Penile duplex sonography in the diagnosis of

Transcription

1 Penile duplex sonography in the diagnosis of venogenic impotence Hisham S. Bassiouny, MD, and Laurence A. Levine, MD, Chicago, Ill. This study tested the hypothesis that measurements of cavernous arterial diastolic velocity and resistance index could provide a quantitative but noninvasive measure of penile corporal venous leakage. Seventy-four men were studied with duplex ultrasonography after intracavernosal injection of 60 mg of papaverine. Fourteen men had normal erection and served as controls. Sixty men had a I-year history of transient fading or incomplete erections. In all subjects the peak systolic velocity and end-diastolic velocity were measured, and the resistance index was calculated (peak systolic velocity - end-diastolic velocity/peak systolic velocity). Men with normal erections had peak systolic velocities greater than 35 cm/sec and end-diastolic velocities less than 4.5 cm/sec (group 1). Patients with incomplete erections (group II) could be classified into three subgroups. Twentythree patients with end-diastolic velocities greater than 4.5 cm/sec and normal peak systolic velocities greater than 35 cm/sec were suspected to have corporal venous leakage (group A). Eighteen patients had normal end-diastolic velocities less than 4.5 cm/sec. Twelve of this group had peak systolic velocities less than 35 cm/sec, and six had peak systolic velocities ranging from 37 to 44 cm/sec. These patients were suspected of having arterial insufficiency (group B). Fifteen patients with end-diastolic velocities greater than 4.5 cm and peak systolic velocities less than 35 cm were suspected of having both venous leakage and arterial insufficiency (group C). Twenty-one patients with abnormal diastolic flow underwent infusion pharmacocavernosometry to determine the saline maintenance infusion rate necessary to maintain an intracavernosal pressure ofg0 to 100 mm Hg or a flail erectile response. In 19 patients venous leakage was confirmed by cavernosography. A significant inverse correlation was found between the resistance index and the infusion maintenance rate (r = 0.66,p < 0.001). These findings suggest that penile duplex scanning is useful in differentiating between venogenic and arteriogenic impotence. The presence of diastolic flow correlated with accepted determinations of venous leakage from the corpora cavernosa. (J VASC SURG 1991;13:75-83.) It is estimated that approximately 10 million American men are impotent) The mechanism of erection is complex and involves intact neurochemical stimulation and an adequate end organ vasculogenic response. 2-s In men with normal erections, neurochemical stimulation produces arteriolar and cavernosal smooth muscle relaxation. This results in increased penile arterial inflow via the paired deep cavemosal arteries into the corpora cavernosa. The corporal trabecular walls expand compressing the subtunical venus plexus against the tunica albuginea. The reduction in venous outflow with subsequent trapping of blood in the corpora results in tumes- From the Sections of Vascular Surgery and Urology, the Depam'aent of Surgery at The University of Chicago Medical Center. Presented at the Forty-fourth Annual Meeting of the Society for Vascular Surgery, Los Angeles, Calif., June 4-6, Reprint requests: Hisham S. Bassiouny, MD, Section of Vascular Surgery, University of Chicago Medical Center, 5841 S. Maryland Ave., Chicago, IL /6/25582 cence and rigidity. 46 Disorders in blood flow to and from the corpora cavernosa are thought to be the most common organic causes of impotence, s# Atherosclerotic and traumatic arterial occlusive disease can diminish arterial blood flow to the lacunar spaces and thus reduce penile rigidity during the erectile response, s'8" Vasculogenic impotence may also result from a defective venoocclusive mechanism such that excessive outflow occurs despite adequate arterial inflow. 6a Duplex ultrasonography has been used previously to assess cavernous arterial peak systolic velocities (PSVs) and luminal diameter changes after intracavernosal injection of papaverine, n'~2 These studies have been performed after injection of papaverine as it mimics the endogenous physiologic mechanisms that produce smooth muscle relaxation, inhibit adrenergic tone, and induce erection. 13 In this study we investigated the value of duplex ultrasonography in diagnosing and quantitating corporal venous leakage using arterial flow velocity parameters. To validate our results we compared our non- 75

2 76 Bassiouny and Levine Journal of VASCULAR SURGERY invasive findings to infusion pharmacocavernosometry and cavernosography. MATERIAL AND METHODS Between April 1988 and March 1990, 270 adult men were evaluated at The University of Chicago Male Sexual Dysfunction Clinic with at least a 1-year history of erectile insufficiency. All patients had a detailed history and thorough physical examination, determination of serum testosterone, glucose, total cholesterol, and serum renal and liver profiles. Those patients suspected of psychogenic impotence underwent psychologic evaluation including formal consultation and psychometric testing. Seventy-four patients underwent penile duplex ultrasonography after intracorporal papaverine injection to study penile arterial velocity flow pattems obtained at the time of maximum erectile response. Sixty were suspected of having vasculogenic impotence. None of these patients had abnormal testosterone levels or neurologic disease. Their mean age was 52 years (range, 26 to 71 years). Patients with diastolic flow on real time velocity spectral analysis were suspected of having corporal venous leakage and were further studied with infusion pharmacocavernosometry and cavernosography to confirm the presence and determine the severity of the venous leakage. In addition, 14 men who had normal erections were studied as a control group; six were volunteers, two had Peyronie's disease, and six had resolving psychogenic impotence with no underlying atherosclerosis risk factors. The mean age of the control group was 52 years (range, 22 to 69 years). Penile duplex uttrasonography and infusion pharmacocavernosometry and cavernosography were performed after informed consent. Duplex ultrasonography of the penile arteries Penile arterial blood flow was evaluated with the patient in the supine position in a dimly lit room. Two technicians were necessary to perform the study. One of us (L.A.L.) was present for all studies. The ultrasonic duplex scanner (Acuson 128; Acuson Inc., Mountain View, Calif.) used to evaluate the deep cavernosal penile arteries consisted of a 7.5 MHz B- mode imager and a 5 MHz pulsed Doppler unit with real-time spectral analysis. Color flow imaging was used to locate the deep cavernosal arteries. The linear array probe was positioned on the dorsal aspect of the flaccid penis, and the corpora cavemosa, deep cavemosal arteries, and tunica albuginea were initially surveyed for any anatomic or structural anomalies. Sixty milligrams of papaverine HC1 (30 mg/ml) was subsequently injected into the corpora cavernosa by means of a 27-gauge needle. Before the intracorporal injection, a rubber band was placed around the base of the penis to reduce leakage ofpapaverine into the systemic circulation and was removed after 3 minutes. After 5 to 7 minutes the deep cavemosal arteries were visualized at the base of the penis by use of color flow imaging, and the sample volume of the pulsed Doppler flow detector was positioned to obtain centerline blood flows. Angle of insonation was minimized to decrease error associated with blood velocity calculations. Patients with incomplete erections were restudied after a period of privacy and self-stimulation using visual erotic stimuli. All controls and patients suspected of vasculogenic impotence underwent the study. Hemodynamic measurements Peak systolic velocities and end-diastolic velocities (EDVs) from the right and left deep cavernosal arteries were measured by use of the electronic calipers of the pulsed Doppler flow detector. The resistance index (ill) was calculated according to Planilo et al.l~ (PSV - EDV/PSV). Flow measurements were recorded before and after self-stimulation. Since penile arterial blood flow varies during the dynamic phase of developing an erection, s only those flow measurements determined at the time of maximal erectile response were considered. Penile brachial pressure index To assess the value of the penile brachial pressure index (PBPI) it was measured during the last 4 months of the study period. Measurements were obtained before duplex scanning by means of a continuous-wave 10 MHz Doppler (Parks) with standard technique, is Infusion pharmacocavernosometry and cavernosography Patients suspected of corporal venous leakage underwent infusion pharmacocavernosometry and cavernosography (IPCC). The technique of IPCC has been previously described. 7,16,17 Penile block is accomplished with 10 ml of 1% xylocaine to reduce patient discomfort during needle placement and when intracorporal pressures become elevated. Two 21-gauge butterfly needles are placed into the corpora cavernosa (one in each corpora) on the lateral surface of the penis. A needle is connected to a pressuretransducer with an online digital readout module for continuous monitoring of intracorporal pressure. The second needle is connected to a roller

3 Volume 13 Number 1 January 1991 Venogenic impotence diagnosed by duplex sonography 77 infusion pump with an adjustable flow rate (Cavropump; Life-Tech, Inc., Houston, Texas). After the device is calibrated, baseline intracorporal pressure is recorded. An intracorporal injection of a mixture of 45 mg of papaverine HCI and 1.5 mg of phentolamine mesylate is then administered via the infusion line to produce cavernous sinusoidal relaxation similar to that achieved during normal erections. Once intracorporal pressures have stabilized, heparinized saline (1000 units/1l of 0.9 N) infusion is initiated at 20 ml/min and incrementally increased by 10 to 20 ml/min until a full rigid erection is attained or intracorporal pressure rises to 90 to 100 mm Hg. The device is then programmed to deliver the volume (ml/min) necessary to maintain the erection at 100 mm Hg. This maintenance rate was selected as the quantitative measurement of penile venous leakage. After cavernosometry, cavernosography is performed. This is accomplished by infusion of 50 to 100 rnl of saline mixed (1 : 1) with diatrizoate or 30% iothalomate meglumine into the corpora until a pressure of 90 to 100 mm Hg, or full erection is reached. If no erection can be attained because of extreme venous leakage, no more than 100 ml of the contrast solution is infused. Three radiographic films are obtained; anteroposterior and two obliques with a separate infusion of contrast in each position. Leakage can be demonstrated from any one or more of the following sites: the superficial and deep dorsal penile veins, the cavernous veins, ectopic penile veins, and via spongio cavernosal shunts into the glans penis or corpus spongiosum. Any one or combination of the above sites may be involved in penile venous leakage.17 Statistical analysis Values were expressed as mean + standard deviation of the mean. Comparison between the various flow' parameters (PSV, EDV, 1LI) of the different groups was performed by use of a paired Student t test. The relationship between the RI and the IPCC maintenance flow rate was assessed by means of a linear regression analysis. Differences were considered to be significant ifp < RESULTS Complete evaluation with duplex ultrasonography after papaverine injection was accomplished in all 14 individuals with normal erections (group I) and in 56 of 60 patients with an incomplete erectile response to intracavernosal injection of papaverine (group II). None of these patients who were offered Fig. 1. Typical deep cavernosal arterial velocity waveform in a normal control with complete erection. Peak systolic velocity is 118 cm/sec and diastolic flow is absent. either procedure refused to undergo the examination. No significant complications occurred as a result of intracorporal cannulation, papaverine injection, or cavernosography. Five patients had small areas of ecchymosis at the needle insertion site (3) or transient discomfort (2). Six subjects in the normal group required reversal of the pharmacologically induced erection. Reversal was indicated when erection was full at least 1 hour after the injection. A diluted solution of 0.1 ml 1:1000 aqueous epinephrine mixed with 0.9 ml sterile 0.9 NaCI was given intracorporally in 0.1 to 0.2 ml aliquots and rapidly caused detumescence. After each aliquot injection the penis was compressed for 5 to 10 minutes before another aliquot was given. In 38 patients (63%) the maximal erectile response was attained after self-stimulation. No more than 0.7 ml of this solution was required to achieve detumescence. Classification of patients: criteria The penile arterial blood flow patterns determined at the time of the maximal erectile response were used to classify patients with an incomplete erection into three subgroups. Corporal venous leakage was suspected if the EDV was greater than that of normal controls (group A). Arterial insufficiency was suspected if the EDV was normal and the PSV was lower than that of normal controls (group B). Both venous and arterial insufficiency were suspected if abnormal EDV and PSV were found (group C). Twenty-four patients with suspected corporal venous leakage by duplex ultrasonography underwent infusion pharmacocavernosometry and cavernosography.

4 78 Bassiouny and Levine Journa! of VASCULAR SURGERY Fig. 2. Pulsed Doppler spectral analysis of deep cavernosal arterial flow at the time of maximal erectile response in a patient suspected of having corporal venous leakage. Peak systolic velocity is greater than 35 cm and continuous diastolic flow with an EDV of 24 cm/sec. Fig. 3. Pulsed Doppler spectral analysis of deep cavernosal arterial flow at the time of the maximal erectile response in a patient suspected of having penile arterial insufficiency. Peak systolic velocity is less than 35 cm/sec and diastolic flow is absent. Duplex ultrasonography Group I: normal controls (n -- 14). All normal controls had PSV >35 cm and EDV <4.5 cm/sec. The mean PSV was _ cm/sec (range, 35 to 120). The mean EDV was 0.62 ± 0.94 cm/sec (range, 0.00 to 3.00). The mean RI was (range, 0.95 to 1.00). Fig. 1 represents a typical arterial velocity waveform in individuals with normal erections. Group II: incomplete erections (n = 56). Patients were subclassified into three subgroups; those suspected of having corporal venous leakage (group A), arterial insufficiency (group B), and both (group C). (i) Patients with suspected corporal venous leakage (group A). Twenty-three patients had EDVs greater than 4.5 cm/sec and PSV greater than 35 cm. The mean EDV was cm/sec, (range, 7.00 to 25.00). The mean PSV was ± cm/sec (range, to 95.00). The mean RI was (range, 0.45 to 0.88). Compared to normal controls, the mean EDV was significantly higher (15.83 _ 5.09 vs 0.62 ± 0.94 cm/sec, p < 0.001), and the mean RI was significantly lower (0.71 vs 0.99, p < 0.001). The PSV was not, however, different (57.13 ± vs ± cm/sec, p > 0.10). Fig. 2 represents a typical arterial

5 Volume 13 Number 1 Januaq, 1991 Venogenic impotence diagnosed by duplex sonography 79 Fig. 4. Pulsed Doppler spectral analysis of deep cavernosal arterial flow in a patient suspected of having both corporal venous leakage and penile arterial insufficiency. Peak systolic velocity is less than 35 cm and EDV is 5 cm/sec. velocity waveform in a patient with suspected corporal venous leakage. (2) Patients with suspected arterial insufficiency (group B). Eighteen patients had EDVs of <4.5 cm/sec. Twelve patients had PSVs <35 cm/sec, and six patients had PSVs >35 cm (37 to 44 cm/sec). The mean EDV was 1.24 _ cm/sec (range, 0.00 to 4.4). The mean PSV was cm/sec (range, to 44.00). The mean RI was (range, 0.77 to 1.00). Compared to normal controls; neither mean EDV nor the RI were different ( vs cm/sec and vs , p>0.03 and p> 0.05, respectively). The mean PSV, however, was significantly lower ( vs cm/sec, p < 0.001). Fig. 3 represents a typical arterial velocity waveform in a patient with suspected arterial insufficiency. (3) Patients with suspected venous leakage and arterial insufficiency (group C). Fifteen patients had EDVs greater than 4.5 cm/sec and PSVs <35 cm/sec. The mean EDV was (range, 4.6 to 18.00), the mean PSV was (range, 15 to 34), and the mean RI was _ 0.13 (0.40 to 0.88). Compared to the control group, the mean EDV was significantly higher ( vs , p < 0.001), the mean RI and PSV were significantly lower ( vs and vs , p < 0.001, respectively). Fig. 4 represents a typical waveform of a patient with both venous leakage and arterial insufficiency. Table I. Mean values of the EDV, PSV, and RI of patients in group I and group II Group I-normal controls (n = 14) Group II-incomplete erections Group A (n = 22) Group B (n = 18) Group C (n = 15) Values expressed as mean + SD. *Differences significant at p < PSV EDV (cm/sec) (cm/ sec) RI 65-27* * 0.99* 57 _ _+ 5* 0.71" " * * 0.63* The mean values of the EDV, PSV, and RI of normal controls (group I) and patients with incomplete erections (group II) are represented in Table I. Penile brachial pressure index The PBPI was measured in 23 subjects. Of these, six were controls with normal erections, and seven were suspected of penile arterial insufficiency by duplex ultrasonography. Although the mean PSV was significantly lower in the latter group compared to controls ( vs , p < 0.02), the PBPI was not different (0.79 vs 0.83, p > 0.5, respectively). Infusion pharrnacocavernosometry and cavernosography Twenty-four patients underwent both complete IPCC and duplex ultrasonography. The maintenance

6 80 Bassiouny and Levine lourn~of VASCULAR SURGERY MAINTENANCE RATE VS 1/RESISTANCE INDEX (RI).~ R =.66 P <.001 s: do do MAINTENANCE RATE Fig. 5. Oblique cavernosogram in patient with maintenance rate of 130 ml/min. Deep dorsal, cavernosal, and spongiosal venous drainage noted into periprostatic plexus and iliac veins. rate varied from 4 to 170 ml/minute. Twenty-one patients had EDVs over 4.5 cm/sec and were suspected of having corporal venous leakage. In 19 of these patients, cavemosography confirmed the presence of corporal venous leakage (Fig. 5). In the remaining three patients with EDVs less than 4.5 cm/sec cavemosography did not demonstrate venous leakage. The positive and negative predictive values of duplex ultrasonography for corporal venous leakage were 90% and 100%, respectively. The maintenance rates of patients with confirmed venous leakage by IPCC ranged from 20 to 170 ml/min. The maintenance rates of patients with negative cavernosograms ranged from 4 to 17 cm/min. We found a strong correlation between the inverse of RI and the maintenance flow rate (r = 0.66, p < 0.001) (Fig. 6). DISCUSSION Erectile dysfunction may occur as a result of neurologic, endocrine, psychogenic, structural, or vasculogenic cause. 2,~sa9 Evaluation of impotence should include a detailed history and physical examination, serum testosterone levels, chemical screening, and psychologic evaluation when indicated. 2 In those patients suspected of vasculogenic impotence, invasive methods such as infusion cavemosography to detect corporal venous leakage, z,16,17 and phalloarteriography have been performed to identify pelvic or penile arterial disease, s,9 More re- Fig. 6. Relationship of the inverse of the resistance index and the IPCC infusion maintenance rate. cently, noninvasive testing with penile duplex ultrasonography has been used to assess penile arterial inflow by measuring arterial diameter changes and PSVs after intracorporal papaverine injection. 11,12,21,22 Those studies have demonstrated that PSVs less than 25 cm/sec, and attenuated diameter changes are indicative of penile arterial insufficiency. The diagnosis of venogenic impotence, however, was made by exclusion. Patients with normal PSVs and an incomplete erectile response to intracorporal papaverine were considered to have corporal venous leakage. In this study we postulated that penile arterial diastolic flow was related to corporal venoocclusive dysfunction and reduced outflow resistance. The RI was used to quantify and integrate diastolic flow relative to systolic flow. ~4 Flow parameters were measured at the time of the maximal erectile response to standardize the timing of measurement and to assess arterial flow when the venoocclusive mechanism was maximally functional. In most patients studied, this occurred after a period of privacy and selfstimulation, the otherwise measured flow parameters after intracavernosal injection of papaverine before self-stimulation were selected. Because of the potential error in measurement of the luminal diameter and the small caliber of the penile arteries, we did not consider this parameter or the derived volumetric flow rates in our analysis. 12'22 Our results indicate that in normal controls with complete erections the EDV was less than 4.5 cm/sec the PSV was 35 cm/sec or greater. In this group the RI approached 1 (mean, )

7 Volume 13 Number 1 January Venogenic irapotence diagnosed by duplex sonography 81 indicating an intact venoocclusive mechanism. We therefore considered EDVs of 4.5 cm/sec or more indicative of corporeal venous leakage and PSVs of 35 cm/sec or less indicative of arterial insufficiency. Based on these criteria, several distinct flow patterns could be identified. In one group significant flow occurred during diastole represented by EDVs greater than 4.5 cm/sec and normal arterial inflow evident by PSVs of 35 cm or greater. In these patients the RI was significantly lower than in normal controls (0.71 +_ 0.10) indicating venoocclusive dysfunction and corporal venous leakage (group A). In another group minimal or no diastolic flow was detected, yet there was a concomitant incomplete erectile response. These patients were suspected of having arterial insufficiency (group B). Sixteen patients in this group had PSVs less than 35 cm/sec, whereas six had PSVs ranging from 37 to 44 cm/sec. Those with PSVs greater than 35 cm/sec were thought to have a psychogenic component associated with excessive adrenergic stimulation and vasoconstriction. This may have resulted in low PSVs and incomplete erections. Other investigators have also described psychogenic inhibition as a cause of erectile dysfunction in the overtly anxious patient. 1'~s'2~ Although we calculated the PBPI in only 23 subjects, this measurement was not different between controls with normal erections and those patients suspected of having penile arterial insufficiency. These findings corroborate with other studies that have questioned the reliability of the PBPI in the evaluation of cavernosal arterial flowy 27 More recently Schwartz et al. 28 have also found a poor correlation between the PBPI and angiography. To validate the duplex ukrasonographic findings indicative of corporal venous leakage, we performed infusion pharmacocavernosometry and cavernosography. Venous leakage during this test was quantitated by determining the intracorporal flow rate of saline required to sustain a full erection or attain an intracavernosal pressure of 90 to 100 mm Hg. 7"16"I7 This test is recognized as the most reliable method for determining the presence and severity of venous leakage. 1"16'17 In 19 of 21 patients suspected of venous leakage, IPCC was confirmatory. In addition, we found a strong inverse correlation between the tli and the maintenance flow rate (p < 0.001). This relationship between RI and pharmacocavernosometry has not been previously reported. The RI appears to be a reliable noninvasive quantitative measure of venous leakage. In conclusion, we have demonstrated that duplex ultrasonography is useful in discriminating between venogenic and arteriogenic impotence. The presence of penile cavernosal arterial diastolic flow strongly suggests corporal venous leakage. In addition, the RI provides a useful measure of the severity of corporal venous leakage. Our findings concur with the previously reported duplex ultrasonographic criteria of arterial insufficiency; however, further study is needed to correlate systolic flow parameters with phauoarteriographic evidence of arterial occlusive disease. Penile duplex ultrasonography appears to be a reliable noninvasive tool in the initial evaluation of patients with suspected vasculogenic impotence. The measured and derived flow parameters may thus help in selecting those patients in need of additional invasive studies and assist in determining the most appropriate treatment of this common form of erectile dysfunction. The authors gratefully acknowledge the technical assistance of Rhonda P. Poindexter, MS. REFERENCES 1. Krane R, Goldstein I, Saenz de Tejada I. Medical Progress: impotence. N Engl J Med 1989;321: de Groat WC, Steers WD. Neuroanatomy and neurophysb ology of penile erection. In: Tanagho EA, Lue TF, McClure RD, eds. Contemporary management of impotence and infertility. Baltimore: Williams & Wilkins, 1988: Lue TF, Tanagho EA. Functional anatomy and mechanism of penile erection. In: Tanagho EA, Lue TF, McClure RD, eds. Contemporary management of impotence and infertility. Baltimore: Williams & Wilkins, 1988: Fournier GR Jr, Juenemann K-P, Lue TF, Tanagho EA. Mechanisms of venous occlusion during canine penile erection: an anatomic demonstration. J Urol 1987;137: Lue TF, Takamura T, Schmidt RA, et al. Hemodynamics of erection in the monkey. J Urol 1983;130: Goldstein AMB, Meechan JP, Zakhary R, Buckley PA, Rogers FA. New observations on microarchitecture of corpora cavernosa in man and possible relationship to mechanism of erection. Urology 1982;20: Wespes E, Schulman CC. Venous leakage: surgical treatment of a curable cause of impotence. J Urol 1985;133: Zorgniotti AW. Padula G, Shaw WW. Selective arteriography for vascular impotence. World J Urol 1983;1: Bookstein JJ, Valji K, Parsons L, Kessler W. Pharmacoarteriography in the evaluation of impotence. J Urol 1987; 137: Delcour C, Wespes E, Schulman C, Struyven J. Investigation of the venous system in impotence of vascular origin. Urol Bad 1984;6: Lue TF, Hricak H, Marich KW, et al. Vasculogenic impotence evaluated by high-resolution ultrasonography and pulsed Doppler spectrum analysis. Radiology 1985; 155: Quam JP, King BF, James EM, et al. Duplex and color Doppler sonographic evaluation of venogenic impotence. AJR 1989;159: Juenemann K-P, Lue TF, Fournier GR Jr, Tanagho EA. He-

8 82 Bassiouny and Levine Journal of VASCULAR SURGERY modynamics of papaverine and phentolamine-induced penile erection. J Urol 1986;136: Planilo T, Pourcelot L, Pottier JM, et al. Etude de la circulation carotidienne par les methodes ultrasoniques et la thermographie. Rev Neurol (Paris) 1972;126: DePalma R. Anatomy and physiology ofmale sexual fimction. In: Giordano J, Trout H, DePalma R, eds. The basic science of vascular surgery. Mt. Kisco NY: Futura, 1988: Puyau FA, Lewis RA. Corpus cavernosography: pressure flow and radiography. Invest Radiol 1983;18: Lue TF, Hricak H, Schmidt RA, et al. Functional evaluation of penile veins by cavemography in papaverine-induced erection. J Urol 1986;135: Herbert J. The role of the dorsal nerves of the penis in the sexual behavior of the male rhesus monkey. Physiol Behav 1973;10: Virag R, Bouilly P, Frydman D. Is impotence an arterial disorder? A study of arterial risk factors in 440 impotent men. Lancet 1985;1: Levine LA. Erectile dysfunction: causes, diagnosis, and treatment. Comprehensive Therapy 1989;15: Lue TF, Mueller SC, Jow YR, Hwang T I-S. Functional evaluation of penile arteries with duplex ultrasound in vasodilator induced erection. Urol Clin North Am 1989;16: Levine LA, Bassiouny HS. Measurement of venogenic impotence by penile duplex ultrasonography. J Urol 1990; 143:211A. 23. Blaivas JG, O'Donnel TF, Gottlieb P, Labib KB. Comprehensive laboratory evaluation of impotent men. J Urol 1980;124: Abber JC, Lue TF, Orvis BR, McClure RD, Williams RD. Diagnostic tests for impotence: a comparison of papaverine injection with the penile-brachial index and nocturnal penile tumenscence monitoring. J Urol 1986;135: Padma-Nathan H, Klavans S, Goldstein I, Krane RJ. The screening efficacy of PBI versus duplex ultrasound versus cavernosal artery systolic occlusion pressure. Proceedings of the Third Biennial World Meeting on Impotence, Boston, Mass., p. 32, October 6-9, Metz P, Bengtsson J. Penile blood pressure. Scand J Urol Nephrol 1981;15: Reiss HF. Difficulties in Doppler auscultation of cavernous arteries of the penis. Urology. 1985;26: Schwartz AN, Lowe MA, Ireton R, et al. A comparison of penile brachial index and angiography: evaluation of corpora cavernosa arterial inflow. J Urol 1990;143: DISCUSSION Dr. Ralph DePalma (Washington, DC). These authors recommend the use of duplex sonography as a means of diagnosing venous leakage, which it accomplished in 19 of 21 instances, and also as a means of differentiating impotence caused by arterial and mixed arterial and venous insttfficiency. A key variable after the injection ofpapaverine is time. The time that one selects to do the ultrasonographic examination is critical. For example, a slower erectile response exists in an older man; if this is studied early one would certainly detect a venous leak in the presence of adequate arterial inflow and normal venoocclusive mechanism. I wish to emphasize the relationship of the timing of the application of the duplex ultrasonic study to the observations that occur after 60 mg of papaverine injected into the corpora. I have two comments. We have limited the use of penile duplex sonography to those cases in which we suspect cavernosal abnormalities such as fibrosis or Peyronie's disease. We hope thereby to avoid unnecessary additional invasive testing by use of cavernosography or arteriography. Second, before any arterial or venous surgical intervention for impotence, we require both dynamic cavernosography and selective arteriography since the screening methods, I believe, are not as accurate as they might be. The rationale for this insistence is that both processes contribute to erectile failure, and the most clear anatomic information is needed for surgical decision making. My questions derive from these concepts. How were the eight controls with abnormalities studied? We considered Peyronie's to exhibit leakage in many instances. Some erectile responses that are normal will be achieved slowly. What were the exact time intervals after papaverine injection when flow velocity diagnostic data were calculated? Was it always 5 to 7 minutes, or what were the times for maximum erections to be obtained in privacy? Second, if penile sonography can select therapy, how would it do so? Would not the equivalent arteriography or venography be required for intelligent surgical intervention? Are you happy with the diagnostic specificity of this test? And finally, how much does the application of penile duplex sonography add to the cost of a routine work up of patients who are candidates for treatment? Dr. Charles Brantigan (Denver, Colo.). The authors group the patients into different groupings according to what they suspected was the cause. In the patients who were suspected of having an arterial cause, which was then confirmed by means of the duplex scanning, how was the conclusion verified that it was in fact an arterial abnormality so that duplex scanning could then confirm that? Mr. Andrew Nieolaides (London, United Kingdom). This work shows yet again the application of duplex scanning. The authors have demonstrated the significance between the means of normal and abnormal. What we would like to see is the overlap between dae groups. Unless you give us a figure of the overlap--if there is a 50% overlap

9 Volume 13 Number 1 January Venogenic impotence diagnosed by duplex sonography 83 for example--this is not a test that is going to become accepted as a routine screening test. Could you please give us an idea of the overlap? Dr. Bassiouny. The timing of duplex ultrasonography after intracavernosal papaverine injection is an important issue, since deep cavernosal arterial flow velocity varies during the course of the erectile response. In this study we elected to measure the flow parameters at the time of the maximal erectile response, which usually occurred 5 to 7 minutes after papaverine injection or after self-stimulation. Only those measurements obtained during the maximal erectile response were used for analysis. With regard to the controls, we recognize that two patients had Peyronie's disease; however, these patients had normal erections and denied any history of transient or fading erections suspicious of venous leakage. In this study penile duplex ukrasonography was useful in screening patients suspected of venous leakage. The presence of significant diastolic flow (>5 cm/sec) and a low resistance index were both indicative of corporal venous leakage. In these patients, further investigation with more invasive studies such as cavernosometry and cavernosography is needed to confirm the severity and site of venous leakage before vascular reconstruction. At this time it is premature to promote this diagnostic modality as the most definitive diagnostic test before surgery. The cost of penile duplex ultrasonography is $450 at our institution. This is considerably less expensive than IPCC, which can vary in cost from $750 to $1200. In response to Dr. Brantigan's question, I agree with you that further studies are needed to correlate PSV measurements with arteriographic evidence of intra- and extrapelvic arterial occlusive disease. We did measure the PBPI in 24 patients. It is important to note that the PBPI did not differ between control subjects and those patients suspected of penile arterial insufficiency, yet the PSV measurements of the latter group were significantly lower than controls. Mr. Nicolaides, we did note minimal overlap between the various groups in regard to the velocity measurements. For example, in patients suspected of arterial insufficiency, four had PSVs over 35 cm/sec. None of these patients had significant diastolic flow. All four patients had psychogenic impotence that can be associated with excessive alpha adrenergic discharge and vasoconstriction. This may explain why their incomplete erections were accompanied with PSVs slightly greater than 35 cm/sec.