Introduction. D Sandhu 1, E Curless 2, J Dean 3, G Hackett 4, S Liu 5, D Savage 6, R Oakes 7 * and G Frentz 8

Transcription

1 International Journal of Impotence Research (1999) 11, 91±97 ß 1999 Stockton Press All rights reserved /99 $ A double blind, placebo controlled study of intracavernosal vasoactive intestinal polypeptide and phenotolamine mesylate in a novel auto-injector for the treatment of non-psychogenic erectile dysfunction D Sandhu 1, E Curless 2, J Dean 3, G Hackett 4, S Liu 5, D Savage 6, R Oakes 7 * and G Frentz 8 1 Department of Urology, Leicester General Hospital, Leicester, UK; 2 Centre, Bolton General Hospital, Bolton, UK; 3 Derriford Hospital, Plymouth, UK; 4 Good Hope Hospital, Sutton Cold eld, UK; 5 Department of Urology, City General Hospital, Stoke on Trent, UK; 6 St Vincents Medical Centre, Doncaster, UK; 7 Senetek plc, Kettering, Northamptonshire, UK; and 8 Department of Urology, Tulane University School of Medicine, New Orleans, Louisiana, USA Three hundred and four patients with non-psychogenic erectile dysfunction (ED) completed a dose assessment phase with intracavernosal injection utilizing 25 mg vasoactive intestinal polypeptide (VIP) combined with phentolamine mesylate 1.0 mg (VIP=P-1) or 2.0 mg (VIP=P-2) in an autoinjector for a response rate of 83.9%. In a sub-group of 183 patients who withdrew from one or more previous ED therapies, 82% responded with an erection suitable for intercourse. One hundred and ninety- ve patients were subsequently treated in a placebo controlled phase. 75.1% responded to VIP=P-1, 12% to placebo (P < 0.001); 66.5% responded to VIP=P-2, 10.3% to placebo (P<0.001), with the median duration of erection of 54 min. The principal adverse event was transient facial ushing in 2770 injections (33.9%). There was no pain post injection and two episodes of priapism (0.05%). Only nine patients withdrew because of adverse events. Over 85% and 95% of patients were satis ed with the drug and auto-injector, respectively. Over 81% of patients and 76% of partners reported an improved quality of life. Keywords: vasoactive intestinal polypeptide; phentolamine; erectile dysfunction; intracavernous injection Introduction Vasoactive intestinal peptide (VIP), or aviptadil, is a 28 amino acid endogenous peptide widely distributed throughout the human body. VIP concentrations in cavernous blood increases up to 20-fold during erection 1 and its presence has been noted to be decreased in the penises of impotent men. 2,3 VIP belongs to the family of nonadrenergic, noncholinergic ancillary neurotransmitters which act by activation of cyclic-adenosine monophosphate (camp). 4 In addition, VIP may act with nitric oxide (NO) as a co-mediator of erection. 5 Its action results in the relaxation of cavernous smooth muscle and a possible effect on the veno-occlusive mechanism. In early studies, intra-cavernosal (IC) injection of VIP produced resulted in varying degrees of tumescence following vibratory and=or visual stimulation. 1,6,7 *Correspondence: Dr R Oakes, Senetek plc, 1400 Montague Court, Kettering Venture Park, Kettering, Northamptonshire, NN15 6XR UK. Received 15 June 1998; accepted 2 November 1998 When combined with the vasodilator phentolamine, Gerstenberg et al 8 reported encouraging results in 1380 IC injections in 52 patients with primarily nonpsychogenic impotence utilizing a dose range of 0.5 ± 2.0 mg phentolamine and 30 mg VIP. All patients achieved a satisfactory erection following sexual stimulation and there were minimal sideeffects, notably no priapism or post-injection pain. The drop out rate of 17% was low compared to that of 24 ± 42% reported in other studies. 9 In a recent study of 20 males with erectile dysfunction (ED), McMahon 10 reported encouraging results utilizing IC injection of 30 mg VIP combined with 1.0 or 2.0 mg phentolamine. An erection suf cient for penetration was achieved in all six patients with psychogenic, 7 out of 9 with arteriogenic, 2 out of 2 with neurogenic, and 1 out of 3 with venous leakage aetiologies, for an overall ef cacy of 80%. There was no priapism, 10% of patients complained of injection pain and 15% noted transient facial ushing. These ndings led to further evaluation of the VIP=Phentolamine combination (INVICORP TM ) for the treatment of ED utilizing IC injection. A novel auto-injector was developed for use with VIP=Phen-

2 92 tolamine (VIP=P). `Injection anxiety' is reduced and in a pilot study of 20 patients previously treated with IC self-injection utilizing VIP=P ampoules, the auto-injector was well received and the injection was described as `painless'. 11 We hereby report the results of a prospective, double-blind, cross-over, placebo controlled trial of the IC use of VIP=P utilizing a novel auto-injector for the treatment of predominantly non-psychogenic ED in 304 men. Methods Patient Selection Three hundred and four men with ED of at least one year duration were evaluated at 10 investigational sites within the United Kingdom. To be included in the study, men were required to be 18 y or older and currently be in a stable heterosexual relationship. Men were excluded from the study if their history suggested overt psychogenic aetiology, ability to perform occasional sexual intercourse without medication or physical aid, hypoactive sexual disorder, nocturnal or early morning erections, or if erection occurred in response to placebo during the training program. In addition speci c exclusion criteria included any known systemic disease which produces an overall weakness (for example, cancer, acquired immune de ciency syndrome), sickle cell trait or disease, hormonal imbalance requiring treatment (except diabetes mellitus (DM)), angina pectoris, uncompensated congestive heart failure, a history of alcohol or drug abuse, current infection of the genitourinary tract, indwelling urethral catheter, surgery which may prevent erectile response to medication, concurrent therapy which in the opinion of the investigator could seriously diminish the effect of study medication (for example, benzodiazipines, sedatives, stimulants, anti- or pro-adrenergic compounds), markedly abnormal ECG, biochemical or haematological results, and participation in any clinical study within the previous 30 d. All patients and their partners gave informed consent and the study was performed in accordance with the Declaration of Helsinki (revised version of Tokyo 1975, Venice 1983 and Hong Kong 1989) as well as the European Directive 91=507=EEC and CPMP Guidelines II.=3976=88-EN dated 11 July The study was approved by the Local Ethics Committee at each of the ten investigative sites. Study Design The study was conducted between January 1996 and December 1997, and was divided into a Dose Assessment Phase and two Placebo Controlled Phases, each of six month duration. In the Dose Assessment Phase 304 men who met the entry criteria were administered VIP=P-1 (25 mg VIP and 1.0 mg phentolamine in 0.35 ml) in the clinic by the investigator utilizing the auto-injector. Patients were as relaxed as possible and provided with suf cient privacy and visual stimulating material as required. A Grade 3 erection achieved in response to this test dose allowed the patient to be immediately admitted to the Placebo Controlled Phase. For purposes of this study, the grade of erection was scored as: 0ˆno effect; 1ˆtumescence only; 2ˆpartial erection, unsuitable for vaginal penetration; and 3ˆerection suitable for vaginal penetration (as judged in clinic) or sexual intercourse in home environment. Following this initial injection, patients underwent a period of training with placebo (0.35 ml) lled auto-injectors and required demonstration of adeptness in self-administration prior to home use. Patients who failed to respond adequately to in clinic VIP=P-1 were given two active auto-injectors for self-administration at home. The rst contained VIP=P-1 and the second VIP=P-2 (25 mg VIP and 2.0 mg phentolamine in 0.35 ml). Patients were instructed to keep a diary and report any response to the injection and any adverse events. Patients who achieved a grade three erection to in home VIP=P-1 were told to immediately notify their investigator and were allowed to enter the Placebo Controlled Phase, and the unused VIP=P-2 autoinjector was returned. Patients who did not achieve a satisfactory response to VIP=P-1 were told to administer the VIP=P-2 dose after a minimum 36 h waiting period. Patients who achieved a grade three erection with VIP=P-2 were allowed to enter the Placebo Controlled Phase; any other response was considered a treatment failure, entered into the study data and analyzed, the patient being withdrawn from the study. The pre- lled auto-injector utilized exclusively in this study was supplied by Senetek plc. This singleuse, disposable device is small ( ) and utilizes an ultra ne 29 gauge needle that is not visible to the patient prior to injection. The device delivers 0.35 ml in <5 s, the injection beginning only after the needle depth of 1.0 cm is achieved. Upon entering the Placebo Controlled Phase, patients were assigned consecutive numbers. Medication was supplied to the investigator in cartons labeled with each patient's unique identi cation number. Each carton contained two packages of six auto-injectors, each of which was randomized to include one placebo and ve active, the dose determined by the investigator and based upon the patient's response in the Dose Assessment Phase. Medication was dispensed to the patient in packages of six auto-injectors with instructions to use the auto-injectors in the exact numbered order, not to use any auto-injector within 36 h of previous use, and not to inject more often than three times in a

3 week. The response to each numbered injection, its duration and any adverse event was recorded by the patient in a diary. Following the twelfth injection, or after six months, all patients were seen by the local investigator, who reviewed diary entries, performed a complete physical examination, biochemical and haematology testing and ECG, following which the second package of twelve auto-injectors was dispensed. The examinations were repeated after using the second set of auto-injectors. On completion of the rst 12 injections (Placebo Controlled Phase I), a dose assessment was made by the patient and the investigator, and if necessary the dose adjusted prior to entry into the Placebo Controlled Phase II, which in all other respects was identical to the rst phase. For a patient to have completed each phase of the study, a minimum of nine injections and corresponding diary completions was required within each six month period. Overall results were assessed with a follow-up questionnaire completed by the patient and their partner. Adverse events were reported for each injection for which there was a corresponding diary entry. Results Dose assessment phase Table 1 Results of dose assessment phase Cumulative response rate (%) Characteristic VIP=P-1 VIP=P-2 AgeÐPatient (y) Mean 58.5 Range 27 ± 79 Aetiology of ED n Arteriogenic % 80.6% Diabetes mellitus % 85.5% Neurogenic % Mixed % 82.9% Total % 83.1% Previous therapy a (Agents) IC Alprostadil % 81.1% IC Papaverine= % 82.2% phentolamine Other (Yohimbine, % 84.4% vacuum devices) Total % 82.0% (Withdrawal Reason) Poor ef cacy % 73.2% Poor tolerability % 84.4% Poor acceptability % 94.4% Other % 91.3% Total % 82.6% a Patients may have been previously treated with more than one therapy. Patients may have recorded therapy but not always the reason for withdrawal. 93 During the dose assessment phase 255 out of 304 men had a grade three erection for an overall response rate of 83.9% with either of the two doses. Prior to commencing the Placebo Controlled Phase, 15 responders withdrew, seven for adverse events and eight for the following miscellaneous reasons: namely, four lost to follow-up, one had spontaneous resolution, and three had withdrawn consent. Patient demographics and response rate by aetiology for the remaining 289 patients is shown in Table 1, with the overall cumulative response being 83.0%. The total response by aetiologic category was calculated by the summation of responses from each successive injection, patients receiving three injections without obtaining a Grade 3 erection being considered non-responders. The lapse from time of injection to onset of tumescence was: 0 ± 5 min (41%), 5 ± 10 min (27%), 10 ± 20 min (14%), and 20 ± 30 min (2%) with 15% not having been recorded. During this phase, it was also noted that 82% of patients who withdrew from previous therapies and 73.2% who withdrew for the reason of poor ef cacy, responded to VIP=P-1 or VIP=P-2. (See Table 1) Placebo-controlled phase Two hundred and forty patients were enrolled into Placebo Controlled Phase I with the following aetiologies: arteriogenic (58), DM (53), Neurogenic (3) and Mixed (126). Forty- ve of these patients never used a single injector and were withdrawn as non protocol compliant. 172 out of 195 (88.2%) patients received at least one injection of active and placebo and could be analyzed under intention to treat, 133 (68.2%) met full Phase I protocol completion criteria. Reasons for withdrawal for the 62 patients who initiated therapy in Phase I were as follows: adverse events (2), non-compliance to protocol (22), withdrew consent (12), unknown (26). Of the 133 patients completing Phase I, 126 (94.7%) participated in Phase II suf ciently for intention to treat analysis, 105 (79.3%) completing the full Phase II protocol. Results are detailed in Table 2. Differential analysis of active vs placebo by underlying aetiology demonstrated no statistical difference in ef cacy of either VIP=P-1 or VIP=P-2. Patients recorded the erection duration; however, this included for many patients the period between onset of tumescence and detumescence, consequently some apparently long durations of erections were recorded. The median duration of erection for Phase I and II are as follows: Phase IÐVIP=P-1: 50 min, placebo 30 min; VIP=P-2: 30 min, placebo 15 min; and Phase IIÐVIP=P-1: 60 min, placebo 45 min; VIP=P-2: 40 min, placebo: 27.5 min. There were only two episodes of priapism throughout all phases of the study and these are described with adverse events below.

4 94 Table 2 Protocol and intention to treat ef cacy analysis Grade III responses Signi cance of difference Drug Study phase patients Active Placebo Active (%) Placebo (%) between active and placebo a Protocol analysis VIP=P (73.5%) 27 (12.7%) P < (77.2%) 18 (11.3%) P < Overall (75.1%) 45 (12.1%) P < VIP=P (62.0%) 4 (9.5%) P < (72.3%) 4 (11.1%) P < Overall (6.5%) 8 (10.3%) P < Intention to treat analysis VIP=P (72.2%) 31 (13.1%) P < (75.0%) 24(12.6%) P < Overall (73.7%) 55 (12.9) P < VIP=P (65.3%) 11 (15.9%) P < (75.0%) 5 (10.6%) P < Overall (69.1%) 16 (13.7%) P < a Wilcoxon signed rank test (95% CI). Overall satisfaction with the drug and the autoinjector was assessed by patients by means of a follow-up questionnaire. Patients and their partners completed a similar questionnaire on quality of life, these results are shown on Table 3. Overall, 85% or more of patients were satis ed or very satis ed with the drug and greater than 95% with the auto-injector. Over 81% of patients and 76% of their partners stated that their lives had been improved or greatly improved by the treatment. Adverse events During the Dose Assessment Phase, 80 of the 304 patients (27.3%) reported a total of 176 adverse events, the majority of which were categorized as mild to moderate. Related events included dizziness (7), headache (7), tachycardia=palpitations (14), ushing (24), bruising (13), bleeding at injection site (9), and priapism (1). Seven patients withdrew from further study due to the following adverse events: ushing (1), ushing and palpitations (2), palpitations (3), dizziness (1), and priapism (1). The onset of tumescence in the patient who developed priapism was 4 h after in clinic injection and responded to simple aspiration. During the Placebo Controlled Phase a total of 3267 injections for which diary entries were completed in 172 patients were analyzed for adverse events, these being summarized in Table 4. The major adverse event was facial ushing calculated per patient or injection, respectively, as follows: VIP=P %, 33.8%; VIP=P %, 33.9%, placebo 13.1%, 6.9%. The majority of patients described this as mild and transient and resulted in no patients withdrawing from this phase of the study. Urethral bleeding (usually characterized by a drop of blood at the meatus) was noted in 1.5%, 1.1% and 1.0% for VIP=P-1, VI:P=P-2 and placebo, Table 3 Patient and partner satisfaction results Responder respondents Score Assessment Patient Partner Phase eligible % % Description Satisfaction with drug X 1 133= Satis ed or very satis ed X 2 105= Satis ed or very satis ed Satisfaction with auto-injector X 1 146= Satis ed or very satis ed X 2 111= Satis ed or very satis ed Quality of life X 1 147= Improved or greatly improved X 1 114= Improved or greatly improved X 1 138= Improved or greatly improved X 2 108= Improved or greatly improved X refers to the responder.

5 Table 4 Related adverse events in placebo controlled phase 95 VIP=P-1 VIP=P-2 Placebo Patients % by Patients % by Patients % by Total number of injections and patients Local Bruising % % % Bleeding at % % % injection site Pain on % % % injection Pain post 0 0.% % % injection Urethral % % % bleeding Priapism % % % Systemic Flushing % % % Dizziness % % % Headache % % % Palpitation % % % Tachycardia % % % respectively. However, when comparing both Placebo Controlled Phases, the overall incidence of urethral bleeding fell from 1.8% in Phase I to 0.9% in Phase 2, suggesting a learning curve in self injection therapy. Only two patients withdrew because of an adverse event; death (1) and priapism (1). The patient who developed priapism had a similar episode following previous IC papaverine (60mg) therapy and responded to aspiration. The single death occurred 10 d after the last injection in a 70 y old with a history of two previous myocardial infarctions. One additional 65 y old patient with diabetes and hypertension suffered an acute myocardial infarction two months after his last injection of VIP=P. Discussion The role of intracavernosal administration of vasoactive compounds in the treatment of male ED has been well recognized. Papaverine, phentolamine and alprostadil, used alone, or in various combinations, have met with varying success rates, but have been accompanied with undesirable adverse effects, primarily priapism, brosis, and post-injection pain, the latter being particularly associated with alprostadil. 9 Although in earlier studies VIP used alone was only associated with tumescence when accompanied by visual or vibratory stimulation, 6 Gerstenberg et al 8 reported encouraging results in 52 patients utilizing 1380 self-injections of 30 mg VIP combined with 0.5 ± 2.0 mg phentolamine. All patients reported achieving an erection suf cient for intercourse, and there was no reported priapism, brosis, pain, or serious complications, though these authors subsequently observed that the 0.5 mg phentolamine combination was largely ineffective. These ndings were reinforced by McMahon's 10 recent report on the treatment of 20 men with ED utilizing VIP=P-1 and VIP=P-2 supplied in ampoules, in which he observed an overall ef cacy of 80% and a low side effect pro le (15% facial ushing, 10% injection pain). The auto-injector used in this study was designed to improve patient acceptance of self-injection as well as safety. The needle is not visible prior to injection therefore reducing patient anxiety. The pre- lled, single dose eliminates the need for reconstitution which requires a degree of dexterity and visual acuity that can challenge the diabetic and older individuals, as well as eliminating dosing errors and contamination from improper reconstitution. Intentional re-injection would be required for potential overdosing. The injection is virtually painless due to the small needle size and injection rate. Injection of the medication does not commence until needle depth has been reached, reducing the possibility of subcutaneous injection, the auto-injector is disposable and not reusable. This current study was designed in two phases: the dose assessment phase mimics clinical practice, in which an attempt is made to rule-out psychogenic, concomitant medication and hormonal imbalance aetiologies, prior to initiating appropriate

6 96 therapy and dosing determination. The cumulative response rates by aetiology of 80.6% in arteriogenic, 85.5% in diabetic, 100% in neurogenic, and 82.2% in mixed clearly attest to the ef cacy of VIP=P-1 and VIP=P-2. In addition, 82.0% of patients withdrawing from previous therapies responded, including a response rate of 81.1% from previous IC alprostadil, 82.2% from previous IC papaverine=phentolamine, and 73.2% of those who had withdrawn because of poor ef cacy. Whether analyzed per protocol or intention to treat, there was powerfully signi cant difference (P<0.001) between active and placebo when examining separately both VIP=P-1 and VIP=P-2. In the per protocol analysis of VIP=P-1 and VIP=P-2, active injections produced an erection suitable for intercourse in 75.1% and 66.5% respectively, with corresponding placebo responses of 12.2% and 10.3%. In the intention to treat analysis of VIP=P-1 and VIP=P-2, active injections produced an erection suitable for intercourse in 73.7% and 69.1% respectively, with corresponding placebo responses of 12.9% and 13.7%. The placebo response is remarkably low and clearly re ects the avoidance of overtly psychogenic patients in this study. There was little remarkable about the adverse event pro le. Although the majority of patients (74.4%) in the Placebo Controlled Phase noted mild transient ushing, the incidence per injection decreased to 33.8% and this resulted in no patients having withdrawn from this phase of the study. In 3 out of 304 patients (1.0%) in the Dose Assessment Phase facial ushing contributed to withdrawal, whereas the overall incidence of ushing was 7.9%. Transient ushing, per se, is not an uncommon feature of injection therapy, but was particularly prevalent in this study by virtue of the forward nature of the questions asked to record adverse occurrences for each administered injection. In view of the vasodilatory properties of the test medication and the high incidence of cardiovascular disorders in the patient population, the frequency of other cardiovascular related events (dizziness, tachycardia and palpitations) was exceptionally low, other events were uncommon. Minor urethral bleeding was noted following 1.4% of active injections and 1.0% of placebos and is possibly related to injection technique. There were two episodes of priapism in 304 patients, the incidence per injection in the Placebo Controlled Phase being 0.05%. Of signi cance no patient complained of post-injection pain throughout the entire study and no brosis was noted. The incidence of these observations was surprisingly low in view of the higher frequencies reported in other studies involving alternative pharmacotherapy. 9 Withdrawals from this study were consistent with those reported elsewhere. 9 Only 16.1% withdrew because of treatment failure and 2.9% patients due to adverse events. The 28.8% rate of non compliance in Placebo Controlled Phase was indicative of the demands of the study protocol. Conclusion This study unequivocally demonstrates the ef cacy and safety of the VIP=phentolamine combination (INVICORP TM ) in the treatment of non-psychogenic ED. There was no evidence of tachyphylaxis when comparing the ef cacy in the two Placebo Controlled Phases, for an overall period of 12 months. Both patient and partner satisfaction were remarkably high with a clear improvement in overall quality of life. Acknowledgements We wish to acknowledge the contribution made to this study by the following additional investigators: Mr MC Foster, Consultant Urologist, Good Hope Hospital, Sutton Cold eld, UK; Mr C Charig, Consultant Urologist, Epsom General Hospital, Epsom, UK; Mr M Hughes, Director of Urology, The Queen Elizabeth Hospital, Birmingham, UK; and Mr C Chapple, Urology Department, The Hallamshire Hospital, Shef eld, UK and nancial support from Senetek plc. References 1 Ottesen B, Wagner G, Virag R, Fahrenkrug J. Penile erection: possible role for vasoactive intestinal polypeptide as a neurotransmitter. Br Med J 1984; 228: 9 ± Gu J et al. Decrease of vasoactive intestinal polypeptide (VIP) in the penises from impotent men. Lancet 1984; 315 ± Ehmke H, Junemann KP, Mayer B, Kummer W. Nitric oxide synthesis and vasoactive polypeptide colocalization in neurons innervating the human penile circulation. Int J Impot Res 1995; 7: 147 ± Fahrenkrug J. VIP and autonomic neurotransmission. Pharmacol Ther 1989; 41: Juneman KP et al. The role of vasoactive intestinal polypeptide as a neurotransmitter in canine penile erections: A combined in vivo and immunohistochemical study. J Urol 1987; 138: 871 ± Wagner G, Gerstenberg C. Intracavernosal injection of vasoactive intestinal polypeptide (VIP) does not induce erection in man per se. World J Urol 1987; 5: Roy JB, Petrone RL, Said SI. A clinical trial of vasoactive intestinal peptide to induce penile erection. J Urol 1990; 143: 302 ± Gerstenberg TC, Metz P, Ottesen B, Fahrenkrug J. Intracavernous self-injection with vasoactive intestinal polypeptide and phenotolamine in the management of erectile dysfunction. J Urol 1992; 147: 1277 ± Manning M, Junemann KP. Pharmacotherapy of erectile dysfunction. In: Hellstrom, WJG, ed. Male Infertility and Sexual Dysfunction. Springer-Verlag: New York, 1997, pp 440 ± 451.

7 10 McMahon CG. A pilot study of the role of intracavernous injection of vasoactive intestinal peptide (VIP) and phentolamine mesylate in the treatment of erectile dysfunction. Int J Impot Res 1996; 8: 233 ± Gerstenberg TC. A novel auto-injector for the administration of vasopotinða new treatment for erectile dysfunction Int J Impot Res 1995; 7: S