Authors from Japan describe their results with penile revascularization in patients who had arteriogenic erectile dysfunction. They recommend that this type of surgery is most suited to younger patients, and suggested that attention be paid to the long-term outcome and the long-term adverse events.
To determine the subjective and objective outcomes (by survival analysis) after penile revascularization surgery in patients with arteriogenic erectile dysfunction (ED), selected by established strict criteria.
PATIENTS AND METHODS
The study included 51 patients diagnosed with arteriogenic ED caused by localised arterial lesions and who had microscopic penile revascularization surgery between January 1996 and March 2002. Before surgery, all patients had a full examination, including a medical and sexual history, laboratory testing, intracavernosal pharmacological tests, colour Doppler ultrasonography (CDU), pharmacodynamic infusion cavernosometry and cavernosography, and digital subtraction angiography (DSA). Penile revascularization surgery was indicated only in patients aged <50 years and with no history of diabetes mellitus, hypertension or hyperlipidaemia. When there were communicating branches between the dorsal and cavernosal arteries, Hauri's procedure was used; when there were none or there was no evidence for them on both CDU and DSA because of severe narrowing or obstruction in the proximal common penile artery, the Furlow-Fisher modification of the Virag V procedure (FFV5) was used. The patency of the neo-arterial blood flow was assessed by CDU and effective rates calculated using the Kaplan-Meier method. The efficacy rate was recalculated whenever there was a recurrence. When occlusion of the neo-arterial blood was confirmed by CDU the date of occlusion was set as that midway between the last examination showing patency of the donor vessel and the latest examination indicating the occlusion. The patency period was the number of days from surgery to the date of occlusion.
Of the 51 patients, 26 had Hauri's and 23 the FFV5 procedure (median age 32 years, range 21–49); in two patients with a previous pelvic fracture surgery was not possible because of scar formation in the dorsal area at the base of the penis. The mean (sd) subjectively estimated efficacy rate was 85.9 (6.3)% after 3 and 67.5 (10.7)% after 5 years of follow-up. The duration at 75% efficacy was 41.0 (5.6) months. The objectively estimated efficacy rate was 84.9 (7.3)% at 3 and 65.5 (13.5)% after 5 years of follow-up. The duration at 75% patency was 42.4 (9.5) months, and at 50% was 60.6 (19.4) months. There was no significant difference in subjective outcome between the FFV5 and Hauri procedures (P = 0.38, log rank test) and none objective outcome after surgery (P = 0.19, log rank test). Thirteen of the 18 patients in the Hauri group had venous dilatation in the deep dorsal, obturator, prostatic and the internal iliac veins. There were operative complications in four patients (hyperaemia of the glans in two, and one each with haemorrhage from the anastomosis site and scar contracture).
The long-term efficacy rates (by the Kaplan-Meier method) of the Hauri and FFV5 procedures were both acceptable. The selection criteria gave acceptable outcomes from both procedures. Penile revascularization surgery is a treatment suitable only for young men and therefore attention must be given not only to the long-term outcome but also to long-term adverse events.
colour Doppler ultrasonography
pharmacodynamic infusion cavernosometry and cavernosography
digital subtraction angiography
Furlow-Fisher (modified Virag procedure).
Penile revascularization surgery is a rational and effective treatment for erectile dysfunction (ED) resulting from localized arterial lesions. This surgery was first adopted in the 1970s . As a result of many consecutive intensive studies, the indicative criteria and surgical mode had largely been established by the late 1990s. According to those early studies, penile revascularization surgery is highly effective when the patients have only a localized arterial lesion with no cardiovascular risk factor, but when the surgery is used in patients with diffuse vascular lesions or with impaired cavernosal function, the outcome of this treatment is significantly worse [2,3].
The purpose of the present study was to evaluate the long-term outcome, one of the remaining issues of penile revascularization surgery. According to relevant reports, the longer the follow-up the worse the effective outcome . However, previous studies have been retrospective and did not consider the censored cases . In this study we evaluated the long-term outcome of penile revascularization surgery, taking censored cases into consideration. We used penile revascularization surgery in patients selected by established criteria and calculated the subjective and objective outcomes using the Kaplan-Meier method.
However, another issue with this surgery that is yet to be resolved is the method of evaluating the outcome. To date, the new blood flow has been assessed mainly by colour Doppler ultrasonography (CDU) , a technique we also used. However, currently the resolution of CT angiography (CTA) has improved significantly and therefore, as in other fields, CTA is expected to become a useful tool for the diagnosis of penile blood flow . Thus we also evaluated a subgroup of patients after surgery using CTA.
PATIENTS AND METHODS
The study included 51 patients diagnosed with arteriogenic ED caused by localized arterial lesions and who had microscopic penile revascularization surgery between January 1996 and March 2002. Before surgery all patients had a full assessment of their ED, including a medical and sexual history, laboratory testing, intracavernosal pharmacological tests, CDU, pharmacodynamic infusion cavernosometry and cavernosography (DICC), and digital subtraction angiography (DSA). The cause of the localized arterial lesions was blunt perineal trauma in 33 and unknown in 35 men. Penile revascularization surgery was indicated only in patients aged <50 years and with no history of diabetes mellitus, hypertension or hyperlipidaemia. Patients with any cardiovascular risk factors other than smoking were excluded. Intracavernosal injection tests were used, with prostaglandin E1 and dose-escalation (from 5 to 20 µg). DICC was used after pre-treating with an intracavernosal injection of 60 mg of papaverine. A flow of <10 mL/min capable of maintaining 90 mmHg of intracavernosal pressure indicated normal cavernosal function, and was considered as an indispensable criterion. Moreover, to eliminate patients with a neurogenic disorder it was essential that the nocturnal penile tumescence records were not a flat trace. When there were communicating branches between the dorsal and cavernosal arteries, Hauri's procedure was used ; when there were none or there was no information about them on both CDU and DSA (because of severe narrowing or obstruction in the proximal part of the common penile artery) the Furlow-Fisher (FFV5) procedure was used . Heparin (5000 units) was administered intravenously for the first 24 h after surgery and 100 mg acetylsalicylic acid daily for 1 year.
The subjective effectiveness was evaluated by direct interview and this outcome classified in two categories. According to the National Institutes of Health definition , when a patient can achieve and maintain an erection of sufficient rigidity and duration to permit satisfactory sexual performance with no additional treatment, the outcome of the surgery is considered a success; when the patient is unable to do so with no additional treatment, the outcome is considered a failure. Patency of the neo-arterial blood flow was examined by CDU; since January 2000 we added CTA evaluation to examine the neo-arterial blood flow. The fundamental follow-up interval was monthly until 1 year, every 3 months until 2 years, and every 6 months from the third year after surgery for ≥ 5 years. When the patient subjectively reported diminished erectile function, additional examinations, including an intracavernosal injection test, CDU and CTA, were used. When occlusion of the neo-arterial blood flow was confirmed by CDU the date of occlusion was determined as that midway between the last examination showing patency of the donor vessel and the latest indicating occlusion. Therefore, the patency period was defined as the time from surgery to the determined date of occlusion. If a patient did not appear on the day of his appointment, he was contacted by telephone or mail. When the patient did not visit by the next appointment he was assumed to be ‘censored’, and the follow-up defined as the time from surgery to the last visit. Patients who subjectively reported being able to maintain erectile function and those who objectively maintained neo-arterial flow were both listed as ‘censored’.
Efficacy rates were calculated by the Kaplan-Meier method; unknown cases and patients whose erectile function was still intact were both considered to be ‘censored’. The efficacy rate was recalculated on each occasion that there was a recurrence. Statistical differences were analysed by the log-rank test.
In all, 65 men were diagnosed as suitable candidates for surgery, of whom 51 finally selected surgical treatment; of these, 26 had the Hauri and 23 the FFV5 procedure (median age 32 years, range 21–49). Surgery was not possible in two patients with a previous pelvic fracture, because of scar formation in the dorsal area at the base of the penis.
Table 1 shows the variables from the subjective and objective evaluations by success and failure, and by the procedure, with the Kaplan-Meier plots shown in Fig. 1. There was no significant difference in subjective outcome between the FFV5 and Hauri's procedures (P = 0.38, log-rank test, Fig. 1d), nor in objective outcome (P = 0.19, log-rank test, Fig. 1e).
|Median (range) follow-up, months||21.3 (0–72.1)||15.1 (0–70.7)|
|Mean (sd) efficacy rate, %|
|3 years||85.9 (6.3)||84.9 (7.3)|
|5 years||67.5 (10.7)||65.5 (13.5)|
|Duration of effective period, months|
|75%||41.0 (5.6)||42.4 (9.5)|
|Follow-up||37.4 (0–72.1)||22.0 (0–66.8)|
|Efficacy rate, %|
|3 years||87.7 (8.2)||75.5 (10.8)|
|5 years||70.9 (12.6)||64.8 (13.6)|
|Duration of effective period, months|
|75%||45.6 (6.56)||41.3 (12.5)|
|Follow-up||15.1 (0–66.8)||12.2 (0–70.7)|
|3 years||87.7 (6.7)||100|
|5 years||65.8 (19.6)||66.7 (27.2)|
|Duration of effective period, months|
|75%||37.9 (26.5)||42.4 (34.6)|
Comparing the subjective evaluation and patency of neo-arterial blood flow in the seven patients whose donor vessels were occluded, four relapsed into ED but three remained potent. These three patients were treated with the FFV5 procedure. However, after Hauri's procedure there were two patients who relapsed into ED instead of patency of the neo-arterial blood flow. Examining the vascular configuration of these two patients with CTA, there were some differences from normal postoperative images (Fig. 2). Blood flow in the proximal portion of the deep dorsal vein was patent but that to the distal portion and to the dorsal artery was not visible (Fig. 3).
CTA was also used in nine patients after FFV5 and 18 in after the Hauri procedure; 13 of the latter showed venous dilatation of the deep dorsal, the obturator, prostatic and internal iliac veins (Figs 4 and 5). These findings were not apparent in the FFV5 group (Fig. 6).
There were operative complications in four patients, i.e. hyperaemia of the glans in two, and one each with haemorrhage from the anastomotic site and scar contracture. The hyperaemia and haemorrhage were controlled conservatively, but the scar contracture required surgical correction.
Deep dorsal vein arterialization surgery for arteriogenic ED was invented by Virag et al. and modified by Furlow et al.. Later, in 1986, Hauri  reported a new surgical procedure which created an anastomosis between the inferior epigastric artery and the deep dorsal vein and dorsal artery of the penis. Lobelenz et al. subsequently modified Hauri's procedure to a triple-vessel anastomosis. This last method was intended to supply neo-arterial flow to both the artery and vein of the penis. Concomitant with these surgical developments the indications for revascularization were also assessed. Although there are minor differences the criteria for surgery for arteriogenic ED have been largely established [13,14]. The accepted criteria include: age <50 or 55 years, a negative response to intracavernosal injection tests, fewer than one or two cardiovascular risk factors, no history of diabetes mellitus, and normal corporal veno-occlusive function.
When the patient agrees with the operative criteria the next step is to select the surgical procedure. Basically, the surgical procedure for revascularization should be determined according to the anatomical variation of the penile artery and the pattern of arterial obstruction. However, clinically it is not possible to obtain detailed information about the arterial pattern. When there are severe stenosis or complete occlusions in the proximal region of the penile artery, reliable images of the peripheral region cannot be obtained. In such cases it is often not possible to determine whether there are significant communicating branches between dorsal and cavernosal arteries, or from which part of the common penile artery the cavernosal arteries arise. We selected Hauri's procedure in patients for whom there was reliable information from CDU and DSA about the communicating branches between dorsal and cavernous arteries. However, when there was no information about these branches or no significant communication between dorsal and cavernosal arteries, we selected the FFV5 procedure, because in these cases there is no assurance that the penile blood flow will increase after surgery to connect an anastomotic donor vessel to the dorsal artery of the penis. Our selection criteria produced satisfactory outcomes from both procedures, although there were too few patients to reach definite conclusions about the selection criteria. However, judging from the outcomes to date, our criteria seem reasonable.
There has been criticism about the evaluation method for the outcome of penile revascularization [15,16]. However, this problem is not easily solved, because evaluating the erectile function of men with arteriogenic ED has some difficulties. In some instances patients with arteriogenic ED might show a normal erectile response during sleep . For an objective evaluation after surgery CDU has been used to assess the patency of the donor vessel, instead of erectile function tests [13,14]. We also examined the neo-arterial flow by CDU, but care is needed when comparing the present results with those previously published, because of differences in accounting for patients who cannot be followed up. Generally, the longer the follow-up the more patients are lost to follow-up. When the long-term outcome of a treatment is assessed there are often difficulties in obtaining the exact outcome because of changes in patients’ residence, death from another illness, or other reasons. Furthermore, as a matter of course, there are many patients who will subjectively maintain that their treatment was effective on the audit day. Patients lost during the follow-up and those who maintained their effectiveness were defined as ‘censored’ in the present survival analysis. Patients who maintain good function lose their motivation to visit the hospital, and therefore have a tendency to be lost from follow-up. However, there are patients who repeatedly visit the hospital seeking additional or alternative treatments for ED. If the efficacy rate of a given treatment is calculated without considering the censored cases the long-term outcome could be underestimated. Thus the long-term outcome of penile revascularization surgery must be evaluated using survival analysis.
The subjectively reported efficacy of penile revascularization surgery is 40–85%[18–24]. These authors noted that prolonged patency of the anastomosis requires a technically sound procedure and that the afferent epigastric artery must not be kinked but linear. However, a proper subjective evaluation of penile revascularization surgery is more difficult to obtain than an objective evaluation, because of the influence of the patient's collaborative status, along with secondarily developed psychogenic factors. Nevertheless, while we cannot entirely eliminate the influences of differences in patients’ attitudes or psychogenic factors, we can minimize any underestimate by using the category of censored cases with survival analysis.
One report  noted differences between the subjective and objective evaluation of penile revascularization surgery. Among the present patients, two had patent new blood flow on CDU but subjectively lacked normal erectile function. In these two patients there was partial occlusion in the triple-vessel anastomosis. It is sometimes difficult to evaluate the patency of the triple-vessel anastomosis by CDU because of pulsing in the arterialized deep dorsal vein and turbulent flow in the vein nearby. While CDU is used as a standard examination for evaluating the donor vessel, it might not be fully reliable. When blood flow in the epigastric artery is diminished or stopped, CTA cannot visualize this artery clearly, although in cases when the blood flow is maintained, CTA can visualize the donor vessel and recipient vessels clearly. Therefore, CTA is necessary for a definitive evaluation in some cases. The significant increase in blood flow detected in the prostatic plexus, the internal pudendal vein, the obturator vein, and the internal iliac vein in Hauri's procedure may also occur in a similar procedure that does not involve ligating the proximal side of the arterialized deep dorsal vein. The long-term effect of increased venous blood flow on the pelvic organs is unknown but this phenomenon would have some adverse effects if these patients later had pelvic or transurethral surgery, or a pelvic injury. The surgeon would not anticipate the increased venous blood flow. To protect against possible risks patients should be informed about this phenomenon. We intend to explain this latent risk to patients before surgery. Penile revascularization surgery is only useful in young patients and therefore sufficient attention must be given not only to the long-term outcome but also to the long-term adverse effects and risks.
We are grateful to Mr Masataka Minese, a radiological technologist who provided the exquisite CTAs of the internal pudendal arteries, and Dr Nobuyuki Kurooka, a radiologist. We also thank my friend Mr Graham Page, who has helped me consistently with this English manuscript.
CONFLICT OF INTEREST