Original Research

International Journal of Impotence Research (2005) 17, 354–358. doi:10.1038/sj.ijir.3901290 Published online 10 February 2005

Combining programmed intracavernous PGE1 injections and sildenafil on demand to salvage sildenafil nonresponders

P Gutierrez1, P Hernandez1 and M Mas1

1Department of Physiology and CESEX, Faculty of Medicine, Campus CC Salud, University of La Laguna, Tenerife, Spain

Correspondence: M Mas, MD, PhD, Department of Physiology and CESEX, Faculty of Medicine, Campus CC Salud, University of La Laguna, Tenerife 38071, Spain. E-mail: mmas@ull.es

Received 9 August 2004; Revised 28 September 2004; Accepted 8 October 2004; Published online 10 February 2005.



In a prospective, placebo-controlled, one group crossover design study, we tested whether adding programmed intracavernous PGE1 injections (IC-PGE1) can improve the effectiveness of sildenafil in erectile dysfunction (ED) patients unresponsive to monotherapy with this drug. In all, 40 ED patients who had experienced unsatisfactory erections with both the 50 and 100 mg sildenafil doses were treated with four bi-weekly 20 mug IC-PGE1 injections given in the clinic and provided with either placebo or 50 mg sildenafil capsules for the next 4 weeks. Thereafter, they were crossed over to the other oral treatment for an additional 4-week period. The IIEF-Erectile Function domain score (IIEF-EFS), the main outcome measure, was found considerably higher (P<0.001) with the combined IC-PGE1–50 mg sildenafil treatment than with IC-PGE1–placebo or sildenafil alone (50 or 100 mg) in a subset of 26 subjects (65%). They thus shifted from the 'severe' or 'moderate' to the 'mild' grading of ED classification.


erectile dysfunction, oral PDE-5 inhibitors, intracavernosal PGE1, combination therapy



Oral therapy with phosphodiesterase type 5 (PDE-5) selective inhibitors has become the preferred first-line treatment for most cases of erectile dysfunction (ED).1 Although these drugs have proved remarkably effective, safe, and easy to use, there are still many ED patients, roughly one out of three,2 who do not respond satisfactorily to monotherapy with PDE-5 inhibitors. For these cases, a second-line, more invasive treatment option is the local administration of a number of vasoactive drugs. Prostaglandin E1 (PGE1, alprostadil) has become the preferred agent for this therapeutic modality because of its high efficacy and safety profile.3 This drug must be administered by intracavernosal (IC) injection or the transurethral route, the former being more commonly used. While being also highly effective IC injections on demand pose several problems for many patients eventually translating into a high dropout rate.3, 4 They include needle aversion, lack of spontaneity, partner rejection of the procedure, pain, and increased risks of priapism or fibrotic changes in the penis after prolonged use.5, 6, 7, 8, 9

In patients not responding to either type of monotherapy, the combination of both has been tried in order to get additive or synergistic effects, which has been carried out by combining PGE1 delivered either transurethrally10, 11 or by IC injection (as triple agent, mixed with papaverine and phentolamine)12 with the oral PDE-5 inhibitor sildenafil at 100 mg dose. These studies reported satisfactory results in many patients previously not responding to monotherapy with either drug. However, as both drugs were taken shortly (within 60 min) before intercourse, this approach shares some of the above-pointed drawbacks of locally delivered PGE1 monotherapy on demand.

Here, we propose that the effectiveness of PDE-5 inhibitors taken on demand can also be improved substantially in previously refractory patients by treating them with a few scheduled IC-PGE1 injections. The rationale for this approach stems from the finding in the animal model that repeated IC-PGE1 injections increased the release of nitric oxide (NO) in the corpora cavernosa and the erectile response to cavernous nerve stimulation. Such changes were associated with an increased expression of the 'constitutive' NO synthase isoforms (nNOS and eNOS) in the corpora.13 We reasoned that if similar mechanisms operate in the human, repeated IC-PGE1 injections would enhance the NO-cGMP pathway and thereby the effectiveness of PDE-5 inhibitors.

To test the above hypothesis, a prospective placebo-controlled study was conducted on ED patients who had responded poorly to sildenafil monotherapy. We assessed whether adding programmed IC-PGE1 injections would improve their response to sildenafil taken on demand. This report summarizes its results.


Patients and methods


All the procedures complied with National Regulations for Human Research and were approved by the Institutional Review Board of the University of La Laguna Hospital and the Spanish Ministry of Health. The subjects were recruited from patients attending de novo the ED clinic at the Andrology Unit of the University of La Laguna Hospital. The inclusion criteria for entering the initial selection phase were: having a clinical diagnosis of ED lasting more than 6 months, being 18 y or older, being in a stable relationship, and not having received any previous therapy for their ED. The diagnosis of ED and its etiological appraisal was based on medical and sexual history, physical examination, and complementary tests as required. Patients were excluded from this study if they had penile deformities, uncontrolled cardiovascular, endocrine, renal, hepatic, hematological or psychiatric disorders, retinitis pigmentosa, or were taking nitrates or androgens.

Study design

A prospective, one group placebo-controlled, crossover design study was planned to assess the effects of programmed IC-PGE1 injections on the erectogenic efficacy of 50 mg sildenafil taken on demand in previous nonresponders to the 50 or 100 mg doses of this drug. The Erectile Function domain score of the International Index of Erectile Function (IIEF-EFS), a well-recognized sensitive and reliable measure of treatment efficacy,14 was used as the main outcome measure.

The study schedule is summarized in Figure 1. In the first phase, patients were selected on the basis of not responding with satisfactory improvement of their erectile function first to 50 mg sildenafil and then to the 100 mg dose taken on demand for 4 weeks each with a minimum of four coital attempts with each dose. Patients failing to respond to these treatments were offered to participate in the second phase of the study, the trial proper. Thus, in the selection phase, we recruited as many patients as needed to obtain a group of 40 patients who had proved refractory to sildenafil monotherapy with both the 50 mg and the 100 mg doses and were willing to participate in the IC-PGE1–sildenafil combination trial.

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Study design (a) and flowchart of patients (b).

Full figure and legend (110K)

The patients selected for the IC-PGE1–sildenafil combination trial received an explanation about the purpose of the investigation and were told that in addition to the IC injections, they will receive different doses of a safe, approved oral agent on a blind basis. They were asked to sign an informed consent form. The consenting subjects received four IC 20 mug PGE1 injections given in the office every 2 weeks. In each of these visits, the patients were also provided with eight black gelatin capsules containing either a 50 mg sildenafil tablet or placebo (talcum powder) to be taken 1 h before intercourse. The subjects were advised not to take the capsules until at least 24 h after receiving the injection. In all, 20 subjects started with sildenafil and the other 20 with placebo and stayed on these treatments for the ensuing 4 weeks. Then they were crossed over to the other oral treatment for an additional 4-week period. The IIEF questionnaire was administered in the office by the same interviewer during the periodical visits at the end of the successive treatment courses.

Statistical analysis

The successive IIEF scorings recorded from each patient were compared with nonparametric tests for repeated measures, namely Friedman's two-way analysis of variance by ranks followed by post hoc comparisons with the Dunn's test. Nonparametric comparisons between two independent sets of values were made using the Mann–Whitney test. All the analyses were made with the GraphPad Prism4 statistical package (GraphPad Software, San Diego, California, USA, www.graphpad.com).



As shown in Figure 1b, up to 147 patients had to enter the selection phase in order to recruit the specified number of subjects (40) for the IC-PGE1–sildenafil 50 mg combination trial. Of the initial sample, 103 (70%) responded satisfactorily to either the 50 mg or the 100 mg sildenafil doses and were thus excluded from this study. Of the 44 patients found unresponsive to both doses of sildenafil monotherapy, four declined to participate in the IC-PGE1–sildenafil 50 mg combination trial.

Of the 40 patients completing the trial, their mean (plusminuss.d.) age was 58.3plusminus12.7 y and the duration of ED 3.3plusminus2.9 y. The etiology of their ED was characterized as organic in 21 subjects, psychogenic in seven and mixed in 12. Significant comorbidities were frequent in these patients: 20 had diabetes, 19 hypertension, 12 dyslipidemia, three neurological conditions, and nine had undergone nerve-sparing prostatectomy.

As for the treatment end points, no 'carry-over effect', that is, differences in the outcome depending on which treatment (placebo or sildenafil) was taken first, was detected. Therefore, the recorded scores for each of these treatments were pooled regardless of the order in which it had been given. Figure 2 summarizes the main findings for the entire cohort of 40 patients completing the study. It can be seen that the EF domain score was significantly higher with the combination of programmed IC-PGE1 and sildenafil 50 mg on demand than with any other treatment. Consistently with this finding both the IIEF 'intercourse satisfaction' (Q6–8) and 'global satisfaction' (Q13–14) scores reported by these patients were also the highest following the combination treatment (P<0.01 and P<0.05, respectively, vs all the other treatments).

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Effects of the different treatments on the IIEF-EFS scores of the 40 subjects participating in the trial. SIL 50: 50 mg sildenafil; SIL 100: 100 mg sildenafil; PGE: 20 mug IC-PGE1; PL: oral placebo. Bars show the mediansplusminusinterquartile ranges. *P<0.01 vs sildenafil 100 mg, **P<0.001 vs sildenafil 50 mg and IC-PGE1–placebo.

Full figure and legend (55K)

We also noticed that the EF domain scores recorded after the IC-PGE1–sildenafil 50 mg combination therapy actually showed a bi-modal distribution, which is reflected in the asymmetry of the interquartile range corresponding to this treatment (viz., a long 'low tail', Figure 2). Thus, whereas a majority of the patients receiving the IC-PGE1–sildenafil 50 mg combination scored fairly high, there were also some subjects reporting a poor response. It was for that reason that we chose nonparametric tests to analyze the present data. Therefore, the 40-subject cohort was split into two subgroups, 'responders' and 'nonresponders' to the IC-PGE1–sildenafil 50 mg combination therapy. The 'responder' category included those patients in whom the IC-PGE1–sildenafil 50 mg treatment resulted in an increase in their EF domain score of >5 points over the value recorded when they were taking sildenafil 50 mg alone. As shown in Figure 3, 26 patients (65%) met such criterion. The median increase in their EF domain score over the value attained with the sildenafil 50 mg alone treatment was 10 points. In terms of clinical severity of ED, it meant shifting from the 'severe' or 'moderate' to the 'mild' grading of ED classification,14 which reflected also in their values for both 'satisfaction scores' of the IIEF, which these patients rated highest when they were having the IC-PGE1–sildenafil 50 mg combination therapy.

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Effects of the different treatments on the IIEF-EFS scores of the 'responder' and 'nonresponder' subsets of patients. See legend to Figure 2 for abbreviations. Mediansplusminusinterquartile ranges. **P<0.001 vs any other treatment. The ED severity grading scale of Capelleri et al14 is shown at the right side.

Full figure and legend (72K)

The difference between the 'responder' and 'nonresponder' groups was especially conspicuous in the ability to achieve penetration following the combined IC-PGE1–sildenafil 50 mg treatment. That was evidenced by their median scores on Q3 ('when you attempted sexual intercourse, how often were you able to penetrate your partner?') whose values were, respectively, 5 and 1 (P<0.001, Mann–Whitney test).

We were unable to identify clearly any single factor predictive of whether our subjects would respond satisfactorily or not to the IC-PGE1–sildenafil 50 mg combination therapy. No differences were detected in the serum androgen levels of the responder and nonresponder subjects. The recorded comorbidities were distributed within the two groups in proportion to their size, for example, of the nine prostatectomized subjects, there were six responders and three nonresponders. Likewise, 12 of our diabetic subjects responded positively to the combination therapy and eight were unresponsive. Interestingly, though, all these unresponsive diabetics had also hypertension, whereas there were only seven diabetic and hypertensive subjects in the larger responder group. It is conceivable that adding significant comorbidities would decrease the likelihood of responding to this, or any other, pharmacological treatment. Yet, a proper assessment of this point would require a quite larger sample size.

The following adverse events were reported throughout the study by the 40 subjects participating in the IC-PGE1–sildenafil 50 mg combination trial: facial flushing, seven patients (18%); headache, four patients (10%); and blue visual aura, one patient. All these disturbances were mild and transient. In fact, they occurred essentially when the subjects had been taking the sildenafil alone 100 mg dose in the selection phase. No noticeable side effects were specifically associated with the combined IC-PGE1–sildenafil 50 mg treatment phase. Neither painful sensation nor prolonged erections following the in-office IC-PGE1 injections were reported.



The present data show that programmed IC-PGE1 injections can increase substantially the erectile response to 50 mg sildenafil taken on demand to a clinically relevant level. This was found in nearly two-thirds of the patients previously not responding to sildenafil monotherapy with either the 50 mg or the 100 mg doses. It is worth noting that such an effect was obtained with a fairly conservative treatment schedule, that is, two or four bi-weekly IC-PGE1 injections and 50 mg sildenafil doses taken on demand. It is conceivable that using a more intensive treatment, that is, more frequent IC-PGE1 injections and/or a higher sildenafil dose (100 mg), the effects of the combination therapy could have been even larger.

Combination therapies with various drugs acting through different mechanisms have long been used in the treatment of ED, especially in the form of IC injections.15 Recently, the combined use of IC12 or intraurethral10, 11 PGE1 and the first effective oral PDE-5 inhibitor sildenafil taken together at the time of intercourse has been reported successful in many patients not responding to either single drug. The present data indicate that such a combination can also be effective when the drugs are administered quite separately. From a practical viewpoint, this new treatment option could help to circumvent some of the difficulties frequently reported by intracavernous injection users such as the lack of spontaneity and naturalness of the sexual act that the procedure involves and its rejection by the partner.5, 7, 8, 9 It would also reduce the risk of priapism, a complication occasionally observed in patients taking IC injections and oral PDE-5 inhibitors within a short interval.16

Regarding the possible biochemical mechanisms underlying this clinical effect, it is worth noting that significant crosstalk phenomena between the signal transduction pathways for cAMP (stimulated by PGE1) and cGMP (enhanced by PDE-5 inhibitors) are being increasingly documented both in human and animal cavernous tissue (reviewed in Steers15). Particularly relevant is the finding in the rat model that repeated IC-PGE1 injections upregulate NO synthase expression (nNOS and eNOS isoforms) and that translates into a substantial increase in NO release and intracavernous pressure following cavernous nerve stimulation.13 The present data are suggestive of similar molecular mechanisms contributing to the improved erectile responses shown by our 'responder' subjects (2/3 of our sample) following the programmed IC-PGE1–sildenafil on demand combination therapy. However, that needs confirmation by studies on human cavernous tissue. It is also tempting to speculate that similar phenomena could account for some intriguing findings such as the recovery of spontaneous erections reported by some ED patients after receiving several IC injections of PGE1 alone or associated with other erectogenic drugs,4, 6, 7, 8, 9, 17 or the choice by some experienced patients to keep using oral PDE5 inhibitors and IC therapy in alternance.18, 19, 20

In conclusion, we have demonstrated that even using a quite conservative treatment schedule, the efficacy of sildenafil, and likely of other PDE-5 inhibitors, can be enhanced to a clinically relevant level in a significant number of previously unresponsive patients by treating them with some programmed IC-PGE1 injections. However, there are still several issues that should be addressed. They include defining the optimal doses of IC-PGE1 injections as well as sildenafil or other PDE-5 inhibitors for this combined treatment. It is also important to optimize the IC-PGE1 treatment schedule, that is, how long can the injections be spaced? How many are needed for best results? The specificity of this phenomenon is also worth studying, that is, is it specific for IC-PGE1 or could it be achieved with other erectogenic treatments, intracavernous, or otherwise? As pointed above, the underlying molecular mechanisms also need to be assessed in the human penis. All these questions remain open to further inquiry.



  1. Fazio L & Brock G. Erectile dysfunction: management update. Canadian Med Assoc J 2004; 170: 1429−1437. | Article |
  2. Porst H. Current perspectives on intracavernosal pharmacotherapy for erectile dysfunction. Int J Impot Res 2000; 12 Suppl 4: S91−S100. | Article | PubMed |
  3. Porst H et al.. Intracavernous alprostadil Alfadex—an effective and well tolerated treatment for erectile dysfunction. Results of a long-term European study. Int J Impot Res 1998; 10: 225−231. | PubMed | ChemPort |
  4. Virag R et al.. Intracavernous self-injection of vasoactive drugs in the treatment of impotence: 8-year experience with 615 cases. J Urol 1991; 145: 287−293. | PubMed | ChemPort |
  5. Sundaram CP et al.. Long-term follow-up of patients receiving injection therapy for erectile dysfunction. Urology 1997; 49: 932−935. | Article | PubMed | ChemPort |
  6. Casabé A et al.. Drop-out reasons and complications in self-injection therapy with a triple vasoactive drug mixture in sexual erectile dysfunction. Int J Impot Res 1998; 10: 5−9. | PubMed |
  7. Rowland DL, Boedhoe HSM, Dohle G & Slob AK. Intracavernosal self-injection therapy in men with erectile dysfunction: satisfaction and attrition in 110 patients. Int J Impot Res 1999; 11: 145−151. | Article | PubMed | ChemPort |
  8. Purvis K, Egdetveit I & Christiansen E. Intracavernosal therapy for erectile failure—impact of treatment and reasons for drop-out and dissatisfaction. Int J Impot Res 1999; 11: 287−299. | Article | PubMed | ChemPort |
  9. Mulhall JP et al.. The causes of patient dropout from penile self-injection therapy for impotence. J Urol 1999; 162: 1291−1294. | Article | PubMed | ChemPort |
  10. Mydlo JH, Volpe MA & Macchia RJ. Results from different patient populations using combined therapy with alprostadil and sildenafil: predictors of satisfaction. Br J Urol Int 2000; 86: 469−473. | ChemPort |
  11. Nehra A, Blute ML, Barrett DM & Moreland RB. Rationale for combination therapy of intraurethral prostaglandin E1 and sildenafil in the salvage of erectile dysfunction patients desiring noninvasive therapy. Int J Impot Res 2002; 14 Suppl 1: S38−S42. | Article | PubMed |
  12. McMahon CG, Samali R & Johnson H. Treatment of intracorporal injection non-response with sildenafil alone or in combination with triple agent intracorporeal injection therapy. J Urol 1999; 162: 1992−1997. | PubMed | ChemPort |
  13. Escrig A, Marin R & Mas M. Repeated PGE1 treatment enhances nitric oxide and erection responses to nerve stimulation in the rat penis by upregulating constitutive NOS isoforms. J Urol 1999; 162: 2205−2210. | PubMed | ChemPort |
  14. Cappelleri JC et al.. Diagnostic evaluation of the erectile function domain of the International Index of Erectile Function. Urology 1999; 54: 346−351. | Article | PubMed | ChemPort |
  15. Steers WD. Viability and safety of combination drug therapies for erectile dysfunction. J Urol 2003; 170: S20−S23. | Article | PubMed |
  16. McMahon CG. Priapism associated with concurrent use of phosphodiesterase inhibitor drugs and intracavernous injection therapy. Int J Impot Res 2003; 15: 383−384. | Article | PubMed | ChemPort |
  17. Montorsi F et al.. Recovery of spontaneous erectile function after nerve-sparing radical retropubic prostatectomy with and without early intracavernous injections of alprostadil: results of a prospective, randomized trial. J Urol 1997; 158: 1408−1410. | Article | PubMed | ChemPort |
  18. McMahon CG, Samali R & Johnson H. Efficacy, safety and patient acceptance of sildenafil citrate as treatment for erectile dysfunction. J Urol 2000; 164: 1192−1196. | Article | PubMed | ChemPort |
  19. Hatzichristou DG et al.. Sildenafil versus intracavernous injection therapy: efficacy and preference in patients on intracavernous injection for more than 1 year. J Urol 2000; 164: 1197−1200. | Article | PubMed | ChemPort |
  20. Buvat J, Lemaire A & Ratajcyk J. Acceptance, efficacy and preference of sildenafil in patients on long-term auto-intracavernosal therapy: a study with follow-up at one year. Int J Impot Res 2002; 14: 483−486. | Article | PubMed | ChemPort |


This work was supported by a Medical Grant from Pfizer SA, Madrid, Spain and Grant PM 99/0144 from the Spanish Ministry of Science and Technology.



These links to content published by NPG are automatically generated


L-DOPA, dyskinesia and striatal plasticity

Nature Neuroscience News and Views (01 May 2003)

Extra navigation