Asian J Androl 2006; 8 (2): 219–224
.Clinical Experience .
Long-term treatment with intracavernosal injections in
diabetic men with erectile dysfunction
P. Perimenis, A. Konstantinopoulos, P. P. Perimeni, K. Gyftopoulos, G. Kartsanis, E. Liatsikos, A. Athanasopoulos
Department of Urology, University Hospital, 26500 Patras, Greece
To assess the behavior of patients with diabetes mellitus (DM) and erectile dysfunction (ED) during 10 con-
secutive years of treatment with self-injection of vasoactive drugs. Methods:
Thirty-eight diabetic men, including 12
with type I and 26 with type II diabetes, were followed up regularly for 10 years after they began self-injecting for
severe ED. Real time rigidity assessment was used for the objective determination of the initial dosage and then doses
were regulated in order to introduce an erection suitable for penetration and maintenance of erection for approximately
30 min. Patients were followed up every two months, and doses were increased only when the treatment response
was not satisfactory. Results:
The number of injections used per year by the patients was reduced each year (mean
numbers: 50 in the first year and 22.5 in the 10th) and treatment shifted towards stronger therapeutic modalities
(mixtures of vasoactive drugs instead of prostaglandin E1 alone). Type I diabetic men were standardized to a level of
treatment as early as 5 years after the initiation of treatment. That level was finally reached by type II patients after
another 4-5 years. Conclusion:
Treatment with self-injections of vasoactive drugs in diabetic men with severe ED is
a safe and effective alternative in the long term. Diabetic men of both types show the same preferences in quality and
quantity of treatment after 10 years. The key point for maintenance in treatment is the adjustment of the therapeutic
method and dosage to optimal levels for satisfactory erections. (Asian J Androl 2006 Mar; 8: 219–224)
: diabetes mellitus; erectile dysfunction; impotence; intracavernosal injections; prostaglandin E1; papaverine
be offered to diabetic men, sexual problems, despite theirrelevance, are still seldom investigated by general practitio-
The incidence of sexual dysfunction in men with dia-
ners and specialists . However, the etiology of diabetic
betes mellitus (DM) is approaching 50% and, as diabetes
ED has been thoroughly investigated [3, 4] and the thera-
is a problem that is increasing at an alarming rate, dia-
peutic management became satisfactory with the use of
betic men already made up one-quarter of those seeking
a wide spectrum of treatments [4–6].
advice for erectile dysfunction (ED) . Although a suc-
The natural history of ED indicates the importance
cessful impotence assessment and treatment service may
of age. In diabetic individuals, ED is more progressiveand usually irreversible. Its etiology is multifactorial, in-cluding neuropathy, vascular disease, metabolic control,
Corresponence to: Dr Petros Perimenis, Department of Urology,
nutrition, endocrine disorders, psychogenic factors, and
University Hospital of Patras, 26500 Rio, Patras, Greece.
Tel: +30-61-999-397, Fax: +30-61-993-981
drugs coadministered for comorbidities, such as
antihypertensives. The role of autonomic neuropathy
Received 2005-02-03 Accepted 2005-06-23
has been emphasized, and is considered the major factor.
2006, Asian Journal of Andrology, Shanghai Institute of Materia Medica, Chinese Academy of Sciences. All rights reserved.
Erectile dysfunction and diabetes
Zhu et al
.  reported a 50% rate of abnormal pudental
investigation with Doppler ultrasonography and tests of
evoked potentials, and recently, Xu et al
.  reported on
vasoactive drugs. They were all proposed to start on self-
the decrease of nitric oxide (NO) synthase content in
injections but six refused treatment. Of the 47 men who
corpus cavernosum of diabetic rats. NO synthase is the
started on self-injections, nine stopped therapy gradually
only enzyme for the synthesis of NO, the neurotransmit-
for several reasons, and 38 completed 10 years of treatment.
ter mediating smooth muscle relaxation and introducing
Twelve (31.6%) of them had type I diabetes and 26
erection. Although ED is a marker for the development
(68.4%) had type II diabetes. The process of enrolment
of generalized vascular disease, diabetic arteriopathy may
for the studied patients is depicted in detail in Figure 1. The
affect blood supply, contributing to the neurogenic fac-
patients had acceptable metabolic control. Glycosylated
tor , but does not appear to result often in entire penile
hemoglobin levels ranged between 6% and 8%.
artery occlusion. The psychological component has alsobeen emphasized. The results of the largest study evalu-
2.2 Treatment for ED and follow-up
ating quality of life in diabetic patients with ED provide
Each patient was initially examined in privacy under
clear evidence that ED is associated with higher levels of
discrete conditions. The response to intracavernosal in-
diabetes-specific health distress and worse psychologi-
jections was evaluated in real time by Rigiscan (Dacomed,
cal adaptation to diabetes, which are, in turn, related to
Minneapolis, MN, USA). The device was applied for
worse metabolic control. Erectile problems are also as-
30 min after the injection with simultaneous audiovisual
sociated with a dramatic increase in the prevalence of
stimulation. The aims of this test were to assess tumes-
severe depressive symptoms and lower scores in mental
cence and rigidity and to determine the proper drug and
dosage for the achievement of an erection for up to
Diabetic men are more likely to achieve a satisfac-
30 min. A response was considered objectively satisfac-
tory response to intracavernosal injections than those with
tory if there was a 30-mm or more increase in circum-
other types of ED . Moreover, diabetic patients ac-
ference and a rigidity of 70% or more, both for at least
cept self-injecting more easily and comply better with
10 min. To determine the response to vasoactive drugs
treatment for ED compared to non-diabetics .
and the therapeutic dose, all patients were initially in-
However, in general, the frequency of non-compliance
jected with 5–10 µg PGE1 and the non-responders were
with self-injecting is high, approaching 50%, and is prob-
given 15–20 µg PGE1 after 1 week. A few patients,
ably the most common event in clinical practice . In
who did not respond to the higher dose of PGE1 (20 µg),
this study we assessed the main characteristics of long-
needed a further mixture of PGE1 and papaverine (PAP).
term treatment with self-injection of vasoactive drugs in
The drugs were prepared and given in the clinic at
diabetic men with ED.
follow-up, and dosages were regulated to provide an erec-
2 Materials and methods
2.1 Study recruitment
Only diabetic patients with ED who had completed
10 years of treatment with self-injection of vasoactivedrugs were included in this study. In 1993 and 1994, 78men with DM and ED were referred or presented to oursexual dysfunction clinic. A detailed history was ob-tained from all patients, who filled out a questionnaireabout their sexual activities. Most of them underwentlaboratory tests and all had a simple test of intracavernosalinjection with 10 mg prostaglandin E1 (PGE1). Of them,25 men achieved satisfactory erections responding toconservative treatment and psychosexual counsellingduring the assessment period and were not managed
Figure 1. Enrolment process of patients in our study of diabetic
further. The remaining 53 patients underwent a detailed
men with erectile dysfunction (ED).
Asian J Androl 2006; 8 (2): 219–224
tion suitable for penetration and maintenance of approxi-
duration, this group represents a typical sample of men
mately 30 min. The patients were asked to complete a
with ED . Seventeen men (32%) had abnormal pe-
consent form because the drugs used were not licensed
nile Doppler assessments (maximum penile systolic ve-
for intracavernosal treatment, and approval was required
locity < 25 cm/s). Overall, during initial real-time Rigiscan
for possible scientific publication of the data. The pa-
evaluation, 19 men responded to low and 12 to high PGE1
tients were taught how to self-inject and were advised to
doses, whereas four men responded to low and three to
use injections not more than once per week, alternating
high MIX doses. Treatment with self-injections was safe
between the two sides of the penis. They were also asked
and well tolerated. Five patients noticed fibrosis in the cor-
to record the results of their attempts for intercourse, to
pora without bend. Episodes of prolonged erections or
bring back the unused injections (in order to record the
priapism were not recorded during the treatment period.
frequency of sexual activity and to verify the number of
The majority of patients responded initially to PGE1,
recorded attempts), and to report any complications
especially to low doses, but with time they needed in-
immediately. All data of the patients' follow-up were pro-
creasing doses of PGE1, and later, increasing doses of
spectively entered into the departmental database.
mixtures of PGE1 and PAP to achieve a satisfactory
Drug doses were increased only when the treatment
erection. After 7 years of treatment, none was treated
response was not satisfactory. The doses of PGE1 were
with low doses of PGE1. After 5 years the majority
increased by 5–10 µg, and PAP by 8–16 mg. The use of
needed a mixture of the vasoactive drugs, and particu-
20 µg PGE1 without satisfactory response was the cri-
larly after 7 years the majority needed high doses of MIX.
terion for switching to a drug mixture. The mixtures
During the first year of treatment, 31 patients used pros-
were combinations of PGE1 20 µg and various doses of
taglandins only and seven used mixture treatments. In
PAP. For practical reasons, treatment with self-injec-
the 10th year, however, only two patients used prostag-
tions was classified as low PGE1, high PGE1, low MIX
landins and the majority, 36, used mixtures.
and high MIX. This classification is shown in detail in
Without taking into consideration the type of DM,
Table 1. Patients were followed up every 2 months to
there was a statistically significant (aP
< 0.001) turn in
reassess their erectile function.
the patients towards stronger treatments (mixtures) af-ter 10 years. The changes in treatment in the long term
are depicted analytically in Figure 2. In the 10th year of
The McNemar testa, Pearson's χ2-testb, the Wilcoxon
treatment, the type of diabetes was not related to the
signed-ranks test for paired observationsc, and the Mann–
treatment used, as there was no statistically significant
-testd were applied for statistical evaluation of
relation between the two variables (bP
= 0.324). All DM
the data where appropriate, using a designated statistical
type I patients (12/12) used mixtures, as did almost all
package (SPSS 12.0 for Windows, SPSS Inc., Chicago,
DM type II patients (24/26). But in the first year, the
Illinois, USA). Statistical significance was set at P
type of diabetes was significantly related to the kind oftreatment: patients with DM type II used only prostag-
landin and patients with DM type I used prostaglandinand mixtures almost equally, 5 of 12 and 7 of 12, re-
Patients' demographic characteristics at baseline are
< 0.001). This relationship between the
shown in Table 2. According to age and dysfunction
type of diabetes and treatment began to weaken as earlyas the sixth year and lost its significance in the ninth year(bP
= 0.151). By definition, bP
values estimate the sta-
Table 1. Classification of treatment with self-injections of vasoac-tive drugs in diabetic men with erectile dysfunction (ED). PAP,
tistical significance of the difference between the ob-
papaverine; PGE1, prostaglandin E1; MIX, combinations of 20 µgPGE1 and various doses of papaverine.
Table 2. Patient demographics at the beginning of our study of men
with diabetes mellitus (DM) and erectile dysfunction (ED).
20 µg PGE1 + 8–16 mg PAP
DM duration (year)
20 µg PGE1 + > 16 mg PAP
ED duration (year)
Erectile dysfunction and diabetes
Table 3. Relationship between patients' type of diabetes mellitus (DM) and the treatment used for erectile dysfunction. *bP
= 0.151. PGE1, prostaglandin E1; MIX, combinations of 20 µg PGE1 and various doses of papaverine; Count, number of patients;Expected count, number of patients according to the null hypothesis.
First year* (ninth year**)
Figure 2. Treatment methods over 10 years applied to the wholestudy group of 38 diabetic men with erectile dysfunction (ED).
prostaglandin E1 (PGE1) low, 5–10 µg PGE1; PGE1 high, 15–20 µg PGE1; MIX low, 20 µg PGE1 + 8–16 mg papaverine (PAP);MIX high, 20 µg PGE1 + > 16 mg PAP.
served and the expected counts of patients in each treat-ment method (Table 3).
Figure 3. Mean number (±SD) of self-injections of vasoactive drugs
The mean number of injections required by the pa-
in the whole study group of 38 diabetic patients over 10 years.
tients as a whole was 50 in the first year and 22.5 in the10th year. The number of injections, regardless of thetype of diabetes, was significantly reduced year by year
< 0.001), with a temporary weakening of significancebetween the fourth and fifth years (cP
= 0.035). The num-
The treatment of severe ED with self-injection of
ber of injections per year is depicted in Figure 3. Both
vasoactive drugs in diabetic patients has been a very com-
groups of diabetic patients significantly reduced the num-
mon alternative. The mixtures of vasoactive drugs in
ber of injections (cP
= 0.001 for each of the groups).
particular, which use different mechanisms of action and
Between the second and fourth years, type I diabetic men
exert pharmacological synergism, are an effective and safe
used fewer injections than the type II patients (dP
treatment for severe diabetic ED. Self-injection is also a
but after the fifth year the type II patients began to close
safe treatment, especially in terms of concerns about the
the gap, standardizing to 22.42 ± 2.67 (mean ± SD) injec-
perceived risk of priapism. It has been reported that pri-
tions at the 10th year. The mean number of injections
apism never occurred during the long-term treatment phase
used per year by both groups is depicted in Figure 4.
of experienced patients . Although the majority of pa-
Asian J Androl 2006; 8 (2): 219–224
patient, to solve practical problems regarding the injec-tions and to encourage patients and their partners to con-tinue and comply with the treatment. The adjustment ofdosage to appropriate levels is also very important, par-ticularly for the patient treated with injections. The pa-tient must be reassured that the treatment works, and to beconfident that when an increase in the dose is needed, it isnecessary to go along with his physician's advice. Menwith type I (insulin-dependent) DM are more familiar withself-injecting on a daily basis. On the other hand, menwith type II DM, who end up using injections, are gen-erally patients who have used oral treatment in the pastunsuccessfully and injections seem the last option leftbefore penile implantation.
The erectile tissue and penile musculature is not
modified negatively or positively by intracavernosal in-jections , but the biochemical and ultrastructuralchanges by DM, as well as aging, affect it in a negativeway [17, 18]. These factors could play a major role inthe observed increased need for stronger remedies (higher
Figure 4. Mean number of self-injections of vasoactive drugs per
doses of PGE1 or more effective mixtures of PGE1 and
year for each type of diabetes mellitus (DM).
PAP). It is well established that the combination of lowor reasonable doses of vasoactive drugs are more effec-
tients with ED strongly prefer oral therapeutic compounds,
tive than high doses of PGE1 to achieve an erection suit-
which represent the first-line treatment because of the po-
able for penetration, with a lower incidence of pain .
tential benefits and lack of invasiveness , diabetic men
In our study, in the 10th year of treatment, there
who have started self-injecting are not likely to switch suc-
was no difference between the two groups of diabetic
cessfully to oral treatment . Therefore self-injection
patients in the number of injections or the kind of treat-
should be considered at this time a long-term therapeutic
ment they used. Insulin-treated men proceeded earlier
option and these patients should be advised accordingly.
than the others towards the final standardization of their
In this study of men who started therapy before, but
treatment. Because they were more familiar with the
continued within, the Phosphodiesterase 5 inhibitors era,
possibility of ED, they may compromise quickly with
we attempted to assess the behaviour of patients with
lower expectations for sexual life, so they find their
either type I or type II DM, towards continuing treat-
quantitave and qualitative balance earlier. DM type II
ment with intracavernosal self-injections of vasoactive
patients continued for longer to make more effort for
drugs. The group presented here are the non-respond-
successful intercourse, which actually meant more in-
ers to oral treatment after the launch of PDE5 inhibitors.
jections per year, but ended up reaching the same levels
Thus continuing injections, they complied with treatment
of effort as the DM type I group. The obvious decrease
because of satisfaction with the response, the quality of
of injection frequency per year for both groups may also
erections and the care undertaken for the treatment
show tiredness by time of having to self-inject and an
attempt to minimize side-effects.
To keep a patient satisfied in the long term with a
In conclusion, the self-injection of vasoactive drugs
semi-invasive treatment, such as the penile injection, is
continues to be, in the long term, a highly effective and
not easy. This issue may be mainly responsible for the
safe treatment for ED in men with DM. The key point
high rate of non-compliance with self-injecting in the
for maintaining the treatment is the adjustment of the thera-
general population of men with ED. Thus, we conclude
peutic method and dosage to optimal levels for satisfactory
that it is very important for the physician to have a
erections. For this reason, systematic follow-up of these
constant, personal, face-to-face communication with the
cases is of the utmost importance. Diabetic men de-
Erectile dysfunction and diabetes
crease the number of self-injections over time, set realis-
1996; 13: 84–9.
tic expectations and create a baseline of satisfactory sexual
10 Perimenis P, Gyftopoulos K, Athanasopoulos A, Barbalias G.
life with aging.
Diabetic impotence treated by intracavernosal injections: hightreatment compliance and increasing dosage of vaso-activedrugs. Eur Urol 2001; 40: 398–403.
11 Althof SE, Turner LA, Levine SB, Risen C, Kursh E, Bodner
D, et al
. Why do so many people drop out from auto-injectiontherapy for impotence? J Sex Marital Ther 1989; 15: 121–9.
Guay AT. Sexual dysfunction in the diabetic patient. Int J
12 Markou S, Perimenis P, Gyftopoulos K, Athanasopoulos A,
Impot Res 2001; 13: S47–50.
Barbalias G. Vardenafil (Levitra) for erectile dysfunction: a
De Berardis G, Franciosi M, Belfiglio M, Di Nardo B,
systematic review and meta-analysis of clinical trial reports.
Greenfield S, Kaplan SH, et al.
Erectile dysfunction and qua-
Int J Impot Res 2004; 16: 470–8.
lity of life in type 2 diabetic patients. A serious problem too
13 Perimenis P, Athanasopoulos A, Geramoutsos I, Barbalias G.
often overlooked. Diabetes Care 2002; 25: 284–91.
The incidence of pharmacologically induced priapism in the
Bemelmans BLH, Meuleman EJH, Doesburg WH, Notermans
diagnostic and therapeutic management of 685 men with erec-
SLH, Debruyne FMJ. Erectile dysfunction in diabetic men:
tile dysfunction. Urol Int 2001; 66: 22–6.
the neurological factor revisited. J Urol 1994; 151: 884–9.
14 Aversa A, Fabbri A. New oral agents for erectile dysfunction:
Veves A, Webster L, Chen TF, Payne S, Boulton AJM.
what is changing in our practice? Asian J Androl 2001; 3: 175–
Aetiopathogenesis and management of impotence in diabetic
males: four years experience from a combined clinic. Diabetic
15 Perimenis P, Markou S, Gyftopoulos K, Athanasopoulos A,
Med 1995; 12: 77–82
Giannitsas K, Barbalias G. Switching from long-term treat-
Heaton JP, Lording D, Liu SN, Litonjua AD, Guangwei L,
ment with self-injections to oral sildenafil in diabetic patients
Kim SC, et al.
Intracavernosal alprostadil is effective for the
with severe erectile dysfunction. Eur Urol 2002; 41: 387–91.
treatment of erectile dysfunction in diabetic men. Int J Impot
16 Wespes E, Sattar AA, Noel JC, Schulman CC. Does prostag-
Res 2001; 13: 317–21.
landin E1 therapy modify the intracavernous musculature? J
Vickers MA, Satyanarayana R. Phosphodiesterase type 5
Urol 2000; 163: 464–6.
inhibitors for the treatment of erectile dysfunction in patients
17 Shi JP, Zhao YM, Song YT. Effect of aging on expression of
with diabetes mellitus. Int J Impot Res 2002; 14: 466–71.
nitric oxide synthase I and activity of nitric oxide synthase in
Zhu GY, Shen Y. Application of pudental evoked potentials
rat penis. Asian J Androl 2003; 5: 117–20.
in diagnosis of erectile dysfunction. Asian J Androl 1999; 1:
18 Lu YL, Shen ZJ, Wang H, Chen SW, Zhou XL, Chen ZD.
Ultrastructural changes of penile tunica albuginea in diabetic
Xu ZS, Fu Q, Zhao ST, Liu HN. Effect of diabetes and insulin
rats. Asian J Androl 2004; 6: 365–8.
treatment on nitric oxide synthase content in rat corpus
19 Bechara A, Casabe A, Cheliz G, Romano S, Fredotovich N.
cavernosum. Asian J Androl 2001; 3: 139–42.
Prostaglandin E1 versus mixture of prostaglandin E1, papav-
Bancroft J, Gutierrez P. Erectile dysfunction in men with and
erine and phentolamine in nonresponders to high papaverine
without diabetes mellitus: a comparative study. Diabetic Med
plus phentolamine doses. J Urol 1996; 155: 913–4.
MASHAV - Israel's Agency for International Development Cooperation The Arava Institute for Environmental Studies invite professionals to participate in the STATE OF ISRAEL International Course: Environmental Management of Nature Parks and Reserves Israel's Agency for International
CANAZEI, VAL DI FASSA, TRENTINO La Val di Fassa si sviluppa per una ventina di chilometri nel cuore delle Dolomiti ed è un autentico concentrato di meraviglie naturali. Un tripudio di vette che da Moena a Canazei lasciano letteralmente senza fi ato. Roda di Vael, Catinaccio, Sassolungo, Sella, Marmolada e Monzoni sono solo alcune delle cime più famose della Val di Fassa. Una