Paper | Published:

Combination therapy for erectile dysfunction: a randomized, double blind, unblinded active-controlled, cross-over study of the pharmacodynamics and safety of combined oral formulations of apomorphine hydrochloride, phentolamine mesylate and papaverine hydrochloride in men with moderate to severe erectile dysfunction


Oral therapy has become first line treatment for patients with mild to moderate erectile dysfunction (ED). Studies have shown that sildenafil may not be effective in all patients, and has been associated with a variety of adverse effects and an adverse interaction with nitrates and inhibitors of cytochrome P450 enzymes. The objective was to compare the efficacy and safety of three different oral combinations with the highest dose of sildenafil in men with moderate to severe ED. Randomized, double blind, unblinded active-controlled, Phase II study was carried out at three sites in Mexico. After a 4-week placebo run-in period, patients received all four of the following treatments using a 4-way cross-over design: 40 mg phentolamine (PM) +6 mg apomorphine (Apo); 40 mg PM +150 mg papaverine (Pap); 40 mg PM +6 mg Apo +150 mg Pap (Tricombo); 100 mg sildenafil (SC). With the exception of sildenafil tablets, all study medication was blinded. Moderate to severe ED was defined as a less than 50% vaginal penetration success rate during the placebo run-in period. A total of 44 patients were enrolled, of whom 36 completed all four treatment periods. All treatments produced a significant effect in primary efficacy variable (Sexual Encounter Profile) compared to baseline, however, no statistically significant differences were found between treatments. A significant period effect was observed. Also, the four treatments were found not to differ significantly in five out of six secondary efficacy variables. The lowest incidence of treatment-related adverse events (AE) occurred in the 40 mg PM +6 mg Apo group (9.8%), followed by 100 mg SC (15%), and the other two combinations (16.7 and 17.5%, respectively). Nasocongestion and headache were the most frequently reported AE. An oral combination of vasoactive agents may provide an alternative approach to sildenafil. Based on these results a combination of phentolamine and apomorphine warrants further clinical investigation.


The Massachusetts Male Aging Study reported a combined prevalence of 52% for minimal, moderate, and complete impotence in non-institutionalized men between the ages of 40 and 70 y.1 A wide variety of etiologies have been described for this condition, however, the most frequently reported is arterial insufficiency associated with cardiovascular disease. Also, studies have shown that the etiology of this condition is not predictive of response to treatment, and choice of therapy depends on underlying etiology, severity of disease, physician's preference, and treatment success and tolerance by individual patients.2,3,4

Currently, there are several single pharmacological agents approved for the treatment of erectile dysfunction (ED), including oral, intraurethral and intracavernosal applications. Other oral vasoactive agents are currently undergoing clinical testing or regulatory review.

Oral sildenafil has been reported as a safe and efficacious treatment for ED, however, it is ineffective in approximately 27% of the population, and has been associated with a variety of adverse effects including headache, flushing, dyspepsia, and adverse interaction with nitrates and inhibitors of cytochrome P450 enzymes.5,6,7,8,9

In an attempt to improve effectiveness in ED patients dissatisfied with single agent therapy, combinations of different pharmacological agents have been explored. Intracavernosal injection combinations of papaverine, phentolamine and alprostadil (prostaglandin-E1) are frequently used off-label, whereas combinations of oral and local (intraurethral) formulations have also shown promise.10,11,12,13,14

These findings triggered interest in evaluating an oral combination of phentolamine, papaverine and apomorphine. This combination was chosen to couple a central acting agent (apomorphine, dopamine receptor agonist) with the predominantly peripheral activities of phentolamine (non-selective alpha-adrenoceptor antagonist) and papaverine (non-specific phosphodiesterase inhibitor).15,16,17,18

In an initial Phase I study, three oral combinations of 40 mg phentolamine mesylate, 150 mg papaverine hydrochloride, and 2, 4 or 6 mg of oral apomorphine, respectively, were safely administered to 10 male volunteers.19

The primary objective of the present Phase II, Proof-of-Concept study was to compare pharmacodynamics and safety of different oral tablet combinations of apomorphine, phentolamine and papaverine, with those of 100 mg sildenafil in men with moderate to severe ED.



Men who were enrolled in this study were aged 40–75 inclusive, had a history of ED of at least 6 months duration, and were engaged in a stable, monogamous heterosexual relationship. All men provided written informed consent and had to inform their partners of study participation. Moderate to severe ED was defined as a less than 50% vaginal penetration success rate in a 4-week placebo run-in period.

Exclusion criteria included ED caused by an untreated endocrine disease; postural hypotension or uncontrolled hypertension (systolic >160, dia-stolic >90 mmHg); significant penile pathology (curvature, fibrosis, implants, etc.); history of clinically significant hepatic or renal disease, coronary artery disease, nervous system diseases, radical prostatectomy, or uncontrolled psychiatric condition; known intolerance to papaverine, phentolamine, apomorphine or sildenafil; concomitant use of organic nitrates; regular use of sympathetic-mimetic medication; use of antiemetics that are central nervous system depressors; blood donations; planned or ongoing participation in another study; any condition which would interfere with the patient's ability to provide informed consent, to comply with study instructions, or which might confound the interpretation of the study results; and, any condition, which in the opinion of the principal investigator could endanger the health or wellbeing of the patient.

Study design and methodology

The study was conducted at three sites in Mexico. The institutional review board of each participating institution approved the protocol.

After an initial 4-week placebo run-in period to determine study eligibility, subjects received all four of the following treatment options during the treatment period using a 4-way randomized, crossover design: 40 mg phentolamine plus 6 mg apomorphine (Ph40+Apo6); 40 mg phentolamine+150 mg papaverine (Ph40+Pap150); 40 mg phentolamine plus 150 mg papaverine and 6 mg apomorphine (Tricombo); 100 mg sildenafil (SC100). Combination tablets were blinded, whereas for logistical reasons no blinding was used for sildenafil.

At the start of each of the four treatment periods, the subject received an in-office dose of the randomized treatment followed by an observation period of 60 min to evaluate safety. Subjects with a decrease of 30% or greater in their diastolic or systolic blood pressure, or an increase of 30% or greater in heart rate from their baseline value were withdrawn from the study.

During the treatment periods, subjects recorded the outcome of an attempt at sexual intercourse in a Sexual Encounter Profile (SEP) diary. Men returned diary cards and received replacement medication at each subsequent visit. At each of these visits, drug accountability was performed, and adverse events (AE) and concomitant medications since last visit were reviewed and recorded. Also, the investigator evaluated the global assessment question (GAQ).

At the last visit, upon completion of all study procedures, the investigator evaluated the GAQ, and recorded subject's response to the following question: ‘Which of the four treatments did you prefer or liked best?’

Statistical methods

No power-based sample size could be determined accurately since response rates for sildenafil and two of the three experimental drug treatments for the study-specific efficacy parameters were not fully known. However, it was determined that approximately 48 patients had to be enrolled and no more than a 8.33% dropout/non-evaluable rate to reach 44 evaluable patients.20

The mean SEP attempt score was the primary efficacy variable. The SEP score is a numerical rating scale used to objectively rate the patients' erectile response (self-assessment) in a stepped fashion during three sexual encounters. The SEP included questions regarding: (1) attempts at sexual activity; (2) patients' ability to achieve at least some erection; (3) the ability to achieve vaginal penetration; (4) erectile duration sufficient for ejaculation; (5) satisfaction with the hardness of the erection; and (6) overall satisfaction with the sexual experience. A positive answer to each question in the SEP was scored as a one, thus allowing a maximum of six points for each sexual encounter. A six-point SEP mean was calculated for the (maximum of) 12 sexual encounters per treatment.

Secondary efficacy variables were: percentage of vaginal penetrations; percentage of penetrations leading to ejaculations; mean erection rigidity as determined on a Visual Analogue Scale (scale 0–100; VAS); mean VAS for duration of erection; mean VAS for satisfaction with sexual experience, and the Global Assessment Question (GAQ).

The efficacy population was defined as the set of all subjects randomized to treatment sequence, who had received all four treatments in their assigned sequence, had no major protocol violations, and had efficacy data for two or more tablets to calculate the primary efficacy variable (SEP attempt mean) for each of the four treatments dispensed.

Monitoring AE assessed safety. The safety population was defined as all subjects who received at least one dose of study medication. The incidence of AE was summarized overall, by COSTART body system and by preferred term for each treatment group. Further summarizations based on severity (mild, moderate, severe), and relation to study treatment (related, not related) were performed.

The SEP mean attempt score at each visit was analyzed using a model appropriate for a 4×4 Latin square design. The model included terms for treatment and treatment by site interaction, if there were sufficient degrees of freedom. In addition, Gould's method was used to account for patients who chose not to continue with a particular treatment due to lack of efficacy or unacceptable side effects. A subgroup analysis was performed stratifying by patient's previous sildenafil use.21

A similar analysis was performed for the secondary analysis variables, except for the GAQ. A corresponding ordinal analogue of the model appropriate for a 4×4 Latin Square Design was employed for the GAQ secondary efficacy variable.

All of the two-group efficacy statistical tests performed were one-sided, controlling for the type I error level at 0.10. The statistical test(s) conducted for the primary efficacy analyses were under the null hypothesis of ‘not as good as’ vs the alternative hypothesis of ‘at least as good as’ for at least one of the three combination treatment groups compared to sildenafil.


A total of 68 patients were included in the 4-week run-in period, of whom 25 were not randomized (vaginal penetration success rate greater than 50%, or lost to follow-up). Of the 43 randomized patients, one was lost to follow-up after randomization and the remaining 42 patients received at least one dose of the study drug. Thirty-six (36) patients completed all four treatment periods with at least two tablets per period (almost all of these patients took four or more tablets per period). No patient was withdrawn during the 60 min observation period after receiving the in-office dose of each of the randomized treatments.


Results for the primary efficacy variable, the mean six-point SEP score are listed in Table 1. Differences between all four treatments were not large numerically and did not reach statistical significance. However, all four treatments produced highly significant increases (P<0.001) in mean SEP response compared to baseline (Table 2). The mean increase from baseline was 1.84 points for SC100, 1.88 for PM40+Apo6, 1.68 for Tricombo and 1.63 for PM40+Pap150, respectively.

Table 1 Primary and secondary efficacy variables with four different oral treatments for erectile dysfunction (LS-means)
Table 2 Change from baseline in mean six-point SEP score by treatment

There was a statistically significant overall period effect (P=0.0111, 3 d.f.). Patients showed improved scores in the last two treatment periods, independently of the type of treatment they received. It should be noted that the third and fourth period effects were larger than the largest treatment effect, for example, SC100 −Ph40+Apo6=0.22 points on six-point SEP scale.

Previous sildenafil users had a slightly lower mean placebo lead-in baseline score than naive sildenafil users (mean SEP score 1.45 vs 1.67, respectively; Table 1). In the active treatment period, the two strata differed in treatment effect, but this did not reach statistical significance which may be due to the small sample size in the previous sildenafil user group (n=7).

Results for five of the six secondary response variables (3 VAS, proportion of vaginal penetration and proportion of orgasms) were similar to SEP; that is the four treatment means were found not to differ significantly (Table 1). The only significant difference was found for the GAQ (five-point ordinal scale ranging from 1=very satisfied to 5=very dissatisfied), where sildenafil showed a statistically significant lower score than the other three treatments.

The percentage of succesful vaginal penetrations, penetrations leading to orgasm, and of the patients' preference for a specific treatment at the end of the study was similar among all four treatment groups (Figure 1).

Figure 1

Rate of successful vaginal penetration, penetration leading to orgasm and overall treatment preference.


The safety population consisted of 42 patients. Overall, 19 patients reported a total of 150 AE during the study.

The highest incidence of treatment-related AE was reported in the Tricombo group (7/40, 17.5%), followed by Ph40+Pap150 (7/42, 16.7%), 100 mg SC (6/40, 15%), and the Ph40+Apo6 group (4/41, 9.8%; Table 3).

Table 3 Adverse events during treatment with different oral combinations for erectile dysfunction

Nasocongestion (rhinitis) and headache were the most frequently reported treatment-related AE (ranges 4.8–15%, and 2.4–5%, respectively).

Only two patients reported a cardiovascular AE: one patient with palpitations and tachycardia on Tricombo, the other tachycardia and coronary artery disorder while taking 100 mg sildenafil.

Only one patient reported a serious AE: hospitalization for a right nephrectomy not related to treatment. All treatment-related AE were mild or moderate in severity.


This was a multi-center, randomized, partly double blind, phase II proof-of-concept study to determine whether pharmacodynamics and safety of different oral combinations of vasoactive agents were at least as good as those of 100 mg sildenafil in the treatment of moderate to severe ED.

Thirty-six out of the 44 patients enrolled by three centers in Mexico received all four treatments in their randomly assigned sequence. The differences between all four treatments were not large numerically and did not reach statistical significance for any of the primary or secondary response variables, except for the global assessment question.

The fact that the sildenafil medication was not blinded most likely introduced a recognition bias in the study in favor of Viagra® especially because the majority of study subjects were familiar with it and had used this medication before study participation.

However, previous sildenafil use only had a small but not significant effect on the primary efficacy outcome (Table 1). Previous users had a lower baseline SEP score, and a numerically weaker response to any of the four treatments, perhaps an indication of a more severe level of ED, or a disappointing experience with the drug in the past. The small sample size (n=7) should be taken into consideration as well.

For the mean six-point SEP score (primary efficacy variable) there was a period effect, which means that patients did increasingly better as the trial periods advanced, independent of the type of treatment.

The observed period effect across all four treatments in the study is most likely due to increased confidence with sexual performance during study participation.

Despite the recognition bias, the four-way, randomized cross-over design allowed for a more accurate determination of treatment preference than a parallel design. Those patients who completed all four study periods were about equally divided over their treatment preference.

Most AE were mild in severity, and the most frequently reported AE, rhinitis and headache, were to be expected based on the pharmacology of these agents.

The results of this study show that there may be a maximum level that single or a combination of vasoactive drugs can achieve in the treatment of ED. This is supported by the observation that the triple drug combination performed as well as the two combinations of apomorphine plus phentolamine, and of phentolamine plus papaverine. However, the triple combination showed more AE than the other treatments.

Coupled with the relative safety of the bi-combo formulations, one can conclude that an oral combination of two vasoactive drugs with different pharmacodynamic activity may provide an alternative approach to oral treatment with the highest approved dose of sildenafil. Especially, the combination of phentolamine and apomorphine warrants further clinical investigation.


  1. 1

    Feldman HA et al . Impotence and its medical and psychosocial correlates: results of the Massachusetts Male Aging Study. J Urology 1994 151: 54–61.

  2. 2

    Hanash KA. . Comparative results of goal oriented therapy for erectile dysfunction. J Urology 1997 157: 2135–2138.

  3. 3

    Hatzichristou DG, Bertero EB, Goldstein I. . Decision making in the evaluation of impotence: the patient profile-oriented algorithm. Sexuality and Disability 1994 12: 29–37.

  4. 4

    Rosen RC, Goldstein I, Padma-Nathan H. . A process of case model evaluation and treatment of erectile dysfunction. Robert Wood Johnson Medical School: New Brunswick, New Jersey, May 1998

  5. 5

    Goldstein I et al . Oral sildenafil in the treatment of erectile dysfunction. New Engl J Med 1998 338: 1397–1404.

  6. 6

    Goldstein I. . A 36-week, open label non-comparative study to assess the long-term safety of sildenafil citrate (Viagra®) in patients with erectile dysfunction. Int J Clin Pract 1999 102: Suppl 8–9.

  7. 7

    Schwartz I, McCarthy D. . Sildenafil in the treatment of erectile dysfunction [letter]. New Engl J Med 1998 339: 699–700.

  8. 8

    McMahon CG, Samali R, Johnson H. . Efficacy, safety and patient acceptance of sildenafil citrate as treatment for erectile dysfunction. J Urol 2000 164: 1192–1196.

  9. 9

    Kloner RA, Zusman RM. . Cardiovascular effects of sildenafil citrate and recommendations for its use. Am J Cardiol 1999 84: 11N–17N.

  10. 10

    Zentgraf M, Baccouche M, Junemann KP. . Diagnosis and therapy of erectile dysfunction using papaverine and phentolamine. Urol Int 1988 43: 65–75.

  11. 11

    Truss MC, Becker AJ, Schiltheiss D, Jonas U. . Intracavernous pharmacotherapy. World J Urol 1997 15: 71–77.

  12. 12

    Soli M et al . Vasoactive cocktails for erectile dysfunction: chemical stability of PGE1, papaverine and phentolamine. J Urol 1998 160: 551–555.

  13. 13

    Mydlo JH, Volpe MA, Macchia RJ. . Initial results utilizing combination therapy for patients with a suboptimal response to either alprostadil or sildenafil monotherapy. Eur Urol 2000 38: 30–34.

  14. 14

    Kaplan SA et al . Combination therapy using oral alpha-blockers and intracavernosal injection in men with erectile dysfunction. Urology 1998 52: 739–743.

  15. 15

    Przedborski S et al . Peripheral and central pharmacokinetics of apomorphine and its effect on dopamine metabolism in humans. Mov Disord 1995 10: 28–36.

  16. 16

    Heaton JP et al . Recovery of erectile function by the oral administration of apomorphine. Urology 1995 45: 200–206.

  17. 17

    Traish AM et al . A heterogeneous population of alpha-1 adrenergic receptors mediates contraction of human corpus cavernosum smooth muscle to norepinephrine. J Urology 1995 153: 222–227.

  18. 18

    Poch G, Kukovetz WR. . Papaverine-induced inhibition of phosphodiesterase activity in various mammalian tissues. Life Sci I 1971 10: 133–144.

  19. 19

    Phase I Safety Study in Healthy Volunteers of Different Oral Combinations of Papaverine, Phentolamine, and Apomorphine; Data on File, Zonagen Inc.

  20. 20

    Machim D, Campbell MJ. . Statistical tables for the design of clinical trials. Blackwell Scientific Publications, Oxford 1987

  21. 21

    Gould L. . A new approach to the analysis of clinical drug trials with withdrawals. Biometrics 1980 36: 721–727.

Download references

Author information

Correspondence to P I Lammers.

Rights and permissions

Reprints and Permissions

About this article


  • erectile dysfunction
  • impotence
  • pharmacodynamics
  • safety
  • oral combination
  • apomorphine
  • phentolamine
  • papaverine

Further reading