A randomized, double-blind, crossover, placebo-controlled comparative clinical trial of arginine aspartate plus adenosine monophosphate for the intermittent treatment of male erectile dysfunction



Thierry Lebret, Department of Urology, Hôpital Foch, 40, rue Worth, Suresnes 92150, France. E-mail: t.lebret@hopital-foch.org


Efficacy and safety of l-arginine aspartate 8 g combined with 200 mg of adenosine monophosphate (AA) with placebo (PL) alone for intermittent treatment of mild-to-moderate erectile dysfunction (ED) were compared. The study design was a double-blind, PL-controlled, two-way crossover randomized clinical trial with 26 patients. Efficacy was assessed by International Index of Erectile Function (IIEF) and two additional validated questionnaires [the Erection Hardness Score (EHS) and the Erectile Dysfunction Inventory of Treatment Satisfaction (EDITS). During each crossover period, separated by a 2-week wash-out period, drugs were administered orally, 1–2 h before sexual intercourse. Primary endpoint was a change in the IIEF. Secondary endpoints were patient and investigator assessments of treatment success. Investigators’ and patients’ assessment of efficacy was significantly improved by the combination vs. PL (= 0.01 and p = 0.04 respectively]. EHS and EDITS questionnaires were both improved by the combination (p = 0.015 and p = 0.017 respectively). There was no significant difference in terms of tolerance between AA and PL or severe adverse events. ED patients demonstrated significant improvements in all IIEF domains with the exception of the Sexual Desire and Orgasmic Domains when treated with AA compared with PL. This pilot phase II study showed that the on-demand oral administration at a high dosage of l-arginine aspartate–adenosine monophosphate combination may be effective in patients with mild-to-moderate ED, is very well tolerated and could be tested as a safe first-line therapy in a larger size phase III study.


Disorders of erectile function are a major health problem according to the Massachusetts Male Aging Study, affecting 52% of men aged 40–70, with approximately 70% self-reporting mild-to-moderate sexual dysfunction and 30% suffering from severe or even complete impotence (Feldman et al., 1994; O'Donnell et al., 2005). Many patients with only mild-to-moderate forms describe their erectile dysfunction (ED) as a transitory, unstable, insufficient and often unpredictable erection failure, possibly embarrassing but often considered as not serious enough to justify the routine use of prescribed drugs with potentially unpleasant side effects. For these patients very few highly safe therapeutic alternatives are available and their efficacy rarely exceeds that of a placebo (PL). Erection is a neurovascular phenomenon mediated by nitric oxide (NO). This key mediator of erection in mammals is released upon various possible stimuli mainly from cavernosal nerve endings and partially from penile artery endothelial cells (Furchgott & Zawadzki, 1980; Palmer et al., 1988a,b; Ignarro et al., 1987; Furchgott, 1988; Burnett et al., 1992; Kim et al., 1991; Rajfer et al., 1992). NO release by counteracting the vasoconstriction effect locally exerted by the adrenergic tone on smooth muscular cells results in cavernous artery vasodilatation. Phosphodiesterase type-5 inhibitors (iPDE5) enhance NO activity and are currently the most effective oral drugs for the treatment of male ED. Nevertheless, these molecules may sometimes induce unwanted adverse effects and are not always the adapted treatment for patients with mild-to-moderate ED, where unpleasant side effects should be avoided (multi-disease patients) or not always accepted with regard to their discomfort (episodic sufferers). There are also patients who do not respond well to drugs that inhibit NO degradation and who could be better improved by an arginine/adenosine supplementation, potentially those where endothelial dysfunction is the predominant causal factor.

Considering the essential role played by NO in the physiology of erection, NO donors and first of all l-arginine, the physiological precursor of NO (Palmer et al., 1988a,b), are among the most attractive agents for patients who suffer from non-severe impotence. The natural origin of l-arginine, its good bioavailability after oral absorption (Bode-Böger et al., 1998) and its excellent tolerance allowing long course treatment (Barbul, 1986) has made it a widely used supplement by male consumers seeking self-medication for ED. However, the level of efficacy reached by the chronic use of arginine alone remains uncertain (Boucharlat et al., 1972; Zorgniotti & Lizza, 1994; Moody et al., 1997; Klotz et al., 1999; Chen et al., 1999). To our knowledge only one double-blind randomized study that combined arginine with low doses of yohimbine administrated intermittently, proved to be effective in mild-to-moderate male ED (Lebret et al., 2002).

Based on the well-documented pro-erectile effects of purines observed in the isolated rabbit and human corpus cavernosum experimental model and also in in vivo studies after intra-cavernous administration of either ATP or adenosine (Kataoka et al., 2007; Tong et al., 1992; Chiang et al., 1994; Filippi et al., 2000), this work hypothesized that a specific high dose combination of l-arginine and adenososine monophosphate could improve the erectile status of patients suffering from mild-to-moderate dysfunction when used ‘on demand’ 1 or 2 h before intercourse. Our pilot study was designed to evaluate the efficacy and safety of this new combination, orally administered, for this subpopulation of patients.

Material and methods


We conducted a randomized double-blind, crossover, single centre, comparative, PL-controlled trial. The main inclusion and exclusion criteria are summarized in Table 1. Three different validated questionnaires (Rosen et al., 1997; Mulhall et al., 2007; Althof et al., 1999) were used for the assessment of erectile function: the International Index of Erectile Function (IIEF), the Erection Hardness Score (EHS) and the Erectile Dysfunction Inventory of Treatment Satisfaction (EDITS). At the screening visit the IIEF scale was employed alone for subject selection: only patients with an Erectile Function Domain score between 14 and 22 (of a 25 maximum–optimal score) were included in the study. All selected patients underwent a physical examination, provided a previous medical history and signed an informed consent form. After randomization, they were either included in arm A (receiving first AA, and after a 2-week wash-out period a PL) or in arm B (receiving first PL, and after a 2-week wash-out period AA).

Table 1. Selection of study population (short list)
Inclusion criteria

1. Signed Informed consent

2. Men over 18 years of age

3. Patients suffering from erectile dysfunction for at least 3 months. This was defined as erections inadequate to maintain (either in terms of frequency and/or quality) the individual's normal sexual behaviour satisfactorily

4. Patients with grade 3 or 4 erections in certain circumstances, including morning erections, during the 4 weeks prior to selection

5. Patients in a stable relationship desiring an active heterosexual life

Exclusion criteria

1. Patients suffering from severe grade 1 or 2 erection of neurological or vascular origin

2. Psychotic patients receiving neuroleptic drugs

3. The following were contraindicated during the trial; nitrates, estrogens, antiandrogens, anxiolytic drugs not taken as sleeping tablets, LH-RH analogues and tricyclic antidepressants. No concomitant treatments, which may have affected erectile disorders and may had been prescribed once a patient had entered the trial

4. Patients with renal insufficiency (serum creatinine > 200 μmol/L) or clinically significant hepatic insufficiency

5. Patients who had previously received one of the trial treatments within 30 days or 5 half-lives (referring to the longest half-life) before inclusion into the trial

The formulation used for this study was prepared by an independent drug products development company CRID (St. Gély du Fesc, France), had known stability and was in the form of white granules soluble in water, of similar size and taste, for both the combination, l-arginine aspartate 8 g–Adenosine monophosphate 200 mg (AA), and the PL.

At inclusion (before randomization) and after each period of treatment the three questionnaires (IIEF, EHS and EDITS) were completed by the patients (without the presence of the investigator). The two comparative periods of treatment were separated by a 2-week wash-out period without assessment to avoid a carry-over effect. The treatment drugs of the first and second period were delivered according to randomization. Patients were instructed to take the medication 1–2 h prior to intercourse, no more than once a day. At the end of the study the overall patient's opinion regarding the two periods of treatment with their comparative impact on the erectile function was recorded by the investigator. He also reported his own opinion after extensive interview of patients' intercourse attempts and successes.

Statistical methods

Determination of sample size and randomization

The number of patients required to demonstrate that the arginine–adenosine monophosphate combination is superior to PL in the treatment of ED was calculated using statistical methods similar to those used in a previous referenced study (Lebret et al., 2002). It was determined that a difference of 50% standard deviation of the outcome of interest could be detected with 80% power and alpha = 0.05 if a minimum of 24–32 subjects were recruited for the study. Finally, 35 consecutive patients were selected to obtain at least 24 evaluable patients.

Statistical analysis

Descriptive statistics (means and standard deviations) were calculated for all efficacy evaluable patients for each of the 15 IIEF questions and the five domain scores based on the IIEF. Conditions required for robust statistical analysis (such as the assumption of normality) were screened: two treatment groups were compared using a general linear model (GML StatsSoft Statistica software; StatSoft, Inc., Tulsa, OK, USA) repeated measures procedure. In addition, the possibility of significant carry-over effects for the primary efficacy variables was evaluated. After assessing these overall comparisons, a series of paired t-tests were performed, comparing the two treatments. All of these analyses were carried out for individual questions as well as for the domain scores of the IIEF. For the analysis of EHS and EDITS questionnaires the same methodology and tests were applied.

Urologist and patient opinions about overall efficacy were analysed using McNemar's test for paired samples data.

All statistical analyses were performed at the 5% statistical significance level (i.e. alpha = 0.05) using SAS software (SAS Institute Inc., Cary, NC, USA).


Thirty-five patients were randomized to treatment. Six of them, who decided to discontinue following the protocol after period 2 (four under PL and two under AA), and three who encountered difficulties in their relationships leading them to interrupt their sexual life were withdrawn from the study. Twenty-six patients completed the trial according to the protocol design and were suitable for statistical analysis. The mean age of the study's population was 56.46 ± 9.26. Eight patients were treated for arterial hypertension and four for hypercholesterolaemia. No patient was withdrawn because of adverse events. Complete demographic, comorbidities and medication data description is shown in Table 2.

Table 2. Demographic, comorbidity and medication data for the 26 patients who completed the trial
  1. BMI, Body Mass Index.

Mean age ± SD (years)56.46 ± 9.26
Mean weight ± SD (kg)79.3 ± 10.4
Mean height ± SD (cm)172.9 ± 6.3
BMI (kg/m²)26.49 ± 3.03
Surgical history (n)
Transurethral resection of prostate2 (8%)
Transurethral resection of bladder1 (4%)
Medical history (n) and medications (n)
Arterial hypertension8 (31%)
Beta adrenergic blocking agent2
Hypercholesterolaemia4 (15%)
Cardiac arrhythmia2 (8%)
Beta adrenergic blocking agent1
Diabetes1 (4%)
Gastroesophageal reflux disease1 (4%)
Mucoviscidosis1 (4%)
Levocetirizine and salbutamol1

The mean score value of the overall population for the Erectile Function Domain at baseline was 19.01. The treatment value for the five IIEF domains in the study population is summarized in Table 3. Significant differences between treatment and PL regarding score variations from baseline, in the overall cohort and for the 15 questions of the IIEF are summarized in Table 4. There was a significant difference between the combined therapy and PL, in favour of the combination (AA), for the Erectile Function Domain (p = 0.004), the Intercourse Satisfaction Domain (p = 0.01) and the Overall Sexual Satisfaction Domain (p = 0.001). The Orgasmic and Sexual Desire Domains were not modified by treatment or PL. Consistent with IIEF results, a significant difference in favour of the combination vs. PL was also observed with the EHS (= 0.015) and EDITS questionnaires (p = 0.017).

Table 3. Treatment values for IIEF domains in the study population
Score of EF ≥14 (n = 26)MeanSDp values (AA vs. Pl)
  1. IIEF, International Index of Erectile Function.

Erectile function
Orgasmic function
Sexual desire
Intercourse satisfaction
Overall satisfaction
Table 4. Significance of positive score variations between the AA combination vs. Placebo for the 15 questions of the IIEF
IIEF Domain Questions p
  1. NS, not significant.

Erectile function Q1: Frequency0.03
Erectile function Q2: Adequacy0.03
Erectile function Q3: Penetration frequency0.05
Erectile function Q4: Maintain erectionns
Erectile function Q5: Difficulty maintaining0.04
Intercourse satisfaction Q6: Frequency of attempts at intercoursens
Intercourse satisfaction Q7: Frequency of satisfaction during intercourse0.02
Intercourse satisfaction Q8: Enjoyment of sexual intercourse0.008
Orgasmic function Q9: Orgasm frequencyns
Orgasmic function Q10: Ejaculation frequencyns
Sexual desire Q11: Frequency of sexual desirens
Sexual desire Q12: Level of sexual desirens
Overall satisfaction Q13: Satisfaction with overall sex life0.025
Overall satisfaction Q14: Sexual relationship with the partner0.002
Erectile function Q15: Confidence0.04

The AA combination showed a significant greater success rate than PL as estimated by the investigator (p = 0.01). Similarly the patients' estimation of success for the administration of AA was better compared with PL (p = 0.04).

There was no significant difference as regards to tolerance between AA and PL as well as no withdrawal related to any adverse event. Only mild adverse effects were recorded in the treatment group (two patients had mild gastrointestinal complaints).

Discussion and conclusion

This study was conducted to evaluate an oral treatment tailored for mild-to-moderate ED patients for whom the iPDE5 therapies have a reduced benefit/risk ratio. As a natural NO donor, l-arginine is one of the most attractive substances to achieve this purpose, provided it is given at the right dosage. Klotz et al. showed that l-arginine is not effective when administered as a supplement for 2 weeks, at 1.4 g/day (Klotz et al., 1999). Nevertheless, a few studies (Boucharlat et al., 1972; Zorgniotti & Lizza, 1994; Chen et al., 1999) have suggested that l-arginine taken alone at higher dosages, between 2.8 and 6 g/day for 2–6 weeks, could have beneficial effects in the treatment of ED. However, the subjective improvements reported by patients in these trials were not assessed by a validated questionnaire, in particular the erectile function domain was not studied as required by current standards. Thus, they remain controversial.

In a previous work (Lebret et al., 2002), a combination of l-arginine glutamate 6 g with yohimbine 6 mg (AY) had been tested as l-arginine-based therapy for ED. The efficacy and safety of this combination administered as needed in a single dose, 1 h before intended sexual intercourse, were assessed in a double-blind, controlled, at home trial, by the IIEF and a patient log. Results showed that the AY combination was significantly effective in improving erectile function in patients with mild-to-moderate ED (EF score ≥14). Unfortunately, because of various regulatory issues involving yohimbine, the development of this combination was abandoned. Interestingly, the rationale supporting the AY combination had considered both of the predominating biochemical mechanisms leading to penile vasorelaxation, that is, the adrenergic tone blockage (induced by yohimbine) and the release of NO from nerve endings and endothelial cells (induced by arginine). However, besides these two pivotal mechanisms a large number of central and peripheral transmitters, such as purines (adenosine, AMP, ATP) can act as modulators.

In a standard male rabbit corpus cavernosum tissue preparation study which focused on the effect of purine-nucleotides AMP and ATP, we confirmed AMP pro-erectile activities and observed that AMP and l-arginine strongly potentiate each other's effect. In fact, a synergistic effect could be obtained by their combined use (Hupertan et al., 2012).

Because of its extremely short half-life in the blood stream (<5–10 sec) adenosine cannot be used orally. Similarly, when considering purine nucleotides and purine nucleosides, only AMP is available for oral use because the half-life of other nucleotides–nucleosides is much too short for it to survive in the gastrointestinal tract and its bioavailability is very poor after oral administration. Once in the bloodstream, AMP is naturally converted over time into adenosine by ecto-nucleotidases enzymes at the surface of endothelial cells. A2a receptors are coupled to Gs proteins, which stimulate the formation of cyclic AMP (cAMP). The cAMP pathway by decreasing intracellular calcium concentration induces smooth muscle cell relaxation and possibly other erectogenic mechanisms (Wyatt et al., 2002).

Another reason to combine AMP and l-arginine was their very close pharmacokinetic profile with a plasma peak between 30 and 60 min for both, followed by a 1–2 h decrease (Bode-Böger et al., 1998; Wu et al., 2007; Tangphao et al., 1999; Malmary-Nebot et al., 1979). From a clinical point of view, these very rapid kinetics suggest, at least in the ED domain, an ‘as needed’ ‘on demand’ use of the two products rather than a chronic use. These PK profiles in addition to l-arginine/AMP potentiating effects, justified our choice, in this study, for an ‘on demand’ method of use.

The dose selected for arginine was based on our previous experience (Lebret et al., 2002), also that of others (Barbul, 1986) suggesting that l-arginine is ineffective below 3 g/day and finally on the excellent track record of this amino acid for oral doses below 20–30 g/day divided into three intakes (Tousoulis et al., 2002). In this study and including known published data we used 8 g of l-arginine aspartate (5.2 g l-arginine base) combined with 200 mg of adenosine monophosphate which is in the low range of doses so far experienced with these two substances. Daily doses of 120–900 mg AMP via the oral route has been commonly prescribed for over three decades and are extremely well tolerated in medical applications such as venous insufficiency and porphyria (Gajdos, 1974).

In this study, in selected patients with mild-to-moderate ED, significant improvements were shown in all domains of the IIEF with the exception of the Sexual Desire and Orgasmic Domains when treated with AA compared with PL. The investigator and patient's own assessment as well as the EHS and EDITS questionnaires on erection also demonstrated significant improvements, which are consistent with IIEF results. It is worth noting that the improvement in the EF domain, corresponding to a 1.85 progression of the mean score, was obtained in a 2-week trial interval only, in a small study population and in patients with mild-or-moderate impotence in whom significant variations are usually more difficult to demonstrate than in patients with severe impotence. Arginine (via NO synthase) and AMP (via adenosine) both have a predominant vascular effect especially in diseases where endothelium-dependent relaxation is impaired (Böger & Ron, 2005; Gür & Öztürk, 2000; Lerman et al., 1998). Both substances are also well tolerated at high dosages. As sexual dysfunction can be related to endothelial dysfunction and decreased availability of both adenosine (as a local vasoregulator), and of l-arginine (as a substrate for endothelial NOS), in ageing, diabetes, atherosclerosis and chronic tobacco addiction, specific subpopulations such as heavy smokers, selected cardiac and diabetic patients with ED of limited severity, could benefit from this AA combination.

The results of our pilot phase II study showed that the on-demand oral administration of l-arginine aspartate–adenosine monophosphate combination may be effective as a safe first-line therapy in patients with mild-to-moderate ED and is very well tolerated. We only found two very mild gastrointestinal complaints in 26 patients who received the combination and no adverse event in patients (n = 8) taking antihypertensive drugs. Oral doses of 9 g/day or less of an arginine salt are usually free of side effects (Tousoulis et al., 2002), whereas AMP side effects are very rare especially at the dose of 200 mg. In comparison iPDE5 adverse events are more frequent (Table 5). As a whole for those patients with mild-to-moderate ED in whom an excellent tolerance profile is a prerequisite for the use of pro-erectile products (some episodic SD sufferers) or desirable (multi-disease patients), the l-arginine /AMP combination could be an interesting alternative to iPDE5. In addition, some patients who do not respond well to iPDE5 could better benefit from a supplementation strategy, possibly those in whom endothelial dysfunction impacts their erectile function.

Table 5. Adverse events and contraindications. iPDE5 vs. AA. Comparative table
 Most frequent side effectsContraindicationsPrecaution
  1. AA, L-ARG/AMP combination, AE, adverse events.

PDE5 inhibitors

PDE5i most frequent adverse events are headache (15–16%), flushing (10–11%), dyspepsia (4–10%)

For some of them, abnormal vision (3%), rhinitis (3–4%), back pain (4%) and dizziness (2%) are quite frequent

PDE5i are contraindicated in patients taking long-acting nitrates and in patients with a recent history of myocardial infarction or strokePrecaution is required in patients with angina taking short-acting nitrates and in patients taking antihypertensive drugs

L-ARG/AMP combination

Only arginine is discussed in this table as AMP 200 mg is totally free of AE

We only found two very mild gastrointestinal complaints in 26 patients who received the combination and no adverse event in patients (n = 8) taking antihypertensive drugs

Oral doses of 9 g/day or less of an arginine salt are usually free of side effects (Tousoulis et al., 2002). Oral doses of 7 g three times per day induce mild GI disorders in less than 8% that disappear with a dose of 7 g twice a day (Lerman et al., 1998)

Oral AMP side effects are very rare

Absolute contraindications with arginine have not been identified

Beneficial effects of supplemental l-arginine have been demonstrated in patients with intractable angina at the dose of 9 g/day, in patients with stable angina with dose going from 6 to 15 g/day given during 3–15 days (Böger & Ron, 2005)

Although arginine does not block or inhibit NO degradation and has a shorter peak of action than PDE5i, the same recommendations as for PDE5i should apply in patients taking nitrates and/or antihypertensive drugs

These preliminary results need to be confirmed in a larger patient population study (phase III), including specific subpopulations of patients such as non-responders to PDE5 inhibitor therapy, episodic sufferers and multi-disease patients. Our pilot phase II study showed that the on-demand oral administration at a high dosage of l-arginine aspartate–adenosine monophosphate combination may be effective in patients with mild-to-moderate ED, is very well tolerated and could be tested as a safe first-line therapy in a larger size phase III study.


The authors are grateful to Richard Medeiros, Medical Editor – Medical Editing International for editing the manuscript.

Conflict of interest or financial support