Moxisylate versus papaverine (see papaverine versus moxisylate, above).

Phentolamine plus PGE₁ versus PGE₁. Two trials compared phentolamine plus PGE1 versus PGE1 alone for efficacy and/or harms (Aversa 1996; studies: a and b).²⁶⁷

Harms. In the single trial that reported prolonged erection, ²⁶⁷ 4.2 percent of participants in each treatment group reported this adverse effect.

Efficacy. In the first trial, 54.2 percent of participants randomized to phentolamine plus PGE₁ reported improved erections versus 20.8 percent of those randomized to PGE₁.²⁶⁷ The corresponding proportions reported for the other study were 60 versus 30 percent, respectively.²⁶⁷

Papaverine plus phentolamine versus placebo. One trial compared the efficacy and harms of papaverine plus phentolamine versus placebo.²⁶⁶

Harms. In total, 15 and 18.3 percent of the participants randomized to papaverine plus phentolamine reported the occurrence of pain and prolonged erection, respectively. None of the participants in placebo group experienced these adverse events.

Efficacy. In total, 56.7 percent of participants randomized to papaverine plus phentolamine reported improved erections versus 0 percent of those randomized to placebo.

Papaverine plus phentolamine versus papaverine plus phentolamine plus sexual counseling. One trial compared the efficacy and harms of papaverine plus phentolamine versus papaverine plus phentolamine plus sexual counseling.²⁵⁷

Harms. Results for harms in this study were pooled for the two treatment groups. 12 percent of the men discontinued the treatment due to prolonged erection. Priapism was reported in three (4.3 percent), hematoma in four (5.7 percent), and curvature of the penis in one participant (1.4 percent).

Efficacy. The mean values on a self-rated erections score (scale 0–100) for papaverine plus phentolamine versus papaverine plus phentolamine plus sexual counseling groups were 79 versus 84 percent, respectively.

Trimix versus PGE₁. Two trials compared the efficacy and harms of trimix versus PGE1 alone.^{255, 272} In the first trial, 32 men (mean age: 61 years) with ED of at least 6 months of duration (etiology not reported) refractory to in-clinic injection of papaverineplus phentolamine, were randomized in a crossover fashion to 40μ g of PGE1 versus 17.64 mg papaverine plus 0.58 mg phentolamine plus 5.8μ g PGE1.²⁷² Another trial enrolled 180 men (mean age: 51 years) with ED at least 6 months' duration, predominately of organic cause, of whom 20 percent had complete ED.²⁵⁵ Men in this trial were randomized in a crossover fashion to 20μ g of PGE1 alone versus one of nine different dose combinations of papaverine (range 5mg–20mg) plus phentolamine (1 mg) plus PGE1 (2.5–10μ g).

Harms. In the first trial, 12.5 percent of participants randomized to trimix reported pain versus 40.6 percent of those allocated to PGE₁.²⁷². In the second trial, corresponding proportions were 14.4 percent (for trimix) versus 17.3 percent (for PGE₁).²⁵⁵ Only the second trial reported data on priapism, which occurred in 5.0 percent of participants allocated to trimix versus 0.6 percent of those allocated to PGE₁.²⁵⁵

Efficacy. About half (50 percent) of the participants randomized to trimix reported grade 4 or 5 erections versus 21.9 percent of those randomized to PGE₁ alone.²⁷²

There was no difference between trimix and PGE₁ for erections of either grade 4 or 5 (trimix: 66.7 percent versus PGE₁: 67.8 percent), or for improvement in self-rated erection compared with home (trimix: 83.2 percent versus PGE₁: 84.9 percent, p = 0.85). There was no statistically significant difference in efficacy outcomes between PGE₁ and any individual trimix dose combination.²⁵⁵

Trimix plus atropine versus trimix. One trial compared the efficacy and harms of trimix versus trimix plus atropine.²⁶⁴ Addition of atropine to trimix did not reduce pain or improve erections compared with trimix alone.

Harms. In total 55.3 percent of participants allocated to trimix plus atropine reported pain versus 50 percent of those allocated to trimix alone.

Efficacy. Each treatment group reported improved erections in 45.6 percent of the participants.

Trimix plus sodium bicarbonate versus trimix. One trial compared the efficacy and harms of trimix injections with and without sodium bicarbonate.²⁸³

Harms. In total pain was reported by 5.3 percent of the participants receiving trimix plus sodium bicarbonate compared with 57.9 percent of those in the group treated with trimix alone.

Efficacy. The difference between the rates of improved erection in participants allocated to trimix plus sodium bicarbonate versus trimix alone was not statistically significant (78.9 percent versus 68.4 percent; RD 11 percent, 95 percent CI: -17.0–38.0).²⁸³

Vasoactive intestinal peptide (VIP) plus phentolamine versus placebo. Two trials compared the efficacy and harms of vasoactive intestinal peptide (VIP) plus phentolamine versus placebo.²⁶⁰

Harms. Compared with participants receiving placebo, those randomized to VIP plus phentolamine reported more frequent bruising (12.3 percent versus 43.1 percent), bleeding at injection site (5.1 percent versus 20.5 percent) urethral bleeding (2.6 percent versus 12.3 percent), flushing (13.3 percent versus 74.4 percent), palpitation (0 percent versus 7.7 percent), and tachycardia (0.5 percent versus 5.1 percent).

The participants in the placebo group reported bleeding at the injection site. There was no statistically significant difference between the treatment groups with respect to pain during injection (4.6 percent versus 8.2 percent), priapism (0.5 percent versus 0 percent), or headache (3.6 percent versus 1.5 percent).

Efficacy. Three hundred and four men with ED were screened for response to in-clinic or home administration of 25μ g VIP plus either 1mg or 2 mg of phentolamine.

Based on the phentolamine dose to which responses were observed, 240 participants were randomized in a crossover design to active treatment versus placebo. Efficacy results were reported only for the 172 men who received at least one dose of active drug and placebo.

Seventy-two percent of injections in men allocated to VIP plus 1 mg phentolamine produced grade 3 erections (suitable for sexual intercourse) versus 13 percent in men allocated to placebo (p < 0.001). The proportions of participants with grade three erections in VIP plus 2 mg phentolamine and placebo groups were 65 and 16 percent, respectively (p < 0.001)

Tables 13–15 illustrate the results presented in this section.