30 to 50% of the patients presenting with Erectile Dysfunction (ED) do not respond to PDE V Inhibitor therapy, which is presently considered as the first choice treatment for most ED patients. Recent reports stated a high prevalence of low serum testosterone levels in such non responders, and an improvement of their response by combining testosterone therapy with the PDE V Inhibitor. This suggests there may be a minimum threshold level of blood testosterone for a full effectiveness of PDE V Inhibitor therapy. Two double blind, placebo controlled studies have added support to this hypothesis but one involved only 20 patients while in the other the benefit of combining testosterone was transient. This is a multi-centric study, double blind placebo controlled and randomized as concerns testosterone administration, that aims to objectively assess the efficacy of co-administering testosterone with the PDE 5 inhibitor Tadalafil to improve the erectile function of a large group of ED patients non-responders to PDE V inhibitors alone. Patients will be screened to ensure inclusion and exclusion criteria completion, including a serum testosterone level < 4 ng/ml for total testosterone or < 1 ng/ ml for bioavailable testosterone. They will then enter a four week run-in period in the meanwhile they will receive Tadalafil 10 mg only, once daily, in order to confirm their non responsiveness to PDE V inhibitors and their eligibility to enter the treatment phase based on IIEF scoring, SEP diaries and a Global Assessment Question (GAQ). The patients still non responders after 4 weeks of Tadalafil 10 mg daily will enter a 12 weeks treatment phase including visits at weeks 4, 8, 12 and 16. Treatment procedure will include: 1. continuation of Tadalafil at 10 mg dose daily followed by routine assessment using SEP diaries, IIEF scoring, GAQ and Aging Male Symptoms scale administered at each study visit. Safety assessments will be performed in addition during the last visit (physical examination including DRE, PSA and BCC). 2. Randomization in 2 parallel arms (Placebo gel + Tadalafil 10 mg daily, and Testosterone gel 50 mg + Tadalafil 10 mg daily). If indicated according to suboptimal clinical response of the patient, the dose of study medication will be increased at the 8 or 12 weeks visit to 100mg of testosterone or to 2 sachets of placebo gel. Up to 430 patients will be screened in order that 172 are enrolled in the double blind treatment phase.

Design and Methodology: This was a multicentre, randomised, double-blind, placebo-controlled study of the effects of co-administering testosterone with the PDE 5 inhibitor tadalafil in ED patients who do not respond to PDE 5 inhibitors alone.

The patients were randomised into two parallel groups: the Control group (tadalafil plus placebo) and the Test group (tadalafil plus testosterone).

The study lasted a maximum of 16 weeks for each patient and was divided into two parts:

• Run-in phase: four week period from visit 1 (V1) to V2 where patients received a dose of 10 mg tadalafil per day. Non-response to tadalafil was assessed by the SEP diary, the IIEF questionnaire and the GAQ. Patients had to attempt to have intercourse at least four times during this phase. Non-responders were randomised at V2.
• Treatment phase: 12 week period from the day after V2 to V5 with visits at weeks 8, 12 and 16. The patient response was assessed by the SEP diary, the IIEF and AMS questionnaires and the GAQ at each visit. Patients were to apply the testosterone or placebo gel to their skin once a day, preferably in the morning. There was an option to increase the dose of testosterone or placebo at V3 or V4 if the patient had a suboptimal clinical response (patient felt insufficiently improved and achieved scores lower than 5 for questions 3 or 4 of the IIEF).

Test product. The products given during this study are licensed under the names of Cialis (tadalafil) and Androgel and Testogel (testosterone gel).

All patients were treated with 10 mg tadalafil tablets once daily over the four week run-in phase then, if randomised, for the following 12 week treatment phase.

Patients in the Test group were treated with 5 g sachets of testosterone gel (50 mg testosterone) once daily for the 12 week treatment phase, with an option to increase to two 5 g sachets (100 mg testosterone) per day from V3 or V4 if the patient had a suboptimal clinical response (patient felt insufficiently improved and achieved scores lower than 5 for questions 3 or 4 of the IIEF questionnaire). Patients in the Control group received 5 g sachets of placebo gel once daily for the 12 week treatment phase, again with an option to increase to two 5 g sachets per day from V3 or V4.

Statistical methods

1. Descriptive statistics:

   1. Categorical variables were summarised using classical frequency statistics: number of non-missing observations (N) and percentages (%) by categories. Percentages were calculated within each treatment group on the number of non-missing observations and were displayed using one decimal.
   2. Continuous variables were summarised using standard quantitative statistics. number of non-missing observations (N), mean, standard deviation (SD), median and range (minimum and maximum observed values).

   The number of missing observations (N missing) was also specified. The 95% confidence interval (CI) was displayed when relevant.

2. Inferential analyses:Comparisons between treatment groups were two-sided. The significance level was set at 0.050. All p-values were rounded to three decimal places.Main inferential analyses used one of the following tests:

   1. Chi two-Test or Fisher's exact Test comparing the distributions of a categorical variable between the level of one factor (e.g. treatment arm), in reference to theoretical distributions. When at least one theoretical frequency is less than 5, then the Fisher's exact test was used in place of Chi two-Test
   2. Analysis of variance (ANOVA): comparing the mean values of a continuous quantitative variable between the level of categorical factors (e.g. treatment group, centre)
   3. Analysis of covariance (ANCOVA) ANOVA adjusted for treatment group and centre effects and baseline value as covariate: The primary efficacy criterion was analysed by an ANCOVA, including the covariate 'baseline value of the primary criterion' and the factor 'treatment group'. The ANCOVA estimated the difference between the two treatments as well as its two-sided 95% CI.

Summary and conclusions

Efficacy results:

Of the 173 patients included in this study, 35 were prematurely withdrawn. Therefore 138 patients completed the study. The ITT population consisted of 167 patients and the PP population of 120; 47 patients in the ITT population were excluded for major protocol deviations.

No statistically significant differences were found between the testosterone and the placebo groups as concerns the primary criterion in either the ITT or the PP populations. The EFD score of the IIEF questionnaire increased between baseline (V2 for efficacy results) and endpoint (V5 or withdrawal visit) for both groups, indicating an overall improvement in erectile function during the study.

Apart from the hormone levels, there were no statistically significant differences for any of the secondary criteria in either the ITT or the PP populations. As expected, for certain hormones there were significant differences between the two groups (total testosterone (TT), bioavailable testosterone (BT), Dihydrotestosterone (DHT), oestradiol, luteinizing hormone (LH), follicle stimulating hormone (FSH), calculated FT (cFT) and calculated BT (cBT)). Significant differences were also found for all these hormones for the treatment responders.

There was no statistically significant difference for the additional analysis; the percentage of successful sexual intercourse attempts amongst treatment responders was similar between groups.

At V2, after the run-in phase of four weeks of tadalafil treatment alone, the responder rate was 17.0% and the rate of those patients with a score > 26 for the EFD of the IIEF i.e. considered as no longer having ED, was 14.8%. For all domains of the IIEF, the score was higher at V2 than at V1, indicating an increase in erectile function after the run-in phase. Almost half of the selected patients (44.8%) thought that the tadalafil treatment had improved their erections. However, the results of these exploratory analyses may be biased as they were performed on the Selected population (who responded to the IIEF at both V1 and V2) and not on the Randomised population.

Additional exploratory analyses were performed to determine the testosterone threshold from which a possible improvement would be obtained by testosterone gel. The results of these analyses found statistically significant differences between the two groups, in favour of the Test group, in the ITT patients with a TT level of 3 ng/ml or less at baseline. The results included a significantly higher increase in the primary criterion (EFD score) at V4 (p=0.027),after 8 weeks of testosterone gel, and significantly greater improvements in various secondary criteria. For the IIEF questionnaire: significant improvements were shown in the score of the Orgasmic Function Domain at V4 (p=0.028), in the Intercourse Satisfaction Domain at V4 (p=0.005), in the Overall Satisfaction Domain at endpoint (p=0.046) and in the total IIEF score at V4 (p=0.008). For the SEP diary a significantly higher increase was shown in the rate of attempts of intercourse resulting in vaginal penetration (SEP 2) at V4 (p=0.033), and in the rate of totally successful intercourses (SEP 3) at V4 (p=0.038) and endpoint (p=0.006). These results suggest that testosterone gel significantly improved erectile function compared to placebo gel in ED patients who are non-responders to PDE5 inhibitors with a baseline testosterone level of 3 ng/ml or less, and that this effect is discernible from the second month of administration (assessments done at V4).

Safety results Overall, 61 of the 173 Safety population patients (35.3%) experienced at least one AE during the study (111 AEs were reported in total) with more AEs reported in the Control group than in the Test group. In the Safety population, 32 pre-treatment AEs were recorded in 23 patients (13.3%) with no significant difference between groups.

During the study, a total of 79 emergent AEs were recorded in 53 patients (30.6%), 34 AEs in 22 Test group patients and 45 AEs in 31 Control group patients. All AEs were of mild or moderate intensity, except four AEs considered as severe, all of which were reported in the Test group (pneumopathy, arrythmia, bowel obstruction and exacerbation of back pain). The latter was the only severe AE related to one of the study products.

Five emergent AEs were considered as serious for three Test group patients (coronary stenosis and diabetes, pneumopathy and arrhythmia, and bowel obstruction). These SAEs were unrelated to the study product. The patient with the bowel obstruction was withdrawn from the study after V4.

A total of 11 patients (four in the Test group -diabetes impaired by corticotherapy, itching, bowel obstruction, nausea- and seven in the Control group) were withdrawn from the study due to an AE.

Conclusions In conclusion, in this study, testosterone gel did not improve the efficacy of tadalafil in a population of ED patients with a low or low-to-normal testosterone level (TT < 4 ng/ml or BT < 1 ng/ml) who were non-responders to tadalafil 10 mg once-a-day alone. However additional exploratory analyses found significant improvements with tadalafil plus testosterone gel compared to tadalafil plus placebo gel in a subgroup of the ITT population restricted to those patients with a serum TT level <3 ng/ml at baseline. These results agree with the scientific literature, which places the threshold level below which a man may be considered hypogonadal (testosterone deficient) at 3 ng/ml (and not at our inclusion criterion of 4 ng/ml).

Inclusion Criteria:

1. ED complaint ongoing for over 3 months;
2. Age comprise between 45 and 80 years old;
3. Had a stable heterosexual relationship for more than 3 months and anticipates having the same partner for all the study
4. Has not responded adequately to the highest available dosage of Tadalafil or other PDE5 inhibitors (20 mg for Tadalafil and Vardenafil, 100 mg for Sildenafil) taken at least at 4 separate occasions, defined as: a score of 2,3 or 4 at Question 3 of the IIEF AND a score of 2 or 3 at Question 4 of the IIEF; measured prior to Visit 1
5. Low or low-to-normal serum testosterone level (either on total or bioavailable testosterone levels) with respect to the range of men under aged than 50 y.o. (TT < 4 ng/ml and/or BT < 1 ng/ml) according to a first assay done prior to Visit 1 and a confirmation by a second assay at central laboratory Biolille on blood sampled at Visit 1
6. Agrees to make at least 4 attempts at sexual intercourse on 4 separate days during the 4 weeks run-in period with daily Tadalafil 10 mg
7. At least 50% of attempts during this period must be unsuccessful according an answer "No" at one of the questions 1 ("were you able to achieve at least some erection (some enlargement of the penis)?"), 2 ("were you able to insert your penis in your partner's vagina?") or 3 ("did your erection last long enough for you to have successful intercourse?").
8. At the end of the run in phase with Tadalafil 10 mg daily, the patient should provide: a score of 2, 3 or 4 at Question n°3 of the IIEF AND a score of 2, 3 at Question n°4 of the IIEF
9. Agrees not to use any other ED drug or non-drug (devices) treatment during the full course of the study;
10. Provides a signed informed consent.

Exclusion Criteria:

1. Impotence caused by other primary sexual disorder (e.g. premature ejaculation);

2. History of penile implant or significant penile deformity;

3. Body mass index >35kg/m2;

4. Diabetes mellitus that is uncontrolled (HbA1c level > 10%). HbA1c will be checked at screening for each diabetic patient or suspected to be;

5. Uncontrolled thyroid disorders;

6. Known hyperprolactinemia (serum prolactin > 30ng/ml in local laboratory);

7. Organic hypothalamic-pituitary pathology;

8. History of alcohol, drug or substance abuse within 6 months before Visit 1;

9. Renal insufficiency defined as receiving renal dialysis, having a creatinine clearance < 30 ml/mn, or serum creatinine > 30 mg/ml;

10. Severe hepatic impairment, Child Pugh class C, elevation of AST and/or ALT > 3 x the ULN;

11. Systolic Blood Pressure > 170 or < 90 mm Hg or diastolic blood pressure > 110 or < 50 mm Hg at screening;

12. Cardiac disease contra-indicating any sexual activity;

13. Unstable angina within 6 months before Visit 1;

14. Angina during sexual intercourse within 6 months before Visit 1;

15. Myocardial Infarction within 90 days before Visit 1;

16. Coronary artery by-pass graft surgery or percutaneous coronary intervention (angioplasty or stent insert) within 90 days before Visit 1;

17. Severe cardiac rhythm disturbances e.g. supraventricular arrhythmia with a ventricular response >100 bpm. at rest despite medical or device therapy, history of refractory spontaneous or induced sustained ventricular tachycardia (heart rate > 100 bpm. for > 30 sec) or fibrillation, automatic internal cardioverter-defibrillator, history of sudden cardiac arrest) within 6 months before Visit 1;

18. Known new and significant conduction defect that was not evaluated with regard to significance within 90 days prior to Visit 1;

19. Congestive heart failure (NYHA Class II or above) within 6 months before Visit 1;

20. History of stroke within the 6 last months;

21. Epilepsy not adequately controlled by treatment;

22. Polycythemia with hematocrit >52% at study entry (i.e. screening visit/visit 1);

23. Suspicion of current, or past history of prostate or breast cancer;

24. Severe symptomatic Benign Prostate Hyperplasia;

25. PSA value exceeding the age specific reference ranges published by Richardson and Oesterling, Urol Clin North Am, 1997, 24: 339-351

26. Diagnosed sleep apnea;

27. Extensive skin abnormalities that could affect absorption of the gel;

28. Any clinically significant chronic disease that might, in the opinion of the investigator, compromise patient's safety, interfere with the evaluations, or preclude completion of the trial (e.g. hemochromatosis, chronic lung disease, chronic malabsorption disease);

29. History of HIV infection;

30. Severe psychiatric disease;

31. Illiteracy, lack of fluency in the language used for the writing of the protocol and questionnaires, unwillingness, medical, psychiatric or other conditions that compromise the patient's ability to understand the patient information, to give informed consent, to understand or complete diary or questionnaires or otherwise comply with the trial protocol, or to complete the study;

32. Known hypersensitivity to Cialis(Tadalafil);

33. Hypersensitivity to the active substances or any of the excipients of Androgel®/ Testogel®;

34. Use of androgen therapy or anabolic steroids within 6 months of entry into the study (i.e. screening visit/visit 1);

35. Concurrent use of the following medications:

    androgens including dehydroepiandrosterone (DHEA) and anabolic steroids, antiandrogens, estrogens, corticotrophin (ACTH), oxyphenylbutazone, clomipramine, Serotonin Reuptake Inhibitors, long or short-acting nitrates, NO donors, potent cytochrome P3A4 inhibitors (e.g. ketoconazole, itraconazole, ritonavir, saquinavir, macrolides like erythromycin), cancer chemotherapy;

36. Patients unwilling to cease use of vacuum devices, intracavernosal injection, Viagra, or other therapy for ED;

37. Patients seeking conception or on treatment for infertility;

38. Concurrent participation in another clinical trial within 1 month of entry into this study (i.e. screening visit/visit 1) or throughout the duration of the study;

39. Previous randomization into this study.

40. History of temporary or permanent partial or complete blindness