Double-Blind, Placebo-Controlled Safety and Ef®Cacy Trial with Yohimbine Hydrochloride in the Treatment of Nonorganic Erectile Dysfunction

Double-blind, placebo-controlled safety and ef®cacy trial with yohimbine hydrochloride in the treatment of nonorganic erectile dysfunction

H-J Vogt1, P Brandl1, G. Kockott2, JR Schmitz2, MH Wiegand2, J Schadrack3 and M Gierend3

1 Dermatologische Klinik und Poliklinik; 2 Psychiatrische Klinik und Poliklinik, Technische UniversitaÈtMuÈnchen, Munich, Germany; and 3 Medicomp Gesellschaft fuÈr Versuchsplanung und Datenanalyse mbH, Ulm, Germany

This double-blind, placebo-controlled clinical trial of yohimbine hydrochloride included 86 patients with erectile dysfunction and without clearly detectable organic or psychologic causes. The patient group ful®lled all entry criteria; 85 of these could be considered for the Safety- respectively 83 for the Intention-to-treat (ITT)-analysis. Yohimbine was administered orally in a dosage of 30mg a day (two 5mg tablets three times daily) for eight weeks. Patients were seen for follow-up after four weeks' treatment, and for a ®nal visit after eight weeks. Ef®cacy evaluation was based on both subjective and objective criteria. Subjective criteria included improvement in sexual desire, sexual satisfaction, frequency of sexual contacts, and quality of erection (penile rigidity) during sexual contact/intercourse. Objective criteria of outcome were based on improvement in penile rigidity determined by use of polysomnography in the sleep laboratory. Overall Yohimbine was found signi®cantly more effective than placebo in terms of response rate: 71 vs 45%. Yohimbine was well-tolerated: Only 7% of patients rated tolerability fair or poor, and most adverse experiences were mild. There was no serious adverse event.

Keywords: yohimbine hydrochloride; erectile dysfunction; nocturnal penile tumescence (NPT); polysomnography; clinical trial

Introduction gic receptor blocker with some loss of selectivity in higher doses. The drug has been used in numerous in vitro receptor binding studies.3±5 Moreover a large Erectile dysfunction affects at least 1±2% of the number of in vivo pharmacology studies are avail- male population and possibly up to 9%, according able to identify three major sites of action: (1) central to Spector and Carey.1 It is a serious problem for action in various regions of the central nervous many patients and therapeutic options that are system (CNS), where yohimbine blocks cerebral a2- convenient, safe, and effective are desirable. Oral adrenergic receptors, increasing norepinephrine treatment of erectile dysfunction with yohimbine release and turnover; (2) direct peripheral action hydrochloride, an a2-adrenergic receptor blocker, through modulation of autonomic nervous system has been the focus of persistent medical interest. (ANS) tone via presynaptic autoinhibitory a2 recep- Yohimbine is an indol alkaloid derived from the tors; (3) direct action on penile tissues and vascular bark of Corynanthe johimbe, a tree indigenous to smooth muscle cells, blocking postsynaptic vaso- Central Africa. It has been used for more than a constrictor a2 receptors. century in treatment of libido and erectile disorders. Although the mode of action of yohimbine As a nonstandardized whole or bark extract, yohim- suggests its effect may not be con®ned to erections bine was usually one ingredient in various com- while awake, clear evidence to support this assump- pounds and preparations. Once yohimbine was tion is still lacking. It was among the aims of the identi®ed as the main active constituent, puri®ed present study to include NPT measurements as a and standardized drug became available. Borelli2 possible additional indicator of the effect of yohim- considered it the prime aphrodisiac in terms of bine on erectile dysfunction. quality of medical documentation and clinical The clinical use of yohimbine has been docu- utility. mented by several open studies and anecdotal Yohimbine is a selective competitive a2-adrener- reports as well as by ®ve double-blind placebo- controlled trials6±10 conducted between 1987 and Correspondence: Dr H-J Vogt. 1990. Patients received daily oral doses of 16.2± Received 18 December 1996; accepted 28 December 1996 43.2mg for 4±10 weeks. While varied in design and Treatment of nonorganic erectile dysfunction H-J Vogt et al 156 outcome, all of these studies consistently indicated group (range, 28±71). The patients in the yohimbine a trend for improved erectile function with yohim- group were slightly shorter (176.5 vs 178.1cm) and bine [for a review see KoÈhler et al11; for meta- weighed slightly less (80.3 vs 82.3kg) than those of analytic considerations see Carey and Johnson12]. the placebo group. Other main characteristics of the Given the above considerations, and since pre- study population are summarized in Table 2. vious studies had focused on subjective aspects Blood counts, blood glucose, erythrocyte sedi- known to be exceedingly dependent on the patient's mentation rate, [serum] creatinine, liver enzymes, momentary suffering, the present clinical trial was urinalysis, and TPHA showed no deviations from designed to evaluate the ef®cacy of yohimbine in normal. The testosterone assay showed slightly terms of both objective and subjective variables. higher mean [serum] concentrations in the yohim- This study also aimed to establish the safety and bine group (14.88nmol/L) than in the placebo group tolerability pro®le of daily dosing with yohimbine (12.81nmol/L). Mean baseline luteinizing hormone over an eight week period of therapy. (LH) levels were essentially identical in the two groups. Nonetheless, all values including those determined for prolactin were within a normal Methods range. Psychiatric/psychologic evaluation using a depressive disorder evaluation questionnaire13 re- Subjects vealed evidence of mild paranoid or depressive ideation (yet without clinical signi®cance) in three yohimbine-treated patients (7.3%) and in one The target population consisted of men aged 18 and patient assigned to the placebo group (2.4%), and up suffering from erectile dysfunction. Criteria for evidence of a history of affective disorders (with no exclusion were: absence of sexual partner; severe current signi®cance) in two patients of the yohim- organic diseases such as severe heart disease, genital bine group. injury or serious intercurrent diseases; severe men- In both groups the baseline mean scores of a tal disease (for example depression); severe emo- Partnership Questionnaire14 and Insecurity in Socio- tional disturbances such as severe partner discord or sexual Situations Questionnaire15 showed no sig- overwhelming fear of sexual incompetence; severe ni®cant deviations from normal (Partnership sleep disorder; known allergy or hypersensitivity to Questionnaire: normal score, 64.9; yohimbine yohimbine; alcohol or drug abuse; participation in group, 59.6; placebo group, 59.0; Insecurity in another clinical trial within the last 30d; expected Sociosexual Situations Questionnaire: normal score, lack of cooperation and failure to give informed 21.64; patients with sexual disorders due to psycho- consent. genic factors: 46.53; yohimbine group, 29.3; placebo A total of 309 patients presented with erection group, 27.6). This con®rms that no patients with inadequacy at the Male Sexual Disorders Clinic of disinterest in sexual partner/marital discord or Dermatologische Klinik und Poliklinik, Technische clearly psychogenic sexual disorders had been UniversitaÈt MuÈ nchen, Munich, Germany, between included. 1993 and mid-1995. Out of the 309 patients 86 men were quali®ed for the trial, 223 patients with erectile impotence were screened and excluded for various Procedure reasons (see Table 1). All patients gave informed consent in writing prior to study entry. The study received institutional approval by the Ethics' Re- After the screening interview, patients were pro- view Board. vided with a diary to be completed and returned at Mean age of the study population was 53.9y for the time of the entry visit. After a thorough history the yohimbine group and 51.3y for the placebo and physical examination, subjects underwent a

Table 1 Reasons for exclusion from the trial

Primary reason for exclusion No. of patients

Poor Compliance (for example no interest, too much hassle) 34 (15.3%) Criterion for Exclusion (for example no sexual partner, already takes yohimbine, 35 (15.7%) participates in another trial, currently receives psychotherapy) Disinterest in sexual partner/marital discord 19 (8.5%) Erectile dysfunction due to mental disorder 42 (18.8%) Erectile dysfunction due to organic cause (for example anorchism, prolactinoma, prostate cancer) 38 (17.1%) Severe serious concomitant disease or need for unacceptable comedication 55 (24.6%) Total exclusions (223 of 309 patients) 223 (100.0%) Treatment of nonorganic erectile dysfunction H-J Vogt et al 157 Table 2 Baseline demographics/population characteristics

Yohimbine hydrochloride Placebo (n 41) (n 42)

Impotence Loss of erection 41 (100.0%) 42 (100.0%) Loss of desire 3 (7.3%) 2 (4.8%) Absence of orgasm 1 (2.4%) 1 (2.4%) Absence of emission Ð 4 (9.5%) Prior treatment for erectile inadequacy None 32 (78.0%) 27 (64.3%) Androgens 4 (9.8%) 4 (9.5%) Vitamin E Ð 1 (2.4%) Yohimbine 2 (4.9%) 3 (7.1%) Cavernous body auto-injection therapy Ð 1 (2.4%) Suppositories 1 (2.4%) Ð Unknown 2 (4.9%) 5 (11.9%) Patients with prior psychotherapy 4 (9.8%) 6 (14.3%) Patients with concomitant diseases 16 (39.0%) 14 (33.3%) Patients with concomitant medications 11 (26.8%) 11 (26.2%) Smokers/exsmokers 23 (56.1%) 23 (54.8%) Sexual partner Spouse 35 (85.4%) 30 (71.4%) Other partner 6 (14.6%) 12 (28.6%) Partner's age (y) Mean 49.6 46.7 s.d. 8.92 9.81

cavernous body injection test using 5±10mg of interfered with erectile function or speci®c study prostaglandine (PGE-1) followed by penile artery procedures (for example sleep quality). Unaccepta- Doppler ultrasonography. Hormonal and serious ble comedications included aphrodisiacs, male sex organic causes were ruled out by a variety of hormones, and other drugs affecting sexual func- laboratory tests. A complete psychiatric/psychologic tion; beta-blockers and antihypertensive drugs; evaluation, including an interview/discussion of the sleeping drugs, tranquillizers, and any other psy- information provided in the completed diary, choactive drugs. All comedication was to be re- Partnership Questionnaire,14 and Insecurity in So- corded in the Case Report Form (CRF), including ciosexual Situations Questionnaire15 was also per- nature, dosage, duration, and indication. Existing formed. acceptable comedications were not to be changed Sleep was polysomnographically recorded during during the trial as far as possible. two consecutive nights at baseline and after eight The sample size calculation for this trial was weeks of treatment. In addition to routine sleep based on assumptions set forth in the study proto- parameters, spontaneous nocturnal erections were col, which were derived from reference to previous monitored using a RigiScan device (Dacomed Corp., clinical trials of yohimbine.6±10 These studies used Minneapolis, MN, USA). This device allows con- 5.4mg of yohimbine (instead of 5.0mg) per tablet, tinuous recording of both penile tumescence and but the current state of medical knowledge and the rigidity without interfering with sleep recordings. nature of the criteria used for ef®cacy evaluation in At the four-week follow-up visit the following this trial suggest this minor difference in tablet variables were recorded: compliance (by counting strength was insigni®cant. The following assump- returned tablets), comedication changes, adverse tions were used in calculating sample size: two events, overall and physical well-being, sexual independent groups, testing for statistically signi®- desire, and best erection. Patients then received cant between-group difference, same number of another four weeks' supply of their study medica- patients in both groups, a 5%, b 20%, one-tailed tion, along with a new diary. test, and difference in response between yohimbine After eight weeks' treatment, patients underwent and placebo after eight weeks: yohimbine response follow-up sleep laboratory studies including mea- rate placebo response rate 30% (in absolute surements of nocturnal penile tumescence/rigidity terms). Based on the above assumptions, the during two consecutive nights. Other ®nal visit calculated sample size ranged from 64±80 evaluable activities included a psychiatric/psychologic patients as a function of assumed placebo response follow-up evaluation, as well as recording of rate (10±50%). A 30% difference (in absolute terms) co-medication changes and any adverse reports. in response rate in favor of yohimbine was assumed Treatment of concomitant diseases that were not to be a clinically signi®cant result. criteria for exclusion was acceptable unless it The basis for response de®nition was a combina- Treatment of nonorganic erectile dysfunction H-J Vogt et al 158 tion of objective and subjective criteria of outcome. out before completing the full eight week treatment The objective end-point was based solely on the period and/or all of the ®nal visit activities. The rigidity measured at the base of the penis, rating the primary reasons for termination were non-compli- quality of erection on a scale from 0±3 (0 0% ance and adverse events such as sleep disturbances. rigidity, 1 1±40% rigidity, 2 41±70% rigidity, The mean baseline values calculated for `sexual 3 71±100% rigidity). Erections of nights 1 and 2 desire' were slightly better in the yohimbine group (baseline recordings) and those of nights 3 and 4 than in the placebo group (2.61 vs 2.43), whereas (end-of-study recordings) were pooled for erection baseline `best erection during sexual contact', quality analysis, using the best rigidity level main- `frequency of sexual contacts', and `sexual satisfac- tained for at least 15 minutes for the baseline versus tion' were essentially identical in the two groups. end-of-study night comparison (objective end- The reported current mean monthly frequency of point). intercourse/sexual contacts was 1.8 (yohimbine) or The subjective end-point re¯ected changes in 2.5 (placebo). All variables indicated considerable four variables, namely `sexual desire', `sexual satis- variation in both groups. faction', `best erection during sexual contact/intercourse', and `frequency of sexual contacts'. These variables were determined from the information Response provided in the diaries and obtained during the patient interviews. To calculate subjective response After eight weeks of treatment 24 out of 41 patients the values of the 14d diary maintained by patients (59%) were responders according to the subjective prior to initial entry into the study and the eight- response de®nition in the yohimbine group com- weeks follow up diaries were used as well as the pared to 16 out of 41 (39%) in the placebo group. interview values. In order to calibrate response rate This difference is associated with a P-value of for these values in an equal manner, qualitative 0.1215. diary values were averaged (sexual desire, sexual The NPT recordings re¯ected a treatment re- satisfaction, best erection during sexual contact) and sponse according to the objective response de®ni- added to the proper interview value, using the tion in 13 out of 41 patients (32%) receiving arithmetic mean for analysis. The quantitative diary yohimbine; in contrast, only 6 out of 42 patients values (frequency of sexual contact) were summar- (14%) receiving placebo exhibited a response ac- ized and added to the proper interview value, again cording to the criteria described above. This differ- using the arithmetic mean for analysis. ence, however, did not reach statistical signi®cance Response was de®ned as a best rigidity improve- either (P-value of 0.1024). ment of at least 1 point in the two end-of-study Using the primary target, namely the overall nights versus the two baseline nights and/or an response de®nition, 29 out of 41 patients of the improvement in at least 2 of the 4 subjective criteria yohimbine group (70.7%) versus 19 out of 42 of outcome by at least 0.5 points (sexual desire, patients of the placebo group (45.2%) were respon- sexual satisfaction, best erection during sexual ders at the end of therapy (Table 3). This difference contact) of a 5-point scale or at least one more in response rates was statistically signi®cant in sexual contact over a two-week period. Fisher's exact test at a 5%, one-tailed (P 0.0163). The between-group difference in response rate The 95% con®dence interval of the likelihood of (primary end-point) was analyzed and interpreted response after eight weeks' treatment was 54.5± con®rmatively by Fisher's exact test at the a 5% 83.8% in the yohimbine group vs 29.9±61.3% in the level, one-tailed. All other ef®cacy and safety placebo group (Figure 1). The statistical power was variables were presented descriptively and inter- 70% at the a 5% level and 81% at the a 10% preted exploratively. level. There was a weak relationship between the outcome of NPT recordings and subjective response Results criteria, with only eight patients receiving yohimbine (20%) classi®ed as responders according to both subjective and objective criteria. In the placebo Of the 86 randomized patients, 85 were evaluated group, only three patients (7%) ful®lled both for safety/tolerability (n 43 in the yohimbine criteria. There was no signi®cant difference between hydrochloride group, n 42 in the placebo group) yohimbine and placebo regarding this relationship. and 83 for ef®cacy (n 41 in the yohimbine group, n 42 in the placebo group). Eighty-one patients (n 41 in the yohimbine group, n 40 in the placebo group) completed eight Safety and tolerance weeks of treatment as scheduled for the trial. Two patients in the yohimbine group and three patients Adverse events (AEs) were reported in 13 patients of in the placebo group were withdrawn or dropped the yohimbine group (30.2%) vs four in the placebo Treatment of nonorganic erectile dysfunction H-J Vogt et al

Table 3 Response rates after eight weeks of treatment 159

Type of Overall response Subjective response Objective response response group Yohimbine Yohimbine Yohimbine hydrochloride Placebo* Hydrochloride Placebo Hydrochloride Placeboa

Response 29 (71%) 19 (45%) 24 (59%) 16 (39%) 13 (32%) 6 (14%) Non-response 12 (29%) 23 (55%) 17 (41%) 25 (61%) 28 (68%) 36 (86%) P-value 0.0163* 0.1215** 0.1024**

* Exact P-values from Fisher's exact test (one-tailed). ** P-values were calculated using Fisher's exact test (two-tailed) and are double the respective one-tailed P-value. a No subjective response evaluation was available for one patient of the placebo group.

stance. Besides the effect on erectile functions, Yohimbine often shows additional improvement in sexual appetite/desire. Apart from this, the improvement in erections may exert a positive in¯uence on the patient's emotional condition and thus support overall success of the treatment. Although subjective improvement in sexual experience tends to be a prime concern for men with erectile impotence, and perhaps for their sexual partners as well6 medical scientists also must concern themselves with measuring and quantifying this subjective experience for objective veri®cation. This is why in our study subjective as well as objective criteria were to be evaluated. This double-blind, placebo-controlled clinical trial of yohimbine hydrochloride was carried out Figure 1 95% con®dence intervals (95% CI) for overall response in patients with erectile dysfunctions with or with- rates and respective mean values after eight week treatment with out reduced sexual desire/libido. Strict exclusion Yohimbine versus Placebo. The upper and lower bars indicate the upper and lower limit of the 95% CI, respectively. The middle criteria were applied to ensure valid conclusions, (mean) bar indicates the study population mean. with the result that only 86 of the 309 screened patients could be included in this study. The mean age of the study population (53y) is that of the group (9.5%). This difference was statistically typical patient with erectile impotence seen by the signi®cant in Fisher's exact test. Two of the 43 GPs, andrologists and urologists. patients of the yohimbine group and one of the 42 Most clinical studies used the patients' (subjec- patients of the placebo group stopped treatment tive) response to rate the therapeutic ef®cacy, prematurely for adverse experiences. relaying, primarily on the patients, as well as on There were no serious AEs in the yohimbine their partners to a lesser extent (Morales et al6, Reid group. Two AEs experienced by patients of the et al7). Patients were asked to rate their erectile placebo group were considered serious (per study capability (Morales et al6,17, Sonda et al9, Susset et protocol de®nition): one patient required hospitali- al10). Additional ratings referred to orgasmic re- zation for intercurrent depression and was with- sponse (Riley et al8) or sexual desire (mostly in older drawn, but the investigator considered the reports). In several previous studies three rating relationship of this AE with the study medication categories were used: ùn-changed', `partial im- questionable. Another placebo-treated patient un- provement' or `no further erection problems' (Mor- derwent an appendectomy during the trial, but this ales et al6, Reid et al7). This study used three event had no signi®cant impact on the course of the qualitative parameters (sexual desire, sexual satis- trial and a causal relationship with the study faction, and best erection during sexual contact/ medication was ruled out. intercourse) and one quantitative variable (frequency of sexual contacts) for subjective response evaluation. To obtain more complete information Discussion the patients were interviewed using a semi-struc- tured schedule, and additionally were asked to keep a speci®c daily diary for two weeks before the start The report of Morales et al16 on the successful medication (baseline) until the end of the study. treatment of impotence in diabetes with yohimbine Data from these two sources were given equal helped lead to intensive investigation of this sub- statistical weight in the ®nal analyses. Treatment of nonorganic erectile dysfunction H-J Vogt et al 160 We used NPT recording as an additional objective improvement in patients with extremely low base- response criterion to evaluate the ef®cacy of yohim- line activity. This led us to choose a response bine. We considered it essential to perform the NPT criterion which takes account of the `ceiling effects' measurements in the sleep laboratory concurrently and other possible kinds of distortions. Based on with polysomnographic recordings during sleep at studies using a similar design,22,23 we used a four night in order to allow differentiation of sponta- class ordinal scale for rigidity classi®cation, re- neous erection while asleep and erections while sponse being de®ned as an improvement by at least lying awake. NPT being rather closely linked to REM one rigidity class. Additionally, we required that sleep, NPT frequency and duration can be properly rigidity of a given class should be maintained for at assessed only when the frequency and duration of least 15 minutes. We are aware that by doing so, REM phases are known. REM sleep deprivation may some of the potential statistical power of the be caused by many different factors including analyses is lost, but elimination of cluster-effects unfamiliar surroundings, sleep fragmentation due and bias were the main arguments leading to our to external or internal causes, drugs, and alcohol. decision. Moreover, certain sleep disorders such as severe Yohimbine tended to be better than placebo in sleep apnea are known to compromise erectile terms of subjective response criteria and the objec- function. tive response criteria looked at these two categories The measurement of nocturnal penile tumescence separately (Table 3), but it is not statistically (NPT) is by no means a standard procedure in this signi®cant. It was not possible to sum up both kind of study. While the merits of this method in the groups, because some patients from the subjective differential diagnosis of erectile impotence are response group responded in the objective response noncontroversial,18 its utility and reliability for group as well. documentation of ef®cacy of drugs used in treatment The combined set of objective responders and of this sexual disorder are not. Jovanovic19 used NPT subjective responders (namely the overall response measurements in a clinical trial of a yohimbine set of patients achieving objective and/or subjective product and observed pronounced effects that response), revealed a statistically signi®cant super- correlated closely with subjective improvement. iority of yohimbine to placebo. Using all due Morales et al6, on the other hand, found signi®cant caution, this result can be interpreted as indicating discrepancies between NPT measurements and that the threshold of a subjective or objective subjective perception of clinical outcome among response shows great between-patient variability. patients and their partners. As subjective response In other words, some patients experienced subjec- appeared more valid from a clinical standpoint, the tive improvement while objective response re- authors decided not to use the NPT recordings as an mained below the `limit of detection', whereas outcome measurement.20 Other workers performed others achieved objective response while subjective polysomnographic measurements including NPT response was not detectable with the rating scales but only used these to establish the differential used for sexual function assessment in our trial. diagnosis and not as part of the response criterion.21 Unlike Morales6 and Condra20 we found no NPT measurements result a variety of parameters substantial discrepancy between subjective percep- including extent and duration of penile tumescence, tion and NPT recordings. However, a couple of extent and duration of penile rigidity, number of important differences between these studies deserve erections, and temporal stability of an erection. special mention. Morales6 studied only patients There is no consensus in the literature on which of with organic erectile dysfunction, while these these variables are useful for the assessment of patients were excluded from our trial. Moreover, spontaneous nocturnal erectile functions (for exam- changes in penile circumference (tumescence) may ple in the differential diagnosis of erectile dysfunc- have less clinical signi®cance than do changes in tion) or which most reliably mirror drug effects. We penile rigidity. Consequently, we used an instru- decided to focus on penile rigidity measurements at ment that allows continuous rigidity measurements the penis base, and to de®ne òbjective' response as (RigiScan), and based our assessment on rigidity an improvement in this parameter. However, when only. Another variable used by Morales6 was the `response' is de®ned using absolute rigidity mea- number of nocturnal erections. As this parameter surements and percentages of improvement from correlates with the number of REM phases, all baseline, a serious methodological problem arises: events producing REM sleep deprivation are also in the present sample, a large percentage of patients liable to reduce the number of erections. Based on had such favorable NPT measures at baseline as to this consideration, we believe this variable to be an leave very little, if any, room for potential improve- unreliable indicator of erectile function. ment attributable to treatment. In these patients, The relationship between `subjective' and òbjec- minor improvements are not adequately mirrored in tive' response criteria merits particular attention. It the perceptual changes. Conversely, even minute is still a matter of debate whether, and to what changes are re¯ected as disproportionally great extent, there is a relationship between nocturnal Treatment of nonorganic erectile dysfunction H-J Vogt et al 161 spontaneous erections/nocturnal penile tumescence 5 Weitzell R, Tanaka T, Starke K. Pre- and postsynaptic effects of and sexual performance in erotic situations. While yohimbine stereo-isomers on noradrenic transmission in the pulmonary artery of the rabbit. Naunyn Schmiedebergs Arch certainly not identical, the underlying neurophysio- Pharmacol 1979; 308: 127±136. logic mechanisms are likely to overlap at least in 6 Morales A et al. Is yohimbine effective in the treatment of some respects. The following conclusions can be organic impotence? Results of a controlled trial. J Urol 1987; drawn from our results regarding this relationship: 137: 1168±1172. The intersection of the set of objective responders 7 Reid K et al. Double-blind trial of yohimbine in treatment of psychogenic impotence. Lancet 1987; 2: 421±423. and the set of subjective responders is a small one 8 Riley AJ, Goodman RE, Kellett JM, Orr R. Double blind trial of but shows a marked difference between the two yohimbine hydrochloride in the treatment of erection inade- treatment groups, suggesting a rather loose link quacy. J Sex Marital Ther 1989; 15: 17±26. between the two criteria. Neither criterion is 9 Sonda LP, Chancellor MB, Mazo R. Treatment of erectile prerequisite or suf®cient for the other to occur. impotence with yohimbine; prospective double blind trial and review of experience in 215 patients. J Sex Marital Ther 1990; The higher incidence of adverse experiences with 16: 15±21. yohimbine than with placebo is not surprising. As 10 Susset JG et al. Effect of yohimbine hydrochloride on erectile yohimbine acts on both the central and autonomic impotence; a double-blind study. J Urol 1989; 141: 1360±1363. nervous systems, side effects on both are likely. 11 KoÈhler LD, Borelli S, Vogt HJ. Yohimbin-HCl in der Behand- Placebo produced only psychosomatic reactions lung von ErektionsstoÈrungen. Sexuologie 1995; 3 (2): 209±217. 12 Carey MP, Johnson BT. Effectiveness of Yohimbine in the while yohimbine also caused reactions to the drug Treatment of Erectile Disorders: Four Meta-Analytic Integra- itself. However, as these adverse reactions were tions. Arch Sex Behav 1996; 25: 341±360. typically mild, they were either acceptable or made 13 Zerssen D von. Paranoid-DepressivitaÈts-Skala. DepressivitaÈts- acceptable by dose reduction. Skala. PD-S Manual Beltz-Verlag, Weinheim, 1976. 14 Hahlweg K. Partnerschaftliche Interaktion; Partnerschaftsfra- gebogen. G. RoÈttgerVerlag, MuÈ nchen, 1986. 15 Fahrner EM. Fragebogen zur Unsicherheit in soziosexuellen Acknowledgements Situationen (FUSS). In: Fahrner EM (ed). Psychologische Behandlung von SexualstoÈrrungen bei maÈnnlichen Alkohol- abhaÈngigen. G. RoÈttger-Verlag; MuÈ nchen, 1985. This study was supported by Glenwood GmbH, 16 Morales A, Surridge DHC, Marshall PG. Yohimbine for treatment of impotence in diabetes. N Engl J Med 1981; 305: Starnberg, Germany, Yohimbine-HCl (Yocon-Glen- 1221. 1 wood ) was provided by Glenwood GmbH, Starn- 17 Morales A, Surridge DHC, Marshall PG, Fenemore J. Non- berg, Germany. hormonal Pharmacological Treatment of organic impotence . J Urol 1982; 128: 45±47. 18 Schiavi RC. Nocturnal penile tumescence in the evaluation of erectile disorders: a critical review. J Sex Marital Ther 1988; References 14: 83±97. 19 Jovanovic UJ. Hormonfrei gegen Impotenz. Sexualmedizin 1980; 9: 453±457. 1 Spector JP, Carey M. Incidence and prevalence of the sexual 20 Condra M et al. The unreliability of nocturnal penile dysfunctions: a critical review of the empirical literature. Arch tumescence recording as an outcome measurement in the Sex Behav 1990; 19: 389±408. treatment of organic impotence. J Urol 1986; 135: 280±283. 2 Borelli S. Die Aphrodisiaka. In: Jadassohn J (Hrsg). Handbuch 21 Montorsi F et al. Effect of yohimbine-trazodone on psycho- der Haut- und Geschlechtskrankheiten. ErgaÈnzungswerk Band genic impotence: a randomized, double-blind, placebo-con- VI 3. 1960, 737±766. trolled study. Urology 1994; 44: 732±736. 3 Borowski E, Starke K, Ehrl H, Endo T. A comparison of pre- 22 Mann K et al. Effects of yohimbine on sexual experience and and postsynaptic effects of the alpha-adrenolytic drugs in the nocturnal penile tumescence and rigidity in erectile dysfunc- pulmonary artery of the rabbit. Neuroscience 1977; 2: 285±296. tion. Arch Sex Behav 1996; 25: 1±16. 4 Starke K, Borowski E, Endo T. Preferential blockade of 23 Warner P, Bancroft J. Assessment of erectile dysfunction: presynaptic alpha-adrenoceptor by yohimbine. Eur J Pharma- consideration of the research issues. Poster presented at the col 1975; 34: 385±388. 20th Annual Conference of the International Academy of Sex