`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`The New England
`Journal
` Medicine
`of
`
`© Copyright, 1998, by the Massachusetts Medical Societ y
`
`VOLUME 338
`
`M
`
`A Y
`
` 14, 1998
`
`NUMBER 20
`
`ORAL SILDENAFIL IN THE TREATMENT OF ERECTILE DYSFUNCTION
`
`, P
` C. R
`, M.D., R
`-N
` P
`, M.D., H
` F. L
`, M.D., T
` G
`I
`OSEN
`AYMOND
`ATHAN
`ADMA
`ARIN
`UE
`OM
`OLDSTEIN
`RWIN
`W
` D. S
`, M.D.,
` P
` A. W
`, M.D.,
`
` S
` S
` G
`*
`ROUP
`IERRE
`AND
`TEERS
`ILLIAM
`ICKER
`FOR
`THE
`ILDENAFIL
`TUDY
`
`.D.,
`
`H
`
`A
`BSTRACT
`Background
`Sildenafil is a potent inhibitor of cy-
`clic guanosine monophosphate in the corpus caver-
`nosum and therefore increases the penile response
`to sexual stimulation. We evaluated the efficacy and
`safety of sildenafil, administered as needed in two
`sequential double-blind studies of men with erectile
`dysfunction of organic, psychogenic, or mixed causes.
`Methods
`In a 24-week dose–response study, 532
`men were treated with oral sildenafil (25, 50, or 100
`mg) or placebo. In a 12-week, flexible dose-escalation
`study, 329 different men were treated with sildenafil
`or placebo, with dose escalation to 100 mg based on
`efficacy and tolerance. After this dose-escalation
`study, 225 of the 329 men entered a 32-week, open-
`label extension study. We assessed efficacy accord-
`ing to the International Index of Erectile Function, a
`patient log, and a global-efficacy question.
`Results
`In the dose–response study, increasing
`doses of sildenafil were associated with improved
`erectile function (P values for increases in scores for
`questions about achieving and maintaining erec-
`⬍
`tions were
`0.001). For the men receiving 100 mg of
`sildenafil, the mean score for the question about
`achieving erections was 100 percent higher after
`treatment than at base line (4.0 vs. 2.0 of a possible
`score of 5). In the last four weeks of treatment in the
`dose-escalation study, 69 percent of all attempts at
`sexual intercourse were successful for the men re-
`ceiving sildenafil, as compared with 22 percent for
`⬍
`0.001). The mean num-
`those receiving placebo (P
`bers of successful attempts per month were 5.9 for
`the men receiving sildenafil and 1.5 for those receiv-
`⬍
`0.001). Headache, flushing, and dys-
`ing placebo (P
`pepsia were the most common adverse effects in the
`dose-escalation study, occurring in 6 percent to 18
`percent of the men. Ninety-two percent of the men
`completed the 32-week extension study.
`Conclusions
`Oral sildenafil is an effective, well-
`tolerated treatment for men with erectile dysfunc-
`tion. (N Engl J Med 1998;338:1397-404.)
`©1998, Massachusetts Medical Society.
`
`E
`
`RECTILE dysfunction, the persistent in-
`ability to achieve or maintain an erection
`sufficient for satisfactory sexual perform-
`ance, is estimated to affect up to 30 million
`men in the United States.
` The disorder is age-asso-
`1
`ciated,
` with estimated prevalence rates of 39 per-
`1-3
`cent among men 40 years old and 67 percent among
`those 70 years old.
` The available treatments include
`2
`vacuum-constriction devices; intracavernosal injec-
`tions of vasoactive agents, including alprostadil (pros-
`)
`; transurethral delivery of alprostadil
`;
`taglandin E
`4
`5
`1
`implantation of penile prostheses; and venous or ar-
`terial surgery. No effective oral therapy for erectile
`dysfunction is currently available.
`6
`Normal penile erection depends on the relaxation
`of smooth muscles in the corpora cavernosa. In re-
`sponse to sexual stimuli, cavernous nerves and endo-
`thelial cells release nitric oxide, which stimulates the
`formation of cyclic guanosine monophosphate (GMP)
`by guanylate cyclase.
` The mechanism by which cy-
`7-9
`clic GMP stimulates relaxation of the smooth muscles
`remains to be elucidated. Sildenafil is a selective inhib-
`itor of cyclic-GMP–specific phosphodiesterase type 5,
`the predominant isozyme metabolizing cyclic GMP
` By selectively inhibiting
`in the corpus cavernosum.
`10
`cyclic-GMP catabolism in cavernosal smooth-muscle
` sildenafil would be expected to restore the nat-
`cells,
`11
`ural erectile response to sexual stimulation but not
`
`From the Department of Urology, Boston University Medical Center,
`Boston (I.G.); the Department of Urology, University of California, San
`Francisco (T.F.L.); the Department of Urology, University of Southern
`California, Los Angeles, and the Male Clinic, Santa Monica, Calif.
`(H.P.-N.); the Department of Psychiatry, University of Medicine and Den-
`tistry of New Jersey, Robert Wood Johnson Medical School, Piscataway,
`N.J. (R.C.R.); the Department of Urology, University of Virginia, Char-
`lottesville (W.D.S.); and Pfizer Central Research, Groton, Conn. (P.A.W.).
`Address reprint requests to Dr. Goldstein at the Department of Urology,
`Boston University Medical Center, 720 Harrison Ave., P (606), Boston,
`MA 02118.
`*The members of the Sildenafil Study Group are listed in the Appendix.
`
`Volume 338 Number 20
`
`ⴢ
`
`1397
`
`ATI 1027-0001
`
`ATI v. ICOS
`IPR2018-01183
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`The New England Journal of Medicine
`
`cause erections in the absence of such stimulation.
`Sildenafil is rapidly absorbed, with maximal plasma
`concentrations occurring within one hour after oral
`administration and a mean terminal half-life of three
` In a placebo-controlled pilot study
`to five hours.
`10
`of 12 men, sildenafil significantly improved the erec-
`tile response during visual sexual stimulation.
` We
`10,12
`therefore undertook two studies to evaluate in a home
`setting the efficacy and safety of sildenafil in men
`with erectile dysfunction.
`
`METHODS
`
`In two sequential studies, we studied a total of 861 men 18
`years of age or older with a clinical diagnosis of erectile dysfunc-
`tion (as defined previously
`) of six months’ duration or longer at
`1
`37 centers in the United States. Each man had to be in a stable
`relationship with a female partner that had begun at least six
`months earlier. The cause of erectile dysfunction was determined
`from the medical history, physical examination, and other diag-
`nostic procedures, including a test involving the intracavernosal
`injection of a vasoactive drug (done in 31 percent of the men), a
`RigiScan test of nocturnal penile tumescence (26 percent), penile
`duplex ultrasonography (21 percent), and endocrine testing (21
`percent). On the basis of these evaluations, the men were classi-
`fied as having organic, psychogenic, or mixed erectile dysfunc-
`tion. Of the 861 men studied, 605 (70 percent) were judged to
`have organic erectile dysfunction, 99 (11 percent) to have psy-
`chogenic erectile dysfunction, and 157 (18 percent) to have
`mixed erectile dysfunction. Men were excluded if they had penile
`anatomical defects, a primary diagnosis of another sexual disorder
`(e.g., premature ejaculation), spinal cord injury, any major psychi-
`atric disorder not well controlled with treatment, poorly con-
`trolled diabetes mellitus, active peptic ulcer disease, a history of
`alcohol or substance abuse, major hematologic, renal, or hepatic
`abnormalities, or a recent (within the previous six months) stroke
`or myocardial infarction or if they were receiving nitrate therapy.
`Other erectile-dysfunction therapies were discontinued at the
`time of screening (four weeks before the subjects received the
`study medication). Sildenafil (Viagra) and an identical-looking
`placebo were supplied by Pfizer. The men were instructed to take
`a dose approximately one hour before planned sexual activity but
`not more than once daily. The protocols were approved by the
`institutional review board at each center, and all the men gave
`written informed consent.
`We assessed efficacy by using the responses to question 3 (fre-
`quency of penetration) and question 4 (maintenance of erections
`after penetration) of the 15-question International Index of
`Erectile Function, a validated, multidimensional, self-adminis-
`tered questionnaire used for the clinical assessment of erectile
`dysfunction and treatment outcomes in clinical studies.
` The re-
`11
`sponses to these two questions pertaining to the ability to
`achieve and maintain an erection sufficient for sexual intercourse,
`as described in the definition of erectile dysfunction,
` were rated
`1
`on a scale of 1 (almost never or never) to 5 (almost always or
`always). A score of 0 indicated no attempt at sexual intercourse.
`The mean score for each of the two questions was 4.3 for 109
`normal men, 31 to 86 years old, with an age distribution similar
`to that of the men with erectile dysfunction (unpublished data).
`Efficacy was also assessed on the basis of the scores for the five
`separate response domains of male sexual function of the Inter-
`national Index
`: erectile function (questions 1 through 5 and
`13
`15; possible total score, 1 to 30); orgasmic function (questions
`9 and 10; possible total score, 0 to 10); sexual desire (questions
`11 and 12; possible total score, 2 to 10); intercourse satisfaction
`(questions 6, 7, and 8; possible total score, 0 to 15); and overall
`satisfaction (questions 13 and 14; possible total score, 2 to 10).
`The domain scores were computed by adding the scores for the
`
`1398
`
`ⴢ
`
`May 14, 1998
`
`individual questions in each domain. Other means of assessing
`efficacy were an event log, in which we asked the men to record
`the date and dose of medication taken, the presence of sexual
`stimulation, the hardness of erections (graded on a four-point
`scale), and whether sexual intercourse was successful, and a glob-
`al-efficacy question (“Did the treatment improve your erec-
`tions?”), with a response of yes or no. The end points of the In-
`ternational Index quantified the magnitude of the response, and
`the global-efficacy question and the event log provided qualita-
`tive assessments of efficacy. Physical examinations and standard
`blood-chemistry and hematologic laboratory tests were performed
`throughout the studies. Adverse effects were recorded by the in-
`vestigators.
`
`Study of Dose–Response, Efficacy, and Safety
`In this double-blind, placebo-controlled, fixed-dose study, 532
`men were randomly assigned to take placebo or 25, 50, or 100
`mg of sildenafil (approximately one hour before planned sexual
`activity but not more than once daily) for 24 weeks. Each dose
`consisted of three tablets from the same row of a blister pack (pla-
`cebo–placebo–placebo; placebo–placebo–25 mg; placebo–place-
`bo–50 mg; or placebo–50 mg–50 mg). The men were instructed
`not to consume more than two alcoholic drinks within one hour
`of sexual activity. Each man completed the International Index of
`Erectile Function at 0, 12, and 24 weeks and was asked about
`global efficacy at 12 and 24 weeks. The event logs were reviewed
`at 0, 2, 4, 8, 12, 16, 20, and 24 weeks.
`
`Study of Flexible Dose Escalation,
`Efficacy, and Safety with a Long-Term,
`Open-Label Extension
`In the flexible dose-escalation study, 329 different men were
`randomly assigned to take placebo or 50 mg of sildenafil approx-
`imately one hour before sexual activity for 12 weeks. At each fol-
`low-up visit, the dose could be doubled or reduced by 50 percent
`on the basis of the therapeutic response and adverse effects. Each
`man completed the International Index of Erectile Function at
`0 and 12 weeks and was asked about global efficacy at week 12.
`The event logs were reviewed at 0, 2, 4, 8, and 12 weeks. The
`men who completed the study and who did not have any serious
`adverse effects were eligible to receive open-label sildenafil for an
`additional 32 weeks.
`
`Statistical Analysis
`The mean frequency of responses to questions 3 and 4 of the
`International Index of Erectile Function for each treatment group
`was calculated. An analysis-of-covariance model was fitted for
`each question, which included main-effect terms for investiga-
`tional center and treatment effect (as ordered categorical vari-
`ables), with base-line score, patient age, smoking, and duration
`and cause of erectile dysfunction as covariates. Mean domain
`scores from the International Index were calculated, and the
`treatment effect was analyzed by using the analysis-of-covariance
`model described above. From the event log, the mean numbers
`of grade 3 and grade 4 erections (in the dose–response study) or
`the percentage of attempts at sexual intercourse that were success-
`ful (in the dose-escalation study) was calculated. Analysis of co-
`variance (dose–response study), with adjustment for the covari-
`ates listed above, or a chi-square test (dose-escalation study) was
`used to determine the association between the treatment groups.
`The answers of each treatment group to the global-efficacy ques-
`tion (yes or no) were analyzed with the use of logistic-regression
`analysis, accounting for the same covariates as those listed for the
`analysis-of-covariance models. Intention-to-treat analyses were
`performed on all variables and included all the men who were
`randomly assigned to treatment (and received treatment) and who
`had any assessments after base line, regardless of protocol devia-
`tions or whether the men completed the study. All statistical tests
`were two-sided.
`
`ATI 1027-0002
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`ORAL SILDENAFIL IN THE TREATMENT OF ERECTILE DYSFUNCTION
`
`T
`ABLE
`
` 1.
`
` D
` E
`
` M
`
` C
` B
`-L
`
`
`RECTILE
`WITH
`EN
`THE
`OF
`HARACTERISTICS
`INE
`ASE
`YSFUNCTION
`T
`
` S
`
` P
`
` T
` S
`.
`IN
`WO
`REATED
`WITH
`ILDENAFIL
`OR
`LACEBO
`TUDIES
`
`C
`HARACTERISTIC
`
`–R
`D
`ESPONSE
`OSE
`PLACEBO
`(
`216)
`⫽
`N
`
` S
`TUDY
`SILDENAFIL
`(
`316)
`⫽
`N
`
`D
`OSE
`
`-E
`SCALATION
`PLACEBO
`(
`166)
`⫽
`N
`
` S
`TUDY
`SILDENAFIL
`(
`163)
`⫽
`N
`
`57
`20–79
`3.2
`
`58
`24–87
`3.2
`
`59
`31–81
`4.7
`
`60
`26–79
`5.0
`
`55
`14
`31
`
`24
`15
`15
`
`98
`
`63
`16
`22
`
`28
`8
`14
`11
`11
`
`77
`10
`13
`
`26
`8
`16
`10
`15
`
`78
`9
`13
`
`30
`8
`19
`12
`13
`
`Age (yr)
`Mean
`Range
`Mean duration of erectile dysfunction (yr)
`Cause of erectile dysfunction (% of men)*
`Organic
`Psychogenic
`Mixed
`Concomitant condition (% of men)
`Hypertension
`Ischemic heart disease (past or present)
`Hyperlipidemia
`History of radical prostatectomy
`Diabetes mellitus
`
`*The cause of erectile dysfunction was determined by the investigators on the basis of the history,
`physical examination, and additional diagnostic studies (see the Methods section). Because of round-
`ing, percentages do not always total 100.
`
`RESULTS
`The base-line characteristics of the men with erec-
`tile dysfunction enrolled in each study were similar,
`but there were differences between the studies (Ta-
`ble 1). The men in the dose–response study had had
`erectile dysfunction for longer periods, and fewer of
`them had organic erectile dysfunction. Among the
`532 men in the dose–response study, 465 (87 per-
`cent) completed the 24-week study (285 of 316 in
`the sildenafil group and 180 of 216 in the placebo
`group). Among the 329 men in the dose-escalation
`study, 307 (93 percent) completed the 12-week study
`(154 of 163 in the sildenafil group and 153 of 166
`in the placebo group).
`After 12 weeks of treatment in the dose-escalation
`study, the proportions of men taking 25, 50, or 100
`mg of sildenafil were 2 percent (4 men), 23 percent
`(38 men), and 74 percent (121 men), respectively.
`For the men taking placebo, the corresponding pro-
`portions were 0 percent, 5 percent (8 men), and 95
`percent (158 men). Two hundred twenty-five men
`who completed the 12-week study were enrolled to
`receive open-label sildenafil for an additional 32
`weeks.
`
`Efficacy
`In the dose–response study, increasing doses of
`sildenafil were associated with higher mean scores for
`the questions of the International Index of Erectile
`Function assessing frequency of penetration (question
`3) and maintenance of erections after sexual penetra-
`tion (question 4) (P
`0.001) (Table 2). The mean
`⬍
`scores for these questions did not vary according to
`
`the cause of erectile dysfunction. For question 3, the
`percentage increases in mean score from base line to
`the end of treatment were 60, 84, and 100 percent
`for the men who received 25, 50, and 100 mg of
`sildenafil, respectively, as compared with an increase
`of 5 percent for the men who received placebo. For
`question 4, the corresponding values were 121, 133,
`and 130 percent for the men who received 25, 50,
`and 100 mg of sildenafil, respectively, as compared
`with 24 percent for those who received placebo.
`In the dose-escalation study, the mean scores for
`questions 3 and 4 of the International Index were
`significantly higher after treatment for the sildenafil
`0.001) (Table
`group than for the placebo group (P
`⬍
`2). The percent increase from base line was 95 per-
`cent for question 3 and 140 percent for question 4
`for the men taking sildenafil, as compared with 10
`percent and 13 percent, respectively, for those tak-
`ing placebo.
`The mean scores for the erectile-function domain
`on the International Index increased with increasing
`doses of sildenafil in the dose–response study (P
`⬍
`0.001). The mean score for the erectile-function do-
`main in the dose-escalation study was significantly
`higher for the men taking sildenafil (22.1) than for
`those taking placebo (12.2, P
`0.001) (Fig. 1A).
`⬍
`The mean scores for the orgasmic-function, inter-
`course-satisfaction, and overall-satisfaction domains
`were also significantly higher in the sildenafil group
`(P
`0.001) (Fig. 1B), whereas the mean scores for
`⬍
`sexual desire were not significantly different in the
`two groups (P
`0.13).
`⫽
`The event-log data on the proportion of men
`
`Volume 338 Number 20
`
`ⴢ
`
`1399
`
`ATI 1027-0003
`
`

`

`The New England Journal of Medicine
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` 3
` Q
`
` R
`
` S
` M
` 2.
`T
`UESTION
`TO
`ESPONSES
`OF
`CORES
`EAN
`ABLE
` Q
` 4 OF THE INTERNATIONAL INDEX OF ERECTILE
`AND
`UESTION
`FUNCTION FOR THE MEN RECEIVING SILDENAFIL OR PLACEBO
`IN TWO STUDIES.*
`
`TREATMENT GROUP
`
`Question 3
`Dose–response
`Placebo (n⫽199)
`Sildenafil
`25 mg (n⫽96)
`50 mg (n⫽105)
`100 mg (n⫽101)
`Dose escalation
`All the men
`Placebo (n⫽138)
`Sildenafil (n⫽138)
`Organic cause
`Placebo (n⫽90)
`Sildenafil (n⫽81)
`Psychogenic cause
`Placebo (n⫽24)
`Sildenafil (n⫽19)
`Mixed cause
`Placebo (n⫽24)
`Sildenafil (n⫽38)
`Question 4
`Dose–response
`Placebo (n⫽199)
`Sildenafil
`25 mg (n⫽96)
`50 mg (n⫽105)
`100 mg (n⫽101)
`Dose escalation
`All the men
`Placebo (n⫽138)
`Sildenafil (n⫽137)
`Organic cause
`Placebo (n⫽90)
`Sildenafil (n⫽80)
`Psychogenic cause
`Placebo (n⫽24)
`Sildenafil (n⫽19)
`Mixed cause
`Placebo (n⫽24)
`Sildenafil (n⫽38)
`
`BASE-LINE
`SCORE†
`
`FINAL
`SCORE†
`
`PERCENT
`CHANGE FROM
`BASE LINE‡
`
`P
`VALUE§
`
`2.1⫾0.1
`
`2.2⫾0.2
`
`2.0⫾0.2
`1.9⫾0.2
`2.0⫾0.2
`
`3.2⫾0.2
`3.5⫾0.2
`4.0⫾0.2
`
`2.1⫾0.1
`2.0⫾0.1
`
`2.3⫾0.1
`3.9⫾0.1
`
`2.0⫾0.2
`1.8⫾0.2
`
`2.0⫾0.2
`3.6⫾0.2
`
`2.2⫾0.2
`2.0⫾0.2
`
`2.3⫾0.4
`4.3⫾0.4
`
`2.3⫾0.3
`2.3⫾0.3
`
`2.8⫾0.3
`3.6⫾0.3
`
`1.7⫾0.1
`
`2.1⫾0.2
`
`1.4⫾0.1
`1.5⫾0.1
`1.7⫾0.1
`
`3.1⫾0.2
`3.5⫾0.2
`3.9⫾0.2
`
`1.6⫾0.1
`1.5⫾0.1
`
`1.8⫾0.1
`3.6⫾0.1
`
`1.4⫾0.1
`1.4⫾0.1
`
`1.4⫾0.2
`3.3⫾0.2
`
`1.7⫾0.2
`1.6⫾0.2
`
`1.9⫾0.3
`3.8⫾0.4
`
`1.8⫾0.3
`1.6⫾0.2
`
`2.3⫾0.4
`3.7⫾0.4
`
`5
`
`60
`84
`100
`
`10
`95
`
`0
`100
`
`5
`115
`
`22
`57
`
`24
`
`121
`133
`130
`
`13
`140
`
`0
`136
`
`12
`138
`
`28
`131
`
`⬍0.001
`
`⬍0.001
`
`⬍0.001
`
`⬍0.001
`
`0.08
`
`⬍0.001
`
`⬍0.001
`
`⬍0.001
`
`⬍0.001
`
`0.005
`
`*Question 3 of the International Index of Erectile Function is, “When
`you attempted sexual intercourse, how often were you able to penetrate
`your partner?” Question 4 is, “During sexual intercourse, how often were
`you able to maintain your erection after you had penetrated your partner?”
`The analysis was by intention to treat.
`†Scores are based on a scale of 1 (almost never or never) to 5 (almost
`always or always), with 0 representing “did not attempt intercourse.” Val-
`ues are means
`SE.
`⫾
`‡Percent differences are between the final (end-of-treatment) mean
`scores and the base-line mean scores.
`§P values are calculated according to analysis of covariance (ordered cat-
`egorical variable), with base-line score, patient’s age, smoking, and dura-
`tion and cause of erectile dysfunction as covariates.
`
`1400
`
`ⴢ
`
`May 14, 1998
`
`achieving erections hard enough for sexual inter-
`course (i.e., grade 3 or 4) during the last four weeks
`of treatment showed a significant dose–response re-
`lation for sildenafil (72 percent, 80 percent, and 85
`percent for doses of 25 mg, 50 mg, and 100 mg, re-
`spectively, as compared with 50 percent for placebo;
`P
`0.001). The mean number of grade 3 and grade
`⬍
`4 erections and the mean number of grade 4 erec-
`tions during the last four weeks of treatment were
`also significantly higher in the sildenafil group (P
`⬍
`0.001) (Fig. 2A), with 80 percent of the grade 3
`erections and 94 percent of the grade 4 erections re-
`sulting in successful sexual intercourse. In the dose-
`escalation study, 69 percent of all attempts at sexual
`intercourse by the men receiving sildenafil were suc-
`cessful in the last four weeks of treatment, as com-
`pared with 22 percent for those receiving placebo
`(P
`0.001) (Fig. 2B). During the last four weeks of
`⬍
`treatment, the mean numbers of attempts at sexual
`intercourse that were successful were 5.9 for men in
`the sildenafil group and 1.5 for men in the placebo
`0.001) (Fig. 2B).
`group (P
`⬍
`After 24 weeks of treatment in the dose–response
`study, improved erections were reported by 56, 77,
`and 84 percent of the men taking 25, 50, and 100
`mg of sildenafil, respectively, as compared with 25
`percent of those taking placebo (P
`0.001 for treat-
`⬍
`ment effect). After 12 weeks of treatment in the
`dose-escalation study, 101 of the 136 men in the
`sildenafil group who responded to the global-effica-
`cy question (74 percent) reported improved erec-
`tions, as compared with 23 of the 118 men in the
`placebo group who responded to the question (19
`0.001).
`percent, P
`⬍
`
`Cessation and Adverse Effects of Treatment
`During the dose–response study, 31 of the 316
`men in the sildenafil group (10 percent) and 36 of
`the 216 men in the placebo group (17 percent) dis-
`continued treatment (Table 3). Four men (1 per-
`cent) in the sildenafil group stopped taking the drug
`because of treatment-related adverse effects (nausea
`and vomiting in one, leg pain and backache in one,
`intermittent headache and dyspepsia in one, and
`1
`headache in one), as compared with one man (
`⬍
`percent) who stopped taking placebo (because of
`headache and nausea). Five men (2 percent) in the
`sildenafil group and 11 men (5 percent) in the pla-
`cebo group discontinued treatment because of in-
`sufficient responses. In the dose-escalation study,
`9 men (6 percent) stopped taking sildenafil and 13
`men (8 percent) stopped taking placebo. One man
`stopped taking sildenafil because of treatment-relat-
`ed headache and flushing, and one man stopped
`because of an insufficient response. Laboratory-test
`results indicated no evidence of sildenafil-induced
`abnormalities.
`The most frequently reported adverse effects of
`
`ATI 1027-0004
`
`

`

`ORAL SILDENAFIL IN THE TREATMENT OF ERECTILE DYSFUNCTION
`
`25
`
`20
`
`15
`
`10
`
`5
`
`0
`
`8
`
`5
`
`2
`
`8
`
`5
`
`2
`
`Erectile Function
`
`Base line
`End of study
`
`*
`
`Placebo
`
`Sildenafil
`
`Sexual Desire
`
`Placebo
`
`Sildenafil
`
`Overall Satisfaction
`
`*
`
`Placebo
`
`Sildenafil
`
`Mean Domain Score
`
`A
`
`Figure 1. Mean (⫾SE) Scores for Domains of the Interna-
`tional Index of Erectile Function for Men Receiving Sil-
`denafil or Placebo in the Dose-Escalation Study at Base
`Line and at the End of the 12-Week Study by Intention-to-
`Treat Analysis.
`Panel A shows the scores for the erectile-function domain
`(six questions; possible total score, 1 to 30) for 134 men in
`the placebo group and 136 men in the sildenafil group.
`Panel B shows the scores for the orgasmic-function do-
`main (two questions; possible total score, 0 to 10); the sex-
`ual-desire domain (two questions; possible total score, 2 to
`10); the intercourse-satisfaction domain (three questions;
`possible total score, 0 to 15); and the overall-satisfaction
`domain (two questions; possible total score, 2 to 10).
`These scores were for 137 to 139 men in the placebo group
`and 134 to 138 men in the sildenafil group. Asterisks de-
`note P⬍0.001 for the comparison with placebo.
`
`Orgasmic Function
`
`Base line
`End of study
`
`*
`
`Placebo
`
`Sildenafil
`
`Intercourse Satisfaction
`
`*
`
`Placebo
`
`Sildenafil
`
`B
`
`8
`
`6
`
`4 2
`
`0
`
`12
`
`8
`
`4
`
`0
`
`Mean Domain Score
`
`Mean Domain Score
`
`sildenafil in the two studies were transient headache,
`flushing, dyspepsia, and rhinitis (Table 3). Transient
`visual disturbances (i.e., changes in the perception of
`color hue or brightness) were reported by some
`men. The frequency of these adverse effects increased
`with increasing doses of sildenafil, but the symptoms
`were usually mild and lasted a few minutes to a few
`hours after dosing. No man reported priapism dur-
`ing the studies.
`Of the 225 men enrolled in the open-label exten-
`
`sion study, 207 (92 percent) completed an addition-
`al 32 weeks of sildenafil treatment. Four men (2 per-
`cent) withdrew because of treatment-related adverse
`effects (headache in two, intermittent flushing and
`blurred vision in one, and groin pain and headache
`in one).
`
`DISCUSSION
`We found that sildenafil improves sexual function
`in men with erectile dysfunction. In keeping with
`
`Volume 338 Number 20
`
`ⴢ 1401
`
`ATI 1027-0005
`
`

`

`The New England Journal of Medicine
`
`2.4
`
`0.7
`
`8
`
`1234567
`
`0
`
`Mean Number of Erections
`
`per Month
`
`Figure 2. Results from Logs of Sexual Function Kept by the Men
`in the Sildenafil and Placebo Groups.
`Panel A shows the mean numbers of erections of grade 3 and
`grade 4 during the last 4 weeks of treatment in the 24-week
`dose–response study for 205 men receiving placebo, 97 men
`receiving 25 mg of sildenafil, 105 men receiving 50 mg of
`sildenafil, and 102 men receiving 100 mg of sildenafil. Grade 1
`indicates that the penis is larger but not hard; grade 2, that it
`is hard but not hard enough for penetration; grade 3, that it is
`hard enough for penetration but not completely hard; and
`grade 4, that it is completely hard and fully rigid. Asterisks de-
`note P⬍0.001 for the comparison with placebo of grade 4 and
`of grade 3 plus grade 4. Panel B shows the percentages of all
`attempts at sexual intercourse that were successful and the
`mean numbers of successful attempts during the last 4 weeks
`of treatment in the 12-week dose-escalation study for 154 men
`receiving placebo and 157 men receiving sildenafil. Asterisks
`denote P⬍0.001 for the comparison with placebo.
`
`Erections
`Grade 3 (hard enough
`for penetration)
`Grade 4 (completely
`hard and fully rigid)
`Grade 3 plus grade 4
`*
`2.1
`
`*
`2.4
`
`4.2
`
`*
`3.1
`
`3.4
`
`50
`Sildenafil (mg)
`
`100
`
`Mean No. of Successful
`
`Attempts per Month
`
`2.2
`
`25
`
`10
`
`8
`
`6
`
`4
`
`2
`
`0
`
`A
`
`Placebo
`
`Sexual Intercourse
`Base line
`End of study
`No. of attempts per month
`
`*
`
`*
`
`Placebo
`
`Sildenafil
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`That Were Successful
`Percentage of Attempts
`
`B
`
`sildenafil’s mode of action (i.e., the drug causes
`erection only in response to sexual stimulation), the
`studies were performed entirely in a natural environ-
`ment, which meant that we had to rely on the men’s
`own reports of efficacy. However, the self-adminis-
`tered International Index of Erectile Function has a
`high degree of sensitivity and specificity for detect-
`ing treatment-related changes in men with erectile
`dysfunction.13 This questionnaire, together with the
`global-efficacy question, provided a comprehensive
`assessment of erectile function during a specified re-
`call period (four weeks), and the event log provided
`information about individual events during treat-
`ment. The men’s partners were questioned in each
`of the studies, and although the results corroborat-
`ed the men’s efficacy assessments, only 25 percent
`of the partners completed the optional question-
`naire.
`Recognizing the fact that in most men erectile
`
`dysfunction is a multifactorial problem, we enrolled
`men with a broad variety of causes of erectile dys-
`function. The identification of an organic abnormal-
`ity does not establish it as the cause of erectile dys-
`function.14 Conversely, failure to identify an organic
`cause does not prove a psychogenic origin.
`In the dose–response study, increasing doses of
`sildenafil (25 to 100 mg) were increasingly effective
`in improving the frequency of penetration and the
`maintenance of erections after penetration, the mean
`score for the erectile-function domain of the Inter-
`national Index, and the percentage of men reporting
`better erections. In the dose-escalation study, silden-
`afil treatment was associated with improvements in
`the frequency of penetration and the maintenance of
`erections after penetration. The therapeutic response
`to sildenafil was similar in men with various causes
`of erectile dysfunction. The mean scores after silden-
`afil therapy approached those of normal men of the
`
`1402 ⴢ May 14, 1998
`
`ATI 1027-0006
`
`

`

`ORAL SILDENAFIL IN THE TREATMENT OF ERECTILE DYSFUNCTION
`
`TABLE 3. REASONS FOR DISCONTINUING TREATMENT AND SUMMARY OF ADVERSE EFFECTS IN THE MEN
`WITH ERECTILE DYSFUNCTION TREATED WITH SILDENAFIL OR PLACEBO IN TWO STUDIES.
`
`VARIABLE
`
`DOSE–RESPONSE STUDY
`
`PLACEBO
`(N⫽216)
`
`25 mg
`(n ⫽102)
`
`SILDENAFIL
`50 mg
`(n ⫽107)
`
`100 mg
`(n ⫽107)
`
`DOSE-ESCALATION
`STUDY
`SILDENAFIL
`PLACEBO
`(N⫽166)
`(N⫽163)
`
`DOSE-
`ESCALATION
`OPEN-LABEL
`EXTENSION
`SILDENAFIL
`(N⫽225)
`
`number of men (percent)
`
`9 (6)
`1 (1)
`1 (1)
`7 (4)
`
`30 (18)
`30 (18)
`9 (6)
`8 (5)
`4 (2)
`
`18 (8)
`4 (2)
`7 (3)
`7 (3)
`
`28 (12)
`22 (10)
`12 (5)
`4 (2)
`9 (4)
`
`13 (8)
`
`13
`
` (2)
`9 (5)
`
`8 (7)
`2 (2)
`
`06
`
` (6)
`
`32 (30)
`21 (20)
`17 (16)
`12 (11)
`10 (9)
`
`6 (4)
`1 (1)
`4 (2)
`1 (1)
`1 (1)
`
`Reason for discontinuation
`All causes
`Treatment-related adverse effect
`Insufficient response
`Other*
`Adverse effect†
`Headache
`Flushing
`Dyspepsia
`Rhinitis
`Visual disturbance‡
`
`36 (17)
`1 (⬍1)
`11 (5)
`24 (11)
`
`14 (6)
`3 (1)
`3 (1)
`4 (2)
`1 (⬍1)
`
`15 (15)
`1 (1)
`3 (3)
`11 (11)
`
`14 (14)
`13 (13)
`3 (3)
`1 (1)
`2 (2)
`
`8 (7)
`1 (1)
`2 (2)
`5 (5)
`
`23 (21)
`29 (27)
`12 (11)
`3 (3)
`6 (6)
`
`*Other reasons for discontinuation included protocol violations, not returning for follow-up, adverse effects not related
`to treatment, withdrawal of consent, and other reasons.
`†The adverse effects listed are those that occurred in 5 percent or more of any treatment group.
`‡The visual disturbances reported were changes in the perception of color hue or brightness.
`
`same age. The men we studied had a normal level of
`sexual desire, as might be expected of men with
`erectile dysfunction who enter a clinical trial,13 and
`sildenafil did not alter that level. Successful sexual
`intercourse was therefore a key end point for these
`men. Overall, the results of the efficacy assessments
`demonstrated that sildenafil significantly improved
`erectile function and quadrupled the success of in-
`tercourse, with effectiveness maintained for at least
`six months.
`Sildenafil treatment was well tolerated. Its main
`adverse effects were headache, flushing, dyspepsia,
`and visual disturbances and were usually mild. Only
`one man with a visual disturbance discontinued treat-
`ment, and he also had flushing. Few men discontin-
`ued sildenafil, suggesting a relatively high level of
`drug tolerability and acceptance. No man had pri-
`apism after sildenafil treatment. The most common
`adverse effects reflect the pharmacologic nature of
`sildenafil as a phosphodiesterase-type-5 inhibitor
`(headache, flushing, and dyspepsia) and as a weak
`phosphodiesterase-type-6 inhibitor (visual effects). Sil-
`denafil has modest vasodilator properties but no ef-
`fect on heart rate.
`The American Urological Association Panel on
`the Treatment of Organic Erectile Dysfunction stat-
`ed that the ultimate goal is a therapy that is reliable,
`has minimal side effects, and is simple to use.6 Sil-
`
`denafil appears to meet these specifications. Oral ther-
`apy permits discreet administration and is less in-
`vasive than some other treatment options, including
`injections into the corpus cavernosum, transurethral
`drug delivery, and prosthesis implantation.
`
`Supported by grants from Pfizer.
`Drs. Goldstein, Lue, Padma-Nathan, Rosen, and Steers have served as
`consultants to Pfizer, manufacturer of sildenafil, as well as to other compa-
`nies making products for the treatment of erectile dysfunction.
`
`We are indebted to all the investigators who participated in the
`trials; to Dr. Ian Osterloh and Dr. David Cox of Pfizer for their help
`with the study design and logistics; to Dr. Michael Collins and Dr.
`Michael Smith of Pfizer for their assistance with data analysis; and
`to Dr. Patricia Leinen and Dr. Michael Sweeney for aid in the prep-
`aration of the manuscript.
`
`APPENDIX
`The other members of the Sildenafil Study Group are S. Auerbach, New-
`port Beach, Calif.; A.L. Burnett, Baltimore; R. Castellanos, Fort Myers,
`Fla.; L. Charles, Stratford, N.J.; F. Eid, New York; R. Feldman, Waterbury,
`Conn.; W. Fitch III, Sa