Current perspectives on intracavernosal pharmacotherapy for erectile dysfunction

Transcription

1 (2000) 12, Suppl 4, S91±S100 ß 2000 Macmillan Publishers Ltd All rights reserved /00 $ for erectile dysfunction 1 * 1 Neuer Jungfernstieg 6a, Hamburg, Germany Thirty percent of males af icted with erectile dysfunction (ED) do not respond to oral drugs, another 15% reveal contraindications to currently available therapy, and a subset of patients actually prefer injection therapy due to its predictable short time-to-onset of erection and reliable rigidity compared to oral drugs. This paper provides both a historical and current perspective on intracavernosal therapy and reviews the popular injectable therapeutic agents that are currently used for the treatment of ED. Emphasis is placed on the ef cacy, mechanism of action and side effect pro les of approved and experimental injectable pharmacotherapy for ED. International Journal of Impotence Research (2000) 12, Suppl 4, S91±S100. Keywords: erectile dysfunction; penis, injectable pharmacotherapy History of intracavernosal pharmacotherapy It is generally acknowledged that the onset of intracavernosal pharmacotherapy began with the publication by Virag in 1982 concerning the erection-including properties of papaverine. 1 Further important early publications in this eld described the erectile potency of the alpha-adrenoceptor blockers phenoxybenzamine and phentolamine 2 and the combination of papaverine=phentolamine for autoinjection of the cavernous bodies. 3 In fact, the work of Zorgniotti and Le eur 3 marked the breakthrough of self-injection therapy as a completely new tool in the management of male impotence. One year later, at the Second World Meeting on Impotence in Prague, PGE 1 (alprostadil) was introduced for self-injection therapy, by both Adaikan 4 and Ishii. 5 Two years later, these favorable preliminary reports on alprostadil were con rmed in larger clinical trials by Porst and Stackl, respectively. 6,7 Alprostadil has since become the global leader in intracavernosal pharmacotherapy. It is marketed by two different companies as alprostadil sterile powder (Caverject 1, Pharmacia & Upjohn) and alprostadil alfadex (Edex 1, Viridal 1, Schwarz Pharma AG); both agents exhibit an identical ef cacy- and side-effect spectrum. Parallel to the development of monotherapy with alprostadil, the trimix combination of PGE 1 = papaverine=phentolamine began to gain broad *Correspondence:, Neuer Jungfernstieg 6a, Hamburg, Germany. PORST20354@aol.com acceptance following the publication of Goldstein in 1990, especially in the USA. 8 In 1989, the relatively selective alpha 1 -adrenoceptor blocker moxisylyte (thymoxamine) was introduced by Buvat et al. 9 Moxisylyte was rst marketed in France as Icavex 1 and Erecnos 1 and the latter has since been withdrawn due to unpro table sales rates. The erectile potential of other vasoactive drugs, eg, calcitonin-gene-relatedpeptide (CGRP), linsidomine (syn. SIN-1), vasoactive intestinal polypeptide (VIP) or sodium nitroprusside, was temporarily publicised but none of the cited compounds gained market approval status. 10 ± 15 In this context, two exceptions deserve mention. Firstly, the combination of papaverine=phentol-zamine (Androskat 1 ) was of cially approved by the local health authorities of several countries and is marketed in the Benelux and German-speaking countries of Europe. Secondly, the combination of VIP=Phentolamine (Invicorp 1 ), originally studied by Gerstenberg et al in 1992, was of cially approved in Denmark but not marketed and can be prescribed on an individual `named' basis in UK. 16 Pharmacological considerations of marketed intracavernosal vasoactive drugs (Table 1) Papaverine Papaverine belongs to the family of non-selective phosphodiesterase inhibitors. Depending on the target-tissue-related dominance of the respective

2 S92 Table 1 Biological effects of marketed vasoactive drugs for self-injection therapy in erectile dysfunction Compound Site of action Effect on cavernous tissue Papaverine Phosphodiesterases (non-selective) camp=cgmp: L-type Ca 2 channels (inhibition) Intracellular Ca 2 ; Angiotensin II secretion (inhibition) Smooth muscle tone; Phentolamine a1 ± 2 adrenoceptors (blockade) Noradrenaline effects; Maxi K channels (stimulation) Hyperpolarization NO-synthase (stimulation) Intracellular NO: Alprostadil (PGE 1 ) Adenylate cyclase camp: Presynaptic a 1 receptors (inhibition) Noradrenaline release; Angiotensin II secretion (inhibition) Smooth muscle tone; Maxi-K channels (stimulation) Hyperpolarization TGF-b 1 (inhibition) Collagen-production; Moxisylyte a 1 -adrenoceptors Noradrenaline effects; VIP Adenylate cyclase camp: phosphodiesterases (PDE 1 to PDE 10 ), administration of papaverine results in camp or cgmp accumulation. Papaverine also appears to exert inhibitory effects on L-type voltage-dependent Ca 2 channels and on angiotensin II secretion. 17,18 Phentolamine receptor activity with LDL degradation, positive effects on rheologic properties and blood viscosity, decrease of cholesterol deposition into the arterial wall and inhibition of platelet aggregation resulting in atherosclerosis prevention, inhibition of noradrenaline-induced lipolysis, and positive effects on cellular nutritive metabolism of ischemic tissue due to decrease of the lactate to pyruvate ratio. Phentolamine represents a non-selective alpha 1±2 - adrenoceptor blocker with impact both on pre- and post-synaptic alpha-adrenoceptors. In the caninemodel, phentolamine administration resulted in a 50 ± 100% increase of penile arterial blood ow without an additional in uence on the venoocclusive mechanism. 19 Beyond this dominant pharmacological action, it was shown that phentolamine has potassium channel opening, endothelin antagonist and direct nitric oxide (NO)-synthase activating properties. 20,21 Alprostadil (PGE 1 ) Proven pharmacological and biological effects of alprostadil in the context of ED are: (1) Stimulation of adenylate cyclase with generation of camp. 22 (2) Inhibition of noradrenaline release at alpha 1 - adrenoceptors via presynaptic prostaglandin receptors. 23 (3) Inhibition of angiotensin II secretion. 18 (4) Stimulation of maxi-k ion channels resulting in membrane hyperpolarization. 24 (5) Neuromodulatory activities on the medial preoptic area. 25 (6) Inhibition of the transforming growth factor b 1 (TGF-b 1 ) resulting in anticollagen and therefore anti brotic effects. 26 In addition, prostaglandins and especially PGE 1 have demonstrated other biological effects including: 27 decrease of low density lipoprotein (LDL) Moxisylyte (thymoxamine) Moxisylyte has alpha 1±2 -receptor blocking properties with preference for alpha 1 -adrenoceptors. 9,28 Moxisylyte is a pro-drug that is immediately converted to ve metabolites, four of which exhibit biological activities. 28 Vasoactive intestinal polypeptide Vasoactive intestinal polypeptide (VIP) is a naturally occurring neurotransmitter which is co-localized with NO at the nonadrenergic, noncholinergic (NANC) nerve-terminals in cavernous tissue. 29 VIP stimulates adenylate cyclase and results in the generation of camp. In animal studies, intracavernosal VIP induced tumescence and partial erection, which was primarily due to decreased penile venous out ow and to a minor extent, increased arterial in ow. 30 Clinical results of self-injection therapy with marketed vasoactive drugs Papaverine Papaverine is extremely stable in solution and loses only 10% of its potency over a 4 y period. Commercial papaverine solutions have a low ph between 3

3 and 3.7; at ph values above 5, the compound precipitates. After intracavernosal application, maximal plasma levels are reached within 10 to 30 min and the half-life is approximately 1 to 2 h. 31 Papaverine is extensively metabolised in the liver. Reported ef cacy rates with dosages between 30 and 110 mg varied between 27 and 78% and were dependent on dosage and the patient population investigated. 27,32,33 A literature analysis of 19 publications that included 2181 patients overall demonstrated that papaverine produced an average response rate of 61% in in-of ce testing. 27 The most important side-effects were priapisms in 3 ± 18.5%, which mostly occurred during the titration phase. Fibrotic alterations were seen in 5 ± 30% of patients with an average of 5.7% in 15 retrospective studies (Table 2). 27,32,33 In long-term self-injection trials, papaverine produced liver enzyme elevations in 1.6% of patients, evidence of hepatotoxic potential. 27 In animal studies, papaverine resulted in the highest brosis-rate among all the investigated vasoactive drugs. For this reason, monotherapy with papaverine has been discontinued in most industrialized countries. 34,35 However, due to its considerably low cost, self-injection monotherapy with papaverine still continues in many developing countries. Papaverine=phentolamine (Androskat 1 ) Table 2 Side-effects of vasoactive drugs in retrospective studies 27 No. patients (No. of Priapism Fibrosis Pain Liver enzymes: Drug publications) (%) (%) (%) (%) Papaverine (15) Pap.=Phentol (22) PGE (Alprostadil) (10) The combination of papaverine=phentolamine gained world-wide popularity with the publication of Zorgniotti and Le eur in The ph values of marketed Androskat 1 solution varies between 3.1 and 3.5 and is stabile for 2 ± 3 y. The current cost in Europe for 52 ml ampoules containing 15 mg papaverine and 0.5 mg phentolamine per ml is approximately 60 US dollars. Combining the camp=cgmp accumulating effects of papaverine and the alpha-adrenoceptor blocking properties of phentolamine results in an increased average response-rate up to 60 ± 70% observed during in-of ce testing. 36 With home use, response rates as high as 90% have been reported. 37 Prospective long-term comparison studies of papaverine alone to the mixture of papaverine= phentolamine conducted under good clinical practice guidelines are not available. In valuable retrospective studies that have considered important issues related to self-injection therapy, ie, success-, side-effect- and drop-out rates, global ef cacy rates were 68.5% in diagnostic use in 3016 assessed patients. 27 Frequent side-effects were similar to those of papaverine. Priapisms were reported in 6 ± 15% and brotic alterations occurred in an average of 12.4% of patients treated (see Table 2), although in some studies, brotic complications occurred in considerably higher percentages of 18 ± 57%. 27,37 ± 40 Hepatotoxic effects with elevation of liver enzymes were reported in an average of 5.4% (43=799) of patients in one review of varied publications. 27 Alprostadil syn. PGE 1 (Caverject 1 and Edex 1 or Viridal 1 ) The pharmacologic pro le of alprostadil is summarized in Table 6. Intracavernously (i.c.)-injected alprostadil may be partially metabolized by the enzymes of the cavernous tissue and although considerable amounts reach the systemic circulation, it is metabolized during the passage through the lung. 41,42 It is therefore not surprising that the half-life of i.c.-injected PGE 1 is only 30 ± 60 sec. In contrast to papaverine or to the papaverine= phentolamine combination, large world-wide prospective studies have been conducted in accordance with good clinical practice guidelines for both alprostadil preparations (alprostadil sterile powder ± Caverject 1 and alprostadil alfadex ± Viridal 1 or Edex 1 ) and long-term follow-up of 4 ± 5 y is available. 43 ± 46 A review of these large studies shows that the ef cacy-rate of alprostadil during in-of ce titration varied between 70 and 75% in more than patients. 27 Meanwhile, the author's experience in more than patients with ED undergoing a 20 mg intracavernosal PGE 1 injection test resulted in an in-of ce ef cacy rate of 72%, con rming the data from the literature. In a variety of prospective self-injection trials, the success-rates (de ned as successful coitus per injection) varied between 89 and 96%; this is higher than any reported ef cacy rate among all the available marketed vasoactive drugs (Table 3). 43,46 Typical side-effects with alprostadil self-injection therapy are the occurrence of penile pain in 8 ± 52%, however, with long-term follow-up and individual dose adjustment, pain-rates decreased to 1 ± 11%. 45,46 Priapisms were almost exclusively observed in the titration-phase and varied between 0.25 and 1% (Table 2). 27,43,46 In the two major trials with both alprostadil-preparations, penile brotic alterations were encountered in 7.5 ± 11.7% of S93

4 S94 Table 3 Success-rates of self-injection therapy with alprostadil (Viridal 1 =EDEX 1 ) in the European prospective 4 year trial patients during the course of the 4 ± 5 y of long-term follow-up. 45,46 The prevailing majority of brotic alterations were small nodules which did not interfere with erectile function or vaginal penetration (Table 4). Between 33 and 47% of these penile broses healed spontaneously, suggesting that the incidence of persistent penile broses in patients on long-term self-injection therapy is between 5 and 7%. 43,44,46 Similar to self-injection trials with all vasoactive drugs, the drop-out rates in the alprostadil trials were relatively high, eg, 55% after 18 months in the alprostadil-sterile powder (Caverject 1 ) study and 54% after 24 months in the alprostadil-alfadex (Viridal 1 ) trial. 44 ± 46 After 4 y of follow-up, 33% of patients continued therapy with alprostadil alfadex and 22% continued regular therapy with alprostadil sterile powder after 5 y of follow-up. 45,46 The satisfaction-rates of both male and female partners in the group who continued were high and ranged between 91 and 95%. 46 In addition, between 78 and 89% of patients and their female sexual partners stated that the self-injection therapy markedly improved both the self-esteem of the af icted males and the partner-relationships (Table 5). 46 Moxisylyte (Icavex 1 ) Year follow-up No. injections No. successful coitus Percent 1st year nd year rd year th year Total Table 5 Impact of alprostadil (Viridal 1 ) self-injection therapy on self-esteem and partnership: results of a prospective, multicenter, 4 year trial in 162 patients 46 Year Completers no. of patients Positive impact on self-esteem (%) Moxisylyte (Table 7), an alpha-adrenergic blocking compound with preference of alpha 1 -adrenoceptors was investigated in comparison to alprostadil alfadex in a prospective trial. 47 The ef cacy rates for 20 mg alprostadil were 75% (56=75) con rmed with a positive Buckling-test and increased to 85% when the same patients performed self-injection therapy at home. The comparable ef cacy rates after 20 mg moxisylyte were 40% (32=81) for the Buckling-test and 61% (37=61) for the self-injection group. Claimed advantages of moxisylyte in other studies were the low risk of priapism (< 1%) and broses (< 2%) but these low side-effect rates were put into perspective by low ef cacy rates if compared to alprostadil or papaverine=phentolamine. 47 ± 49 In addition, in all published moxisylyte trials, clinically relevant drops in blood pressure accompanied by orthostatic symptoms and dizziness were described in 5 ± 8% of patients. 48,49 Moxisylyte was originally marketed under two different trade names ± Icavex 1 and Erecnos 1, but due to its low market acceptance, the latter one was withdrawn some time ago. VIP=phentolamine combination (Invicorp 1 ) Positive impact on partnership (%) As VIP alone injected i.c. in volunteers did not result in rigid erections, a combination of VIP and phentolamine was developed for self-injection therapy (Table 8). 13,50 In an early prospective 6 month trial with 30 mg VIP and 0.5 ± 2 mg phentolamine in 52 patients who underwent a total of 1380 scheduled injections, no priapisms, pain or broses were reported. 16 In a larger prospective trial with 289 patients (mean age ˆ 58.5 y), 77% responded with grade 3 erections, considered by the investigators to be suf cient for intercourse. 51 In this study, two priapisms (0.6%) were observed. Final results of the UK multicenter placebo-controlled Invicorp 1 -trial were published by Sandhu in 1999 and are summarized in Table In this prospective trial, the total drop-out rate (calculated from the date of rstpatient-in and last-patient-out after 6 months was 65.5% (199=304) and was thus considerably higher than in all other alprostadil injection trials. 43,46,52 There were no comments about the reasons for this high drop-out rate. 43 ± 45 Table 4 Penile brosis in prospective self-injection trials with alprostadil preparations Brand name No. patients Follow-up (months) Fibrosis (all) Outcome (spontaneaus healing) Nodules= plaques Deviations Cavernous brosis Viridal 1 =Edex % (19) 47.% (19) Caverject 1 = % (34) NA NA NA NA Europe % (26) NA NA NA NA Caverject 1 =US % (51) 33.%

5 The greatest advantage of the Invicorp 1 preparation is its availability in a ready-for-use automatic single injection device equipped with a 29 gauge needle (Figure 1). This device may be preferred by many patients because the necessity to reconstitute alprostadil preparations and the requirement for manual injections is avoided. The author's personal experiences with the VIP=phentolamine combination are limited to 25 patients in whom the Invicorp 1 1=2-combination Table 6 Pharmacological pro le of alprostadil (Caverject 1 or Virdal 1 =Edex 1 ) Sites of action Adenylate cyclase? camp: Presynaptic on alpha 1 adrenoceptors? noradrenaline; Angiotensin secretion;?muscle tone; Stimulation Maxi-K -channels?hyperpolarization Neuromodulation preoptic area Inhibition TGF b 1?Collagen; Dosages 2.5 ± 40 mg Pharmacokinetics Metabolism intracavernosal and during lungpassage Half-life 30 ± 60 sec PGE 0 active metabolite Ef cacy 70 ± 80% Side-effects Pain 10 ± 20%, priapism < 1%, brosis 5 ± 11% Target-group Non-responders to or contraindications for oral (Sildena l, Apomorphine) therapy, In combination with oral drug therapy for severe ED-patients (salvage therapy) Table 7 Pharmacologic pro le of moxisylyte (thymoxamine) (Icavex 1 ) Site of action Alpha 1 -adrenoceptors (blockade) Impact on erection Sympathetic tone; Smooth muscle tone; Dosages 10=20 mg Pharmacokinetics Four active metabolites Excretion with urine Ef cacy Low (20 ± 40%) Side effects Priapism < 1%, brosis < 2% Hypotonic reactiondizziness 5 ± 8% Target group Patients with contraindications to sildena l and inconvenient painful erections to PGE 1 was compared against Alprostadil 20 mg (Caverject 1 ). In this small series, the response rates for Invicorp 1 were 60% (15=25) compared to 72% (18=25) for Caverject 1. The automatic injection device for the Invicorp 1 -combination represents a very convenient and user-friendly solution for intracavernosal injections. As mentioned above, Invicorp 1 is approved but not marketed in Denmark and can be prescribed on an individual namedpatient basis in the UK. The dosages for Invicorp 1 1 are 25 mg VIP=1 mg phentolamine or 2 mg phentolamine for Invicorp 1 2, respectively. Clinical experiences with experimental vasoactive drugs (Table 10) Linsidomine (SIN 1) Linsidomine belongs to the family of NO-donors and generates NO non-enzymatically, subsequently resulting in stimulation of guanylate cyclase and ultimately, cgmp accumulation. Very promising Table 9 Ef cacy-rates of VIP=phentolamine (Invicorp 1 after in-of ce testing plus visual sexual stimulation. Number of patients ˆ 289, mean age ˆ 58.5 years (range 27 ± 79) 52 ED-etiology No. of patients Responders to Invicorp 1 Responders to Invicorp 2 Arteriogenic % 80.6% Diabetes % 85.5% Neurogenic % ± Mixed % 82.9% Total % 83.1% S95 Table 8 Pharmacological pro le of VIP=phentolamine (Invicorp 1 ) Site of Action VIP: Adenylcyclase? camp: Phentol: alpha 1 =alpha 2 receptors (blockade) Maxi-K -channels (activation) Endothelin-antagonism, blockade 5-HT-rec. Dosage 25 mg VIP=1 mg or 2 mg Phentolamine Ef cacy 60 ± 70% mixed ED-patients partially nonresponders to PGE 1 Side effects Flush 70 ± 80%, Priapism < 1%, Fibrosis? Target group Non-responders or contraindications to Apomporphine=Sildena l Principally all ED-patients Figure 1 Automatic single injection device for use with Invicorp 1.

6 S96 Table 10 Non-marketed vasoactive drugs with potential for self-injection therapy Compounds Dosage Results (ef cacy) Conclusion PGE 1 =CGRP 20 mg=5 mg 30% of non-responders to 40 mg PGE1 Salvage-therapy potential Linsidomine (SIN-1) 1 mg 35 ± 74% Inferior to PGE 1 Sodium nitroprusside 300 ± 600 mg 64 ± 84% More side-effects than PGE 1 Triple-drug (trimix) Up to 80 ± 90% Salvage therapy Pap=Phentol=PGE 1 30 mg=1mg=20 mg results with success-rates of up to 69% (78=113) were reported by authors of one study group, 11,53 but could not be con rmed by others. 12,54 In a small comparative trial conducted by the author in 40 patients, the ef cacy-rates (partial or full rigidity) after 1 mg SIN-1 were 35% compared to 82.5% after alprostadil 20 mg. 12 Due to these signi cantly lower ef cacy-rates with admittedly low side effect rates, the compound was not pursued further and did not enter a full clinical trial program toward market development. Sodium nitroprusside Sodium nitroprusside, a NO-donor similar to SIN-1 was evaluated in a comparative trial with alprostadil. 15 In a total of 95 patients, 49% responded with partial and 15% with complete rigidity to 300 ± 400 mg doses of nitroprusside compared to 54% and 20%, respectively, after 20 mg alprostadil. With nitroprusside doses of 600 mg, global response rates of 84% were achieved. Because alprostadil produced better response rates and sodium nitroprusside was incriminated with hypotonic blood pressure reactions in up to 15% of patients, this compound did not enter the phase of multicenter trials. Calcitonin gene related peptide combined with PGE 1 Calcitonin gene related peptide (CGRP) a potent vasodilator, increased penile blood ow and resulted in tumescence but not rigidity when given intracavernously in dosages of 500 mg. 10 In a population of 65 patients, of whom 91% were nonresponders to the papaverine (30 mg)=phentolamine (1 mg) combination, CGRP (5 mg) combined with PGE 1 (10 mg) produced rigid erections in 55%. 55 In another study performed by Schwarzer et al, 30% of patients who failed 40 mg PGE 1 or 80 mg papaverine or the combination of 60 mg papaverine plus 2 mg phentolamine obtained rigid erections following treatment with 5 mg CGRP plus 20 mg PGE Since these reports which appeared nearly a decade ago, no further reports on the validity and safety of the CGRP=PGE 1 combination have been published. From this we can surmise that no pharmaceutical company opted to bring this combination product through the rigorous path toward market approval. Triple-drug (trimix) ± papaverine=phentolamine= PGE 1 Combining the three vasoactive drugs: papaverine, phentolamine and PGE 1 allows one to reduce the dosages of the individual compounds and therefore to lower the risks of side-effects attributed to the respective single drugs while increasing ef cacy rates. Reported dosages of single compounds in several publications varied between 4.4 and 15 mg papaverine, 0.15 and 0.5 mg phentolamine, and 1.5 and 20 mg PGE 1 per ml. 8,57,58 Response rates with the trimix combination in non-selected patients with ED approached 80 ± 90% and were thus 10 ± 15% higher than after 20 mg alprostadil monotherapy. 57,59 Because rigid erections following trimix administration were achieved in 50 ± 62% of socalled non-responders, some authors suggested that trimix would be useful rescue therapy for nonresponders to monotherapy with alprostadil or the combination of papaverine=phentolamine. 8,58 In two prospective trials the trimix combination was preferred by 46 ± 67% of patients compared to 19 ± 23% for the approved alprostadil sterile powder (Table 11). 60,61 In the author's hands, the trimix combination produces a response in 30 ± 40% of non-responders to 40 mg alprostadil. It can easily be easily be prepared by mixing one ampule of Androskat 1 (30 mg papaverine=1 mg phentolamine) with 10 ± 20 mg Caverject 1. Within the context of this extemporaneous reconstitution, it must be considered that the stability of the three drug combination varies with a 30% degradation of the PGE 1 -concentration within 60 days. 62 The trimix self-injection therapy should be reserved for patients who do not respond to alprostadil or for those who experience painful erections after alprostadil alone. Trimix is principally associated with the same side-effect and riskpro le of papaverine and phentolamine, especially in terms of priapism and brosis.

7 Table 11 Intra-individual comparison between alprostadil (Caverject 1 ) and trimix (Pap=Phentol=PGE 1 ) S97 Alprostadil study Trimix study No. preference study No. patients study 1 a 2 b 1 a 2 b 1 a 2 b 1 a 2 b Better ef cacy 22% 14% 37% 62% 41% 24% General preference 23% 19% 46% 67% 31% 14% a Ref. 60; b ref. 61. Special topics on self-injection therapy (Table 12) Patients treated concurrently with anticoagulants The safety of vacuum-therapy and self-injection therapy was compared in a cross-over study of 33 patients being treated with warfarin. 63 In only 11 (1.6%) of 706 documented vacuum applications and three (0.5%) of 605 self-injections, ecchymoses of the penile skin were recorded in patient diaries, indicating that the risk of bleeding, both in vacuum therapy and self-injection therapy, is not higher in the group receiving anticoagulants than it would be in a normal patient population. In the author's clinical practice, following 10 patients treated with the anticoagulant marcoumar for a number of years, no severe bleeding complications have been observed. Therefore, patients treated with anticoagulants (either platelet aggregation inhibitors or warfarin=marcoumar) do not represent contraindications for self-injection therapy provided they are carefully instructed in self-injection techniques and use thin (27 ± 30 gauge) needles. Transplant recipients A study conducted with 26 renal transplant patients did not reveal any increased risk for self-injection therapy, 64 corroborating the author's own experiences with three transplanted patients (two kidneys, one heart). Patients with diabetes mellitus In a comparative trial among 16 diabetic and 29 nondiabetic patients with a mean follow-up of 9.4 months and an average injection frequency of three per month, focal brotic alterations were more frequently observed in the diabetic cohort (25%) compared to 3% in the control-group. 65 This series supports the author's impressions that insulin-dependent diabetics are more prone to the occurrence of brosis than other patient subsets. Managing complications of self-injection therapy Priapism It is generally accepted that priapisms lasting > 6h should be interrupted to prevent tissue damage. Electromicroscopical investigations by Spycher (1986) suggested that irreversible tissue impairment may be occurring after 12 ± 18 h. 66 According to the author's personal observation of more than 150 patients with drug-induced priapisms, direct injection of a sympathomimetic antidote is suf cient if the priapism did not last longer than 12 h. In priapisms longer than 12 ± 18 h duration, it is reasonable to evacuate the entrapped hypoxemic blood with i.c. insertion of a butter y cannula, wait 10 ± 15 min until reoxygenation has established, then inject the adrenergic antidote if it is still deemed necessary. Among all the available adrenergic drugs, etilefrine and phenylephrine have the least impact on cardiac b-receptors and are therefore probably the safest choice. 67 The recommended dosages are 5 ± 20 mg for etilefrine or 0.1 ± 0.5 mg for phenylephrine. It is reasonable to start with a low dosage and massage the penis between the ngers after removal of the needle to promote circulation of the injected antidote within the entrapped blood in the cavernous bodies. If the erection persists after 15 min, a repeat dosage should be injected in to the opposite cavernous body. If this Table 12 Self-injection therapy in special risk-groups Risk-group No. of patients Ref. Conclusion Antikoagulants (Warfarin) No increased risk of bruising Kidney transplant No increased risk of infection or brosis Diabetes mellitus Considerable increased risk of pain Non-diabetes 29 (19% vs 3%), cavernitis (19% vs 0%) and brosis (25% vs 3%)

8 S98 injection also fails, evacuation is advisable. After the antidote injection, blood pressure readings must be taken to detect blood pressure increases, which are successfully treated by oral or fast-acting sublingual nifedipine. Fibrosis As described earlier, most brotic alterations that develop during the course of self-injection therapy are mild or moderate and concern small nodules in the tunica albuginea or in the penile septum (Table 4). In this case, the patient should be re-instructed in the correct injection technique and taught to avoid the affected area when performing future injections. In the event real brotic plaques with or without penile curvature develop, self-injections should temporarily be discontinued for 3 ± 4 months. Following reassessment by an in-of ce intracavernosal injection test, the option of surgical interventions (corporoplasty or penile implant) versus continuation of injection therapy should become evident. If corporoplasty seems unavoidable, sexual intercourse may be resumed after 3 ± 4 months, supplemented again with self-injection or transurethral therapy. Alternatively, and preferably, oral sildena l (Viagra 1 ) or apomorphine SL, may be used, provided that these novel oral compounds prove to be effective in these patients. Future role of self-injection therapy versus oral drug therapy There is no question that, since ef cacious oral drug therapy like sildena l has become available, management of male impotence has dramatically improved. Notwithstanding the observation that most ED patients can be managed by oral drug therapy, there will be a considerable number of candidates for self-injection therapy. These candidates will likely be represented in the following categories: (1) About 15 ± 20% of all patients with erectile dysfunction show contraindications to sildena- l-like nitrate or NO-donor medications, ie, those with severe cardiac disease or those treated with antihypertensive polypharmacotherapy. (2) According to the author's personal experiences, about 15% of alprostadil responders are sildena l non-responders. In particular, this observation will be true in patients who are likely to have cavernous nerve impairment (pelvic surgery or trauma, insulin dependent diabetes). (3) According to the author's personal experiences and recent reports, 30 ± 40% of patients satis ed with long-term self-injection therapy will stay on this therapy despite sildena l (Table 13). 68 (4) In a considerable number of patients ( 50%) who do not respond to either sildena l or injection-therapy, the combination of both methods has the potential to rescue these patients and preserve them from vacuum therapy or penile implants. 70,71 References 1 Virag R. Intracavernous injection of papaverine for erectile failure. Lancet 1982; 2: Brindley GS. Cavernosal alpha-blockade: a new treatment for investigating and treating erectile impotence. Br J Psychiatry 1983; 143: 332 ± Zorgniotti, AW, Le eur RS. Autoinjection of the corupus cavernosum with a vasoactive drug combination for vasculogenic impotence. J Urol 1985; 133: 39 ± Adaikan PG, Kottegoda SR, Ratnam SS. A possible role for prostaglandin E 1 in human penile erection. In: Abstract Book Second World Meeting on Impotence, Prague, 1986, Abstr Ishii N et al. Therapeutic trial with prostaglandin E 1 for organic impotence. In: Abstract Book Second World Meeting on Impotence, Prague, 1986, Abstr Porst H. Comparative usefulness of prostaglandin E 1, papaverine and papaverine=phentolamine for the diagnosis of erectile dysfunction in 61 patients. Urolog 1988; 27: 22 ± Stack W, Hasun R, Marberger M. Intracavernous injection of prostaglandin E 1 in impotent men. J Urol 1998; 140: Goldstein I et al. Rescuing the failed papaverine= phentolamine erection: a proposed synergistic action of papaverine, phentolamine and prostaglandin E 1. J Urol 1990; 143: 304A. 9 Buvat J et al. Safety of intracavernous injections using an alpha-blocking agent. J Urol 1989; 141: 1364 ± Stief CG et al. Calcitonin gene related peptide: a possible role in human penile erection and its therapeutic application in impotent patients. J Urol 1991; 146: 1010 ± Table 13 Self-injection therapy vs oral sildena l: preference of patients previously familiarized with self-injection technique Ref. No. of patients Drug used Preference of injection therapy Both methods alternating Combination of both methods Alprostadil 28% 13%? or trimix (31) (14) Trimix 48%?? (32) Porst 173 Alprostadil 39% 6% 4% Unpublished (57) (10) (6)

9 11 Stief CG et al. Preliminary report on the effect of the nitric oxide donor SIN-1 on human cavernous tissue in vivo. World J Urol 1991; 9: 237 ± Porst H. Prostaglandin E 1 and the nitric oxide donor linsidomine for erectile failure: a diagnostic comparative study of 40 patients. J Urol 1993; 149: 1280 ± Wagner G, Gerstenberg T. Intracavernosal injection of vasoactive intestinal polypeptide (VIP) does not induce erection in man per se. World J Urol 1987; 5: 171 ± Brock G, Breza J, Lue TF. Intracavernous sodium nitroprusside: inappropriate impotence treatment. J Urol 1993; 150: 864 ± Martinez-Pineiro L et al. Preliminary results of a comparative study with intracavernous sodium nitroprusside and prostaglandin E 1 in patients with erectile dysfunction. J Urol 1995; 153: 1487 ± Gerstenberg TC et al. Intracavernous self-injection with vasoactive intestinal polypeptide and phentolamine in the management of erectile failure. J Urol 1992; 147: 1277 ± Iguchi M et al. On the mechanism of papaverine inhibition of the voltage-dependent Ca current in isolated smooth muscle cells from the guinea pig trachea. J Pharmacol Exp Ther 1992; 263: Kifor J et al. Tissue Angiotensin II as a modulator of erectile function. I Angiotensin peptide content, secretion and effects in the corpus cavernosum. J Urol 1997; 157: 1920 ± Juenemann KP et al. Hemodynamics of papaverineand phentolamine-induced penile erection. J Urol 1986; 136: 158 ± Gupta S et al. Relaxation of corpus cavernosum smooth muscle by phentolamine via a non-adrenergic mechanism. Int J Impot Res 1998; 10(Suppl 1): S Traish A et al. Phentolamine mesylate relaxes penile corpus cavernosum by adrenergic and non-adrenergic mechanisms. Int J Impot Res 1998; 10: 215 ± Paoletti R. Biochemistry and pharmacology of prostaglandin E 1 : Introductory remarks. In: Sinzinger H, Rogatti W, (eds). Prostaglandin E 1 in Atherosclerosis. Springer-Verlag: New York 1986, pp 3 ± Molderings GJ et al. Modulation of noradrenaline release in human corpus cavernosum by presynaptic prostaglandin receptors. Int J Impot Res 1992; 4: Zhang P et al. PGE 1 -induced alterations in Maxi-K -channel activity in cultured human corporal smooth muscle cells. J Urol 1996; 155: 678 A. 25 Moltz H. E-series prostaglandins and arginine vasopressin in the modulation of male sexual behavior. Neurosci Biobehav Rev 1990; 14: Moreland RB et al. PGE 1 supresses the induction of new collagen synthesis by transforming growth factor b 1 in human corpus cavernosum smooth muscle: mechanism of penile ischemia associated brosis. J Urol 1994, 151(Part 2): 413 A. 27 Porst H. Review article. The rationale for prostaglandin E 1 in erectile failure a survey of world-wide experience. J Urol 1996; 155: 802 ± Bressole F et al. Pharmacokinetics of moxisylyte in healthy volunteers after intracavernous injection of increasing doses. Eur J Clin Pharmacol 1996; 49: 411 ± Ehmke H et al. Nitric oxide synthase and vasoactive intestinal polypeptide co-localisation in neurons innervating the human penile circulation. Int J Impot Res 1995; 7: 147 ± Juenemann KP et al. The role of vasoactive intestinal polypeptide as a neurotransmitter in canine penile erection: a combined in vivo and immunohistochemical study. J Urol 1987; 138: 871 ± Hakenberg O et al. Systemic pharmacokinetics of papaverine and phentolamine: comparison of intravenous and intracavernous application. Int J Impot Res 1990; 2: 247 ± Juenemann KP, Alken P. Pharmacotherapy of erectile dysfunction: a review. Int J Impot Res 1989; 1: 71 ± Zentgraf M, Baccouche M, Juenemann KP. Diagnosis and therapy of erectile dysfunction using papaverine and phentolamine. Urol Int 1988; 43: 65 ± Aboseif SR et al. Local and systemic effects of chronic intracavernous injection of papaverine, prostaglandin E 1 and saline in primates. J Urol 1989; 142: Stackl W, Loupal G, Holzmann A. Intracavernous injection of vasoactive drugs in the rabbit. Urol Res 1988; 16: 455 ± Juenemann KP et al. Offene Multicenterstudie zur Differentialdiagnostik der erektilen Dysfunktion mit einer Papaverin= Phentolamin-Kombination (By 023). Urologe A 1988; 27: 2±7. 37 Witjes WPJ et al. The ef cacy and acceptance of intracavernous autoinjection therapy with the combination of papaverine=phentolamine. A prospective multicenter trial in 60 patients. Int J Impot Res 1992; 4: 65 ± Girdley FM et al. Intracavernous self-injection for impotence: a long-term therapeutic option? Experience in 78 patients. J Urol 1988; 140: Lakin MM et al. Intracavernous injection therapy: analysis of results and complications. J Urol 1990; 143: Levine SB et al. Side effects of self-administration of intracavernous papaverine and phentolamine for the treatment of impotence. J Urol 1989; 141: Roy AC et al. Prostaglandin 15-hydroxydehydrogenase activity in human penile corpora cavernosa and its signi cance in prostaglandin mediated penile erection. Br J Urol 1989; 64: 180 ± Cawello W et al. Pharmacokinetics of prostaglandin E 1 and its main metabolites after intracavernous injection and short-term infusion of prostaglandin E 1 in patients with erectile dysfunction. J Urol 1997; 158: 1403 ± Linet OI, Ogrinc FG, and Alprostadil Study Group. Ef cacy and safety of intracavernosal Alprostadil in men with erectile dysfunction. New Engl J Med 1996; 334: 873 ± Linet OI, Ogrinc FG. Penile brosis during 18 months of intracavernosal therapy with Alprostadil (Caverject 1 ). Int J Impot Res 1996; 8: D Linet OJ. Long-term safety of Caverject TM (Alprostadil S.PO, PGE 1 ) in erectile dysfunction (ED). Int J Impot Res 1998; 10: S Porst H et al. Intracavernous Alprostadil Alfadex Ð an effective and well tolerated treatment for erectile dysfunction. Results of a long-term European study. Int J Impot Res 1998; 10: 225 ± Buvat J et al. Double-blind multicenter study comparing Alprostadil alpha cyclodextrin with Moxisylyte chlorhydrate in patients with chronic erectile dysfunction. J Urol 1998; 159: 116 ± Buvat J et al. Reduced rate of brotic nodules in the cavernous bodies following auto-intracavernous injection of moxisylyte compared to papaverine. Int J Impot Res 1991; 3: 123 ± Costal P, Iacovella JA, Bouvet AA. Ef cacy and tolerability of Moxisylyte and placebo injected intracavernously in patients with erectile dysfunction (ED): a multicenter, double-blind study. J Urol 1997; 157(Suppl): Roy JB, Petrone RL, Said SJ. A clinical trial of intracavernous vasoactive intestinal peptide to induce penile erection. J Urol 1990: 143: 302 ± Hackett G et al. The results of a 6 month multi-center placebo controlled study of Invicorp 1 in the treatment of nonpsychogenic erectile dysfunction. J Urol 1998; 159: Sandhu D et al. A double blind, placebo-controlled study of intracavernosal vasoactive intestinal polypeptide and phentolamine mesylate in a novel autoinjector for the treatment of non-psychogenic erectile dysfunction. Int J Impot Res 1999; 11: 91 ± Truss MC et al. Erfahrungen mit dem Stickoxid-donor Linsidomin (SIN-1) in Diagnostik und Therapie der erektilen Dysfunction. Akt Urol 1995; 26: 181 ± Wegner HE, Knispel HH. Effect of nitric oxide-donor Linsidomine chlorhydrate in treatment of human erectile dysfunction caused by venous leakage. Urology 1993; 42: 409 ± Djamilian M et al. Follow-up results of a combination of calcitonin gene-related peptide and prostaglandin E 1 in the treatment of erectile dysfunction, J Urol 1993; 149(Part 2): S99

10 S Schwarzer UJ et al. Calcitonin-gene-related-peptide for therapy of erectile importance. Int J Impot Res 1992; 4: 219 ± Bennett AH, Carpenter AJ, Barada JH. An improved vasoactive drug combination for a pharmacological erection program. J Urol 1991; 146: Derouet H, Meeth M, Bewermeier H. Die Behandlung von SKAT-Non-Respondern mit einem Papaverin=Phentolamin- Prostaglandin E 1 ± Gemisch. Akt Urol 1996; 27: 271 ± Hamid S, Dhabuwala CB, Pontes EJ. Combination intracavernous pharmacotherapy in the management of male erectile dysfunction. Int J Impot Res 1992; 4: 109 ± Kulaksizoglu H et al. Comparison of Alprostadil Sterile Powder (Caverject 1 ) with Trimix. Nomogram and patient satisfaction. J Urol 1997; 157: Mellinger BC, Abbatiello S, Tarnow SL. Tri-mix versus Caverject 1. Effective doses and patient preferences. J Urol 1997; 157: Soli M et al. Chemical evaluation of the stability of vasoactive cocktails with PGE 1, papaverine and phentolamine. Eur Urol 1999; 35: Limoge JP et al. Minimally invasive therapies in the treatment of erectile dysfunction in anticoagulated cases: a study of satisfaction and safety. J Urol 1996; 155: 1277 ± Mansi MF, Ackhudair WK, Huraib S. Treatment of erectile dysfunction after kidney transplantation with intracavernosal self-injection of prostaglandin E 1. JUrol1998; 159: 1927 ± Plekhanov AY, Zhivovo AV, Goryachew LA. Comparative complications rate after Alprostadil intracavernous pharmacotherapy for erectile dysfunction in diabetes mellitus vs nondiabetic patients. Int J Impot Res 1998; 10(Suppl 1): S Spycher MA, Hauri D. The ultrastructure of the erectile tissue in priapism. J Urol 1986; 135: Lee M, Cannon B, Shari R. Chart for preparation of dilutions of alpha-adrenergic agonists for intracavernous use in treatment of priapism. J Urol 1995; 153: 1182 ± Apostolidis A et al. Sildena l vs intracavernosal injection (ICI): ef cacy and preference in patients following ICI for > 1y. 3rd Meeting of the European Society for Impotence Research, 30 Jan ± 2 Feb, 2000, Barcelona, Abstract book, p Brannigan RE et al. Comparison of Sildena l citrate (Viagra 1 ) versus Trimix intracavernosal injection (ICI) as treatment for erectile dysfunction (ED). J Urol 1999; 161: McMahon CG, Samali R, Johnson H. Treatment of intracorporeal injection non-response with Sildena l alone or in combination with triple agent intracorporeal injection therapy. J Urol 1999; 162: 1992 ± Porst H. Editorial Comment to McMahon CG, Samali R, Johnson H. Treatment of intracorporeal injection non-response with Sildena l alone or in combination with triple agent intracorporeal injection therapy. J Urol 1999; 162: 1997 ± Appendix Open discussion following Dr Porst's presentation Dr Pryor: One of the questions, not only with injection therapy, but also with oral therapy and penile implants is, why don't people continue with therapy? I predict the same thing will happen with oral therapies that happened with injection therapy; over half the patients will ultimately drop out within 3 to 5 y. Dr Porst: In the oral trials we have no long-term data. There are long-term trials, two, four and ve years with Caverject. With sildena l or with apomorphine, we have only 18 months. There are ve or six major reasons for drop-outs with any kind of therapy in erectile dysfunction, including loss of partner, cancer, other diseases and unhappiness with injection therapy. Dr Pryor: I agree, it's probably not going to be one answer. In our study published in Urology a few years ago, the patients liked injection therapy, but even if they liked it, they still dropped out. Dr Heaton: These are multi-dimensional problems and we're only providing a solution to one small part of them. We're happy to treat ED pharmacologically, but we're shying away from treating all the other multi-factoral things, so we lose a lot of people. Chronic diseases are managed with chronic therapies and we haven't got into that mode yet. Dr Nehra: Did you say that patients who've had implants that are not working don't want them? Dr Pryor: I've had some patients who get them and they want to know that they can get an erection and that the prosthesis works. They have the opportunity to use it and they just don't. Dr Nehra: Keep in mind that the mean age is 59. You have patient issues, concurrent medical problems, you may have partner problems, marital discord, and all these things need to be in perspective. I discuss this with all my patients preoperatively.