.9% of placebo recipients (IMPRESS I and II).^[1][2] The registry lists 34 sites per trial, 29 in the United States and 5 in Australia.^[3]

The efficacy population (modified intention to treat, 612 of 832 men) required a baseline and at least one later curvature and PDQ assessment, and vaginal intercourse within 3 months before each PDQ assessment. Curvature was analyzed with analysis of variance and last observation carried forward. Both arms had plaque modeling; there was no sham-only group.^[2][4] Approximately one-third of men did not complete the PDQ at one or both time points.^[4]

Results

In IMPRESS I, mean curvature changed by −35.0% with CCH and −17.8% with placebo (difference −17.2 percentage points, 95% CI −26.7 to −7.6; P < 0.01); in IMPRESS II the figures were −33.2% and −21.8% (difference −11.4, 95% CI −19.5 to −3.3; P < 0.01). The bother domain score (0 to 16, higher is worse) fell by 2.8 versus 1.6 points in IMPRESS I (difference −1.2, 95% CI −2.4 to −0.03; P < 0.05) and by 2.6 versus 1.5 in IMPRESS II (difference −1.1, 95% CI −2.1 to −0.002; P < 0.05). The lower confidence limits for bother lie close to zero. Improvement in curvature was numerically similar in the 30° to 60° and 61° to 90° strata, and the label records no clinically meaningful difference by baseline erectile dysfunction severity or concomitant phosphodiesterase type 5 inhibitor use.^[2]

In the post hoc pooled analysis, mean curvature improved by 34% (−17.0 ± 14.8°) with CCH and 18.2% (−9.3 ± 13.6°) with placebo (P < 0.0001), and the mean change in bother score was −2.8 ± 3.8 versus −1.8 ± 3.5 (P = 0.0037).^[1] The mean difference between arms was 7.7°.^[1] In the combined analysis the International Index of Erectile Function (IIEF) overall satisfaction score rose from 5.6 to 6.6 with CCH and from 5.6 to 6.0 with placebo; the difference was not statistically significant when the trials were analyzed separately.^[4]

A pooled subgroup analysis (n = 612) found reductions in curvature and bother in every subgroup. Curvature reductions were significantly greater with CCH than placebo for baseline curvature of 30° to 60° (n = 492) and 61° to 90° (n = 120), disease duration of more than 2 to 4 years (n = 212) and more than 4 years (n = 199), no plaque calcification (n = 447), and IIEF erectile function of 17 or higher (n = 480). The differences were not significant for disease duration of 1 to 2 years (n = 201), non-contiguous stippling (n = 103), contiguous calcification that did not interfere with injection (n = 62), or IIEF 1 to 5 (n = 22) and 6 to 16 (n = 106). Bother reductions were significantly greater with CCH for curvature of 30° to 60°, duration over 4 years, no calcification, and IIEF 1 to 5 and 17 or higher.^[5]

A post hoc analysis of all 832 men found that 29.9% of CCH recipients had a reduction of at least 20% from baseline curvature after one cycle. Among those who had not responded earlier, 60.8% of cycle 1 nonresponders, 42.7% of cycle 1 to 2 nonresponders and 23.5% of cycle 1 to 3 nonresponders reached that threshold by cycle 4. Mean cumulative percent reduction was greater with CCH than with placebo after every cycle (P < 0.001).^[6]

In a phase 4 follow-up of 280 men previously treated with CCH in IMPRESS or two open-label studies, mean curvature in the 247 men in the curvature response population had improved from 51.8° to 31.0° by the end of the original study. Among 180 men assessed at year 5 without further collagenase, curvature had improved by a further 4.3 ± 13.4° (95% CI 2.3 to 6.2), and the bother score fell from 6.5 to 3.4 by the end of the original study (183 men) and to 2.4 ± 2.9 at year 5 (123 men). The study had no comparison group, and 204 of 280 men (73%) completed it.^[7]

Surgery and treatment delivery

Treatment was an office procedure. Each cycle consisted of two injections of 0.58 mg CCH or placebo 24 to 72 hours apart, followed by clinician plaque modeling 24 to 72 hours after the second injection; up to 4 cycles were given 6 weeks apart (weeks 0, 6, 12 and 18), with non-treatment follow-up to week 52, for a maximum of 8 injections and 4 modeling procedures.^[1][2] Men were instructed to model at home for 6 weeks after each cycle. If curvature fell below 15° after the first, second or third cycle, or further treatment was not indicated, later cycles were omitted.^[1][2][3] The AUA guideline describes the IMPRESS protocol as the one that established the FDA-approved regimen.^[4] Completion was 477 of 551 CCH recipients (86.6%) and 251 of 281 placebo recipients (89.3%) across the two registry records.^[3]

The labeled technique follows that protocol. The penis is made erect (a single intracavernosal injection of alprostadil 10 or 20 mcg may be used) and the plaque is marked at the point of maximum concavity. The vial is reconstituted with 0.39 mL of the supplied calcium-containing diluent, and 0.25 mL (0.58 mg) is drawn into a 1-mL syringe with a 27-gauge, 1/2-inch needle. With the penis flaccid, the needle enters the edge of the plaque at the point of maximum concavity and is advanced transversely through the width of the plaque without passing through it, with resistance to gentle plunger pressure confirming intraplaque position; the dose is deposited along the needle track as the needle is withdrawn, and the second injection of the cycle is placed 2 to 3 mm from the first. At modeling, the surgeon grasps the flaccid penis about 1 cm proximal and distal to the injection site and, using the plaque as a fulcrum, bends the penis opposite to the curvature to moderate resistance for 30 seconds, three times with 30-second rests. Home modeling is gentle straightening during spontaneous erections (hold 30 seconds) and gentle stretching of the flaccid penis three times daily. The label contraindicates treatment of a plaque involving the urethra, advises avoiding injection into the urethra, nerves, vessels and corpora cavernosa, advises against use with anticoagulants other than low-dose aspirin, and tells patients to avoid sexual activity between the two injections and for at least 4 weeks after the second until pain and swelling have resolved, and not to use a vacuum erection device during treatment.^[2]

Adverse events

In the pooled placebo-controlled trials (551 CCH, 281 placebo), most adverse reactions were local to the penis and groin, mild or moderate, and 79% resolved within 14 days of injection. Penile hematoma occurred in 65.5% versus 19.2%, penile swelling in 55.0% versus 3.2%, penile pain in 45.4% versus 9.3%, penile ecchymosis in 14.5% versus 6.8% and erectile dysfunction in 1.8% versus 0.4%. Severe penile or injection-site hematoma occurred in 33 of 551 CCH recipients (6.0%) and none of 281 placebo recipients, and a penile popping sound or sensation, sometimes with detumescence, hematoma or pain, in 73 of 551 (13.2%) versus 1 of 281.^[2]

Three corporal ruptures, all surgically repaired, occurred among the 551 CCH recipients (0.5%).^[1] All three happened during intercourse, one during the required 14-day period without intercourse after treatment. Six men (1.1%) had a serious adverse event, three with corporal rupture and three with penile hematoma; one hematoma resolved spontaneously, one was aspirated and one was surgically addressed.^[1][4] Across all Peyronie's disease studies (1,044 CCH-treated men, 7,466 injections), corporal rupture was reported in 5 (0.5%); in another 9 (0.9%) ecchymosis or hematoma, sudden detumescence or a popping sound occurred and rupture could not be excluded, and severe penile hematoma occurred in 39 (3.7%).^[2] Penile shortening was not associated with treatment.^[2]

Among men who received at least two cycles, injection when curvature was under 30° (289 cycles, versus 1,204 cycles at 30° or more) was associated with more penile swelling (21.1% versus 14.5%, P = 0.007), penile hemorrhage (12.8% versus 8.9%, P = 0.046) and skin hyperpigmentation (1.0% versus 0.1%, P = 0.025); other treatment-related events, including serious ones, were similar.^[8] In the phase 4 follow-up, adverse events were reported in 49 of 280 men (17.5%) and none was considered treatment related.^[7]

Limitations

Both arms received injections and plaque modeling, so the net effect is the difference between arms and the 18.2% placebo-arm reduction is the reference for it.^[1][3] The AUA panel described the effect beyond placebo as modest, with a statistically significant mean difference of 7.7° between arms.^[1][4] Efficacy was analyzed in 612 of 832 men because the population required recent vaginal intercourse and complete assessments, and the PDQ was missing in about one-third of men at one or both time points and had not been fully validated psychometrically at the time of the guideline.^[2][4] The pooled analysis was post hoc, and the per-trial bother differences had confidence limits close to zero. Men with ventral curvature, an isolated hourglass deformity, a calcified plaque that would prevent injection, plaque near the base of the penis, or erectile dysfunction unresponsive to phosphodiesterase type 5 inhibitors were not enrolled, and the controlled follow-up ended at 52 weeks; later outcomes come from an uncontrolled phase 4 study with 73% completion.^[3][4][7] The registry lists Endo USA as sponsor.^[3]

In conversation

Same setting RestoreX (2019)

The comparator in IMPRESS was placebo injection with the same modeling, not observation. The EAU guideline notes that the 18.2% reduction in the placebo arm raises questions about the contribution of plaque injection and penile modeling regardless of the medication, and describes a modified short protocol (one 0.9-mg vial per cycle, distributed along three lines around the point of maximum curvature, up to three cycles four weeks apart, with home stretching, modeling and vacuum use replacing office modeling) whose results were reported as comparable to IMPRESS in small, largely uncontrolled series.^[9] The US label advises against vacuum erection device use during treatment.^[2]

Guidelines and approvals

• AUA (Peyronie's disease guideline, approved April 2015). Clinicians may administer intralesional CCH with clinician and patient modeling to reduce curvature in men with stable Peyronie's disease, curvature >30° and <90°, and intact erectile function, with or without medication . The panel describes IMPRESS I and II as the definitive trials that established the FDA-approved protocol. Clinicians should counsel men about ecchymosis, swelling, pain and corporal rupture before treatment .^[4]
• EAU (Sexual and Reproductive Health, limited update March 2026). Use intralesional CCH in men with dorsal or lateral curvature >30° who request non-surgical treatment . The guideline states that CCH was approved by the European Medicines Agency in 2014 and has been withdrawn from the European market by its manufacturer, and keeps the recommendation for completeness.^[9]
• FDA (December 6, 2013). Supplement approval for XIAFLEX for adult men with Peyronie's disease with a palpable plaque and curvature of at least 30°, together with a modified risk evaluation and mitigation strategy (REMS) requiring certification of prescribers and of dispensing pharmacies or settings because of the risk of corporal fracture and other serious penile injury. The letter also required a prospective long-term observational safety study of men treated in the three phase 3 studies (AUX-CC-802, -803 and -804).^[10]

Report history

IMPRESS report history

Published	Report	Cutoff	Headline results	Citation
2013	Two phase 3 trials, 52 weeks · primary	—	Mean percent change in curvature from baseline to week 52 (adjusted, last observation carried forward, mITT 199 versus 104); percentage-point difference (CCH minus placebo): −35.0% vs −17.8%; difference -17.2; CI -26.7–-7.6; P <0.01 Met Mean percent change in curvature from baseline to week 52 (adjusted, last observation carried forward, mITT 202 versus 107); percentage-point difference (CCH minus placebo): −33.2% vs −21.8%; difference -11.4; CI -19.5–-3.3; P <0.01 Met Mean change in Peyronie's Disease Questionnaire bother domain score (0 to 16, higher is worse) from baseline to week 52 (mITT 199 versus 104); difference (CCH minus placebo): −2.8 vs −1.6; difference -1.2; CI -2.4–-0.03; P <0.05 Met Mean change in Peyronie's Disease Questionnaire bother domain score (0 to 16, higher is worse) from baseline to week 52 (mITT 202 versus 107); difference (CCH minus placebo): −2.6 vs −1.5; difference -1.1; CI -2.1–-0.002; P <0.05 Met Post hoc pooled mean percent improvement in curvature at week 52: 34% vs 18.2%; P <0.0001 Post hoc pooled mean change in curvature at week 52, degrees (mean ± SD); difference (CCH minus placebo): −17.0 ± 14.8° vs −9.3 ± 13.6°; difference -7.7 Post hoc pooled mean change in PDQ symptom bother score at week 52 (mean ± SD); difference (CCH minus placebo): −2.8 ± 3.8 vs −1.8 ± 3.5; difference -1; P 0.0037 Mean IIEF overall satisfaction score, baseline to week 52, combined analysis of both trials: 5.6 to 6.6 vs 5.6 to 6.0; —	Gelbard M, Goldstein I, Hellstrom WJ, et al. Clinical efficacy, safety and tolerability of collagenase clostridium histolyticum for the treatment of Peyronie disease in 2 large double-blind, randomized, placebo controlled phase 3 studies. *J Urol.* 2013;190(1):199-207. doi:10.1016/j.juro.2013.01.087 DOI
2015-05-18	Pooled subgroup analysis · post-hoc	—	CCH versus placebo, baseline to week 52, by baseline curvature, disease duration, calcification and erectile function: Reductions occurred in every subgroup. Curvature reductions were significantly greater with CCH for baseline curvature 30° to 60° (n=492) and 61° to 90° (n=120), disease duration over 2 to 4 years (n=212) and over 4 years (n=199), no calcification (n=447) and IIEF 17 or higher (n=480); not significant for duration 1 to 2 years (n=201), non-contiguous stippling (n=103), contiguous calcification not interfering with injection (n=62), IIEF 1 to 5 (n=22) and IIEF 6 to 16 (n=106).; P <0.05 CCH versus placebo, baseline to week 52, by subgroup: Bother reductions were significantly greater with CCH for curvature 30° to 60°, disease duration over 4 years, no calcification, and IIEF 1 to 5 and 17 or higher.; P <0.05	Lipshultz LI, Goldstein I, Seftel AD, et al. Clinical efficacy of collagenase Clostridium histolyticum in the treatment of Peyronie's disease by subgroup: results from two large, double-blind, randomized, placebo-controlled, phase III studies. *BJU Int.* 2015;116(4):650-656. doi:10.1111/bju.13096 DOI
2017	Safety by curvature at injection · post-hoc	—	—	Hellstrom WJG, Tan RBW, Liu G. Safety profile of collagenase clostridium histolyticum stratified by degree of penile curvature in patients with Peyronie disease. *Urology.* 2017;106:237.e9-237.e14. doi:10.1016/j.urology.2017.05.004 DOI
2020	Phase 4 follow-up to 5 years · updated	—	Mean penile curvature in men previously treated with CCH who received no further treatment: 51.8° to 31.0° at the last visit of the original study (reference; 247 men); a further 4.3° improvement at year 5 (180 men); P <0.02 Descriptive Mean PDQ bother domain score in men previously treated with CCH: 6.5 to 3.4 at the last visit of the original study (reference; 183 men); 2.4 ± 2.9 at year 5 (123 men); P 0.0003 Descriptive	Goldstein I, Lipshultz LI, McLane M, et al. Long-term safety and curvature deformity characterization in patients previously treated with collagenase clostridium histolyticum for Peyronie's disease. *J Urol.* 2020;203(6):1191-1197. doi:10.1097/JU.0000000000000743 DOI
2023	Response by treatment cycle · post-hoc	—	Proportion of CCH recipients with a 20% or greater reduction from baseline after one cycle: 29.9%; — Proportion reaching a 20% or greater reduction by cycle 4 among CCH recipients who had not responded earlier: 60.8% of cycle 1 failures; 42.7% of cycle 1-2 failures; 23.5% of cycle 1-3 failures; —	Ziegelmann M, Hu Y, Xiang Q, et al. Incremental treatment response by cycle with collagenase clostridium histolyticum for Peyronie's disease: a pooled analysis of two phase 3 trials. *Urology.* 2023;175:126-131. doi:10.1016/j.urology.2023.02.019 DOI

---

References

1. Gelbard M, Goldstein I, Hellstrom WJ, et al. "Clinical efficacy, safety and tolerability of collagenase clostridium histolyticum for the treatment of Peyronie disease in 2 large double-blind, randomized, placebo controlled phase 3 studies." J Urol. 2013;190(1):199-207. doi:10.1016/j.juro.2013.01.087

2. Endo USA, Inc. XIAFLEX (collagenase clostridium histolyticum) for injection: prescribing information and Medication Guide. Revised 4/2026. Sections 2.2, 4, 5, 6.2 and 14.2, boxed warning. endodocuments.com/XIAFLEX/PI

3. ClinicalTrials.gov. NCT01221597 (AUX-CC-803) and NCT01221623 (AUX-CC-804): phase 3, double-blind, randomized, placebo-controlled studies of AA4500 in men with Peyronie's disease. Sponsor: Endo USA. NCT01221597; NCT01221623

4. Nehra A, Alterowitz R, Culkin DJ, et al. "Peyronie's disease: AUA guideline." J Urol. 2015;194(3):745-753. doi:10.1016/j.juro.2015.05.098. Guideline statements 8 and 9 and discussion, pp 12-14 of the AUA-approved document (April 2015).

5. Lipshultz LI, Goldstein I, Seftel AD, et al. "Clinical efficacy of collagenase Clostridium histolyticum in the treatment of Peyronie's disease by subgroup: results from two large, double-blind, randomized, placebo-controlled, phase III studies." BJU Int. 2015;116(4):650-656. doi:10.1111/bju.13096

6. Ziegelmann M, Hu Y, Xiang Q, et al. "Incremental treatment response by cycle with collagenase clostridium histolyticum for Peyronie's disease: a pooled analysis of two phase 3 trials." Urology. 2023;175:126-131. doi:10.1016/j.urology.2023.02.019

7. Goldstein I, Lipshultz LI, McLane M, et al. "Long-term safety and curvature deformity characterization in patients previously treated with collagenase clostridium histolyticum for Peyronie's disease." J Urol. 2020;203(6):1191-1197. doi:10.1097/JU.0000000000000743

8. Hellstrom WJG, Tan RBW, Liu G. "Safety profile of collagenase clostridium histolyticum stratified by degree of penile curvature in patients with Peyronie disease." Urology. 2017;106:237.e9-237.e14. doi:10.1016/j.urology.2017.05.004

9. Salonia A, Boeri L, Capogrosso P, et al. EAU Guidelines on Sexual and Reproductive Health. Arnhem: EAU Guidelines Office; limited update March 2026. Section 8.2.3.a.4.

10. U.S. Food and Drug Administration. "Supplement approval, BLA 125338/61 (XIAFLEX)." December 6, 2013. accessdata.fda.gov