Psychosocial Outcomes After Tadalafil or Sildenafil for ED
Psychosocial Outcomes After Tadalafil or Sildenafil for ED
Of 770 patients randomized, 767 (99.6%) took ≥1 dose of study drug, 662 (86.0%) completed the study. Overall, 44.0% of all patients stayed on their randomized treatment and 42.0% switched at least once to another study treatment. Study discontinuation rates were significantly higher in patients randomized to sildenafil PRN (18.0%) compared with patients starting with tadalafil OaD (11.3%, P=0.035). Study discontinuation in the tadalafil PRN group was 12.7%. The most frequent reasons for study discontinuation in the sildenafil group were patient decision (6.1%) and loss to follow-up (5.0%; all other reasons in any treatment group ≤3.5%). From all patients randomized to tadalafil OaD, 45.5% completed on tadalafil OaD and 43.2% switched to a PRN treatment. In the group starting treatment with sildenafil PRN, 30.7% completed on sildenafil and 51.3% switched to either tadalafil PRN or OaD. In the tadalafil PRN group, 56.3% of randomized patients completed on tadalafil PRN and 31.0% switched.
Demographic, clinical and psychosocial characteristics relating to ED (IIEF, PAIRS and SEAR domain-scores) were balanced between treatment groups at baseline (Table 1).
Patients initiating ED treatment with tadalafil OaD or PRN reported significantly greater improvements of sexual self-confidence than patients randomized to sildenafil PRN (Figure 1). Least-square (LS) mean (s.e.) changes of PAIRS Sexual Self-Confidence domain score from baseline to ERT (last measurement while still receiving the initial randomized treatment) were +0.90 (0.048) with tadalafil OaD and +0.93 (0.050) with tadalafil PRN, vs +0.73 (0.049) with sildenafil PRN. LS mean (s.e.) changes of PAIRS Spontaneity domain scores at ERT were +0.11 (0.035) for tadalafil OaD, +0.13 (0.035) for tadalafil PRN and +0.02 (0.035) for sildenafil PRN. When the final values measured at the end of study (regardless whether patients still remained on originally randomized treatment or had changed=LOIS analysis) were compared, LS mean (s.e.) changes in the Sexual Self-Confidence (tadalafil OaD: 0.95 (0.048), tadalafil PRN: 0.93 (0.049), sildenafil PRN: 0.87 (0.048)) and Spontaneity (0.12 (0.034), 0.14 (0.035), 0.10 (0.034)) domain scores were not significantly different between treatment groups, and the overall treatment effect was not statistically significant (P=0.348 and P=0.674, respectively).
(Enlarge Image)
Figure 1.
Changes in Psychological and Interpersonal Relationship Scales ((range 1–4) PAIRS) scores (analysis of covariance, ANCOVA) from baseline to end of randomized treatment (ERT; domain scores range from 1 to 4; higher scores represent higher self-confidence, more time-concerns and more spontaneity, respectively). Significant interaction between baseline scores and randomized treatment (P=0.047 for ERT, P=0.018 for LOIS), see results. LS, least square; N, number of patients per treatment group; OaD, once-a-day; PRN, on demand; s.e., standard error; SIL, sildenafil; TAD, tadalafil.
LS mean (s.e.) changes of PAIRS Time-Concerns domain-scores at ERT were −0.20 (0.038) and −0.08 (0.039) for patients starting treatment with tadalafil OaD and PRN, respectively, and +0.04 (0.038) for patients randomized to sildenafil PRN. The statistical model revealed no overall treatment effect (P=0.310; Figure 1) and there was a significant interaction between randomized treatment and Time-Concerns scores at baseline (P=0.047 for ERT). Baseline Time-Concerns domain score and randomized treatment had an effect on the change measured during the study, and the two factors influenced each other. In the tadalafil OaD group, baseline scores had a higher impact on Time-Concerns reduction at ERT than in the tadalafil or sildenafil PRN groups. For each 1 point increase in baseline Time-Concerns scores, there was a reduction by 0.89 (95% confidence interval (CI): 0.75, 1.03) at ERT in the tadalafil OaD group, compared with 0.67 (0.53, 0.80) and 0.67 (0.53, 0.81) with tadalafil and sildenafil PRN.
In the LOIS analysis of the Time-Concerns domain (including data after treatment switch), the overall treatment effect (P=0.039) and the interaction between baseline score and randomized treatment (P=0.018) were statistically significant. Time-Concerns domain-scores were reduced in all treatment groups with different effect-sizes depending on the baseline score.
Country was included as covariate in the models. Although there was a significant country effect on the PAIRS Time-Concerns domain at the end of randomized treatment, no significant country effect on the PAIRS Sexual Self-Confidence and Spontaneity domain were observed.
SEAR domain scores for sexual relationship, confidence, self-esteem and overall relationship improved in all three treatment groups, with LS mean increases to ERT by 31–34 points for sexual relationship, 26–29 for confidence, 29–32 for self-esteem and 24–25 for overall relationship (Table 2). No statistically significant differences between treatment groups were detected at ERT or LOIS when using ANCOVA (Table 2) or MMRM (data not shown).
All three treatment groups showed similar improvements in the IIEF domain scores for EF, orgasmic function, sexual desire and overall satisfaction (Table 3), no statistically significant differences between treatment groups were detected at ERT or LOIS when using ANCOVA (Table 3) or MMRM (data not shown).
However, LS mean changes of the IIEF domain score for 'intercourse satisfaction' from baseline to ERT were significantly higher for the two PRN treatments (+4.1 points) compared with tadalafil OaD (+3.3 points; ANCOVA: tadalafil PRN vs OaD P=0.003, sildenafil PRN vs tadalafil OaD P=0.001; Table 3). A separate analysis of the individual IIEF-questions 6, 7 and 8 that compose the 'intercourse satisfaction' domain, showed a more pronounced increase in sexual intercourse attempts during PRN treatment (IIEF-question 6). Changes in LS mean (s.e.) scores for IIEF-question 6 from baseline to ERT were +0.9 (0.09) for tadalafil PRN, +1.0 (0.09) for sildenafil PRN, compared with +0.5 (0.09) for tadalafil OaD (P<0.001). There were no significant treatment group differences for the two other questions contributing to the intercourse satisfaction domain score (IIEF-question 7: 'When you attempted sexual intercourse, how often was it satisfactory for you?' and IIEF-question 8: 'How much have you enjoyed sexual intercourse?'). Also, no significant difference in the improvement of the intercourse satisfaction domain score or IIEF-question 6 was found in the LOIS analysis (that is, after switch to preferred treatment).
Patient assessment of satisfaction with erection and overall sexual experience based on SEP questions 4 and 5 revealed no significant differences between the three treatment groups. At ERT/LOIS, respectively, 56.1%/61.9% of patients randomized to tadalafil OaD, 57.2%/60.9% of patients randomized to tadalafil PRN, and 56.9%/59.6% of patients randomized to sildenafil PRN were satisfied with the hardness of erection (ANCOVA: P=0.930/P=0.714). Also, 50.3%/56.2% of patients randomized to tadalafil OaD, 52.3%/55.5% of patients randomized to tadalafil PRN and 54.5%/56.9% of patients randomized to sildenafil PRN were satisfied with the overall sexual experience (ANCOVA: P=0.383/P=0.870). MMRM results (data not shown) were consistent.
Patients' global assessment of efficacy by GAQ-1 and GAQ-2 also revealed no significant differences between the three treatment groups. At ERT/LOIS, respectively, 84.1%/90.5% of patients randomized to tadalafil OaD, 91.5%/91.6% randomized to tadalafil PRN and 86.1%/85.7% randomized to sildenafil PRN stated that treatment had improved their erection (logistic regression, ERT: P=0.052). Respectively, 96.0%/96.8%, 95.4%/93.7% and 95.4%/96.2% of patients stated that treatment had improved their ability to engage in sexual activity (ERT: P=0.865).
Treatment satisfaction, assessed by EDITS total scores, increased slightly during the study, without any significant difference between treatments at ERT or LOIS (Figure 2). Based on the MMRM analysis, LS mean (s.e.) EDITS total scores at ERT were 77.1 (1.21) for both tadalafil groups and 74.8 (1.22) for sildenafil PRN (MMRM, P=0.226). LS mean scores based on ANCOVA were not significantly different between treatment groups at ERT or end of study (P=0.125 and P=0.107, respectively).
(Enlarge Image)
Figure 2.
Treatment satisfaction, as assessed by Erectile Dysfunction Inventory of Treatment Satisfaction (EDITS) total score (mixed model for repeated measures (MMRM); scores range from 0 to 100, higher scores represent a more favorable response). LOIS value based on analysis of covariance. CI, confidence interval; ERT, end of randomized treatment; LOIS, last observation in study; LS mean, least square mean; N, number of patients per treatment group; OaD, once-a-day; PRN, on demand; SIL, sildenafil; TAD, tadalafil.
The only individual treatment-emergent AE (other than 'overdose') reported by >5% of patients was headache (tadalafil OaD 5.4%, tadalafil PRN 4.0%, sildenafil PRN 8.8%; P=0.055). There were no significant treatment group differences regarding the overall frequency of AEs, serious AEs, or AEs leading to discontinuation.
Results
Patient Disposition and Baseline Characteristics
Of 770 patients randomized, 767 (99.6%) took ≥1 dose of study drug, 662 (86.0%) completed the study. Overall, 44.0% of all patients stayed on their randomized treatment and 42.0% switched at least once to another study treatment. Study discontinuation rates were significantly higher in patients randomized to sildenafil PRN (18.0%) compared with patients starting with tadalafil OaD (11.3%, P=0.035). Study discontinuation in the tadalafil PRN group was 12.7%. The most frequent reasons for study discontinuation in the sildenafil group were patient decision (6.1%) and loss to follow-up (5.0%; all other reasons in any treatment group ≤3.5%). From all patients randomized to tadalafil OaD, 45.5% completed on tadalafil OaD and 43.2% switched to a PRN treatment. In the group starting treatment with sildenafil PRN, 30.7% completed on sildenafil and 51.3% switched to either tadalafil PRN or OaD. In the tadalafil PRN group, 56.3% of randomized patients completed on tadalafil PRN and 31.0% switched.
Demographic, clinical and psychosocial characteristics relating to ED (IIEF, PAIRS and SEAR domain-scores) were balanced between treatment groups at baseline (Table 1).
Psychosocial Outcomes (PAIRS, SEAR)
Patients initiating ED treatment with tadalafil OaD or PRN reported significantly greater improvements of sexual self-confidence than patients randomized to sildenafil PRN (Figure 1). Least-square (LS) mean (s.e.) changes of PAIRS Sexual Self-Confidence domain score from baseline to ERT (last measurement while still receiving the initial randomized treatment) were +0.90 (0.048) with tadalafil OaD and +0.93 (0.050) with tadalafil PRN, vs +0.73 (0.049) with sildenafil PRN. LS mean (s.e.) changes of PAIRS Spontaneity domain scores at ERT were +0.11 (0.035) for tadalafil OaD, +0.13 (0.035) for tadalafil PRN and +0.02 (0.035) for sildenafil PRN. When the final values measured at the end of study (regardless whether patients still remained on originally randomized treatment or had changed=LOIS analysis) were compared, LS mean (s.e.) changes in the Sexual Self-Confidence (tadalafil OaD: 0.95 (0.048), tadalafil PRN: 0.93 (0.049), sildenafil PRN: 0.87 (0.048)) and Spontaneity (0.12 (0.034), 0.14 (0.035), 0.10 (0.034)) domain scores were not significantly different between treatment groups, and the overall treatment effect was not statistically significant (P=0.348 and P=0.674, respectively).
(Enlarge Image)
Figure 1.
Changes in Psychological and Interpersonal Relationship Scales ((range 1–4) PAIRS) scores (analysis of covariance, ANCOVA) from baseline to end of randomized treatment (ERT; domain scores range from 1 to 4; higher scores represent higher self-confidence, more time-concerns and more spontaneity, respectively). Significant interaction between baseline scores and randomized treatment (P=0.047 for ERT, P=0.018 for LOIS), see results. LS, least square; N, number of patients per treatment group; OaD, once-a-day; PRN, on demand; s.e., standard error; SIL, sildenafil; TAD, tadalafil.
LS mean (s.e.) changes of PAIRS Time-Concerns domain-scores at ERT were −0.20 (0.038) and −0.08 (0.039) for patients starting treatment with tadalafil OaD and PRN, respectively, and +0.04 (0.038) for patients randomized to sildenafil PRN. The statistical model revealed no overall treatment effect (P=0.310; Figure 1) and there was a significant interaction between randomized treatment and Time-Concerns scores at baseline (P=0.047 for ERT). Baseline Time-Concerns domain score and randomized treatment had an effect on the change measured during the study, and the two factors influenced each other. In the tadalafil OaD group, baseline scores had a higher impact on Time-Concerns reduction at ERT than in the tadalafil or sildenafil PRN groups. For each 1 point increase in baseline Time-Concerns scores, there was a reduction by 0.89 (95% confidence interval (CI): 0.75, 1.03) at ERT in the tadalafil OaD group, compared with 0.67 (0.53, 0.80) and 0.67 (0.53, 0.81) with tadalafil and sildenafil PRN.
In the LOIS analysis of the Time-Concerns domain (including data after treatment switch), the overall treatment effect (P=0.039) and the interaction between baseline score and randomized treatment (P=0.018) were statistically significant. Time-Concerns domain-scores were reduced in all treatment groups with different effect-sizes depending on the baseline score.
Country was included as covariate in the models. Although there was a significant country effect on the PAIRS Time-Concerns domain at the end of randomized treatment, no significant country effect on the PAIRS Sexual Self-Confidence and Spontaneity domain were observed.
SEAR domain scores for sexual relationship, confidence, self-esteem and overall relationship improved in all three treatment groups, with LS mean increases to ERT by 31–34 points for sexual relationship, 26–29 for confidence, 29–32 for self-esteem and 24–25 for overall relationship (Table 2). No statistically significant differences between treatment groups were detected at ERT or LOIS when using ANCOVA (Table 2) or MMRM (data not shown).
Efficacy (IIEF, SEP, GAQs)
All three treatment groups showed similar improvements in the IIEF domain scores for EF, orgasmic function, sexual desire and overall satisfaction (Table 3), no statistically significant differences between treatment groups were detected at ERT or LOIS when using ANCOVA (Table 3) or MMRM (data not shown).
However, LS mean changes of the IIEF domain score for 'intercourse satisfaction' from baseline to ERT were significantly higher for the two PRN treatments (+4.1 points) compared with tadalafil OaD (+3.3 points; ANCOVA: tadalafil PRN vs OaD P=0.003, sildenafil PRN vs tadalafil OaD P=0.001; Table 3). A separate analysis of the individual IIEF-questions 6, 7 and 8 that compose the 'intercourse satisfaction' domain, showed a more pronounced increase in sexual intercourse attempts during PRN treatment (IIEF-question 6). Changes in LS mean (s.e.) scores for IIEF-question 6 from baseline to ERT were +0.9 (0.09) for tadalafil PRN, +1.0 (0.09) for sildenafil PRN, compared with +0.5 (0.09) for tadalafil OaD (P<0.001). There were no significant treatment group differences for the two other questions contributing to the intercourse satisfaction domain score (IIEF-question 7: 'When you attempted sexual intercourse, how often was it satisfactory for you?' and IIEF-question 8: 'How much have you enjoyed sexual intercourse?'). Also, no significant difference in the improvement of the intercourse satisfaction domain score or IIEF-question 6 was found in the LOIS analysis (that is, after switch to preferred treatment).
Patient assessment of satisfaction with erection and overall sexual experience based on SEP questions 4 and 5 revealed no significant differences between the three treatment groups. At ERT/LOIS, respectively, 56.1%/61.9% of patients randomized to tadalafil OaD, 57.2%/60.9% of patients randomized to tadalafil PRN, and 56.9%/59.6% of patients randomized to sildenafil PRN were satisfied with the hardness of erection (ANCOVA: P=0.930/P=0.714). Also, 50.3%/56.2% of patients randomized to tadalafil OaD, 52.3%/55.5% of patients randomized to tadalafil PRN and 54.5%/56.9% of patients randomized to sildenafil PRN were satisfied with the overall sexual experience (ANCOVA: P=0.383/P=0.870). MMRM results (data not shown) were consistent.
Patients' global assessment of efficacy by GAQ-1 and GAQ-2 also revealed no significant differences between the three treatment groups. At ERT/LOIS, respectively, 84.1%/90.5% of patients randomized to tadalafil OaD, 91.5%/91.6% randomized to tadalafil PRN and 86.1%/85.7% randomized to sildenafil PRN stated that treatment had improved their erection (logistic regression, ERT: P=0.052). Respectively, 96.0%/96.8%, 95.4%/93.7% and 95.4%/96.2% of patients stated that treatment had improved their ability to engage in sexual activity (ERT: P=0.865).
Treatment Satisfaction (EDITS)
Treatment satisfaction, assessed by EDITS total scores, increased slightly during the study, without any significant difference between treatments at ERT or LOIS (Figure 2). Based on the MMRM analysis, LS mean (s.e.) EDITS total scores at ERT were 77.1 (1.21) for both tadalafil groups and 74.8 (1.22) for sildenafil PRN (MMRM, P=0.226). LS mean scores based on ANCOVA were not significantly different between treatment groups at ERT or end of study (P=0.125 and P=0.107, respectively).
(Enlarge Image)
Figure 2.
Treatment satisfaction, as assessed by Erectile Dysfunction Inventory of Treatment Satisfaction (EDITS) total score (mixed model for repeated measures (MMRM); scores range from 0 to 100, higher scores represent a more favorable response). LOIS value based on analysis of covariance. CI, confidence interval; ERT, end of randomized treatment; LOIS, last observation in study; LS mean, least square mean; N, number of patients per treatment group; OaD, once-a-day; PRN, on demand; SIL, sildenafil; TAD, tadalafil.
Safety Data
The only individual treatment-emergent AE (other than 'overdose') reported by >5% of patients was headache (tadalafil OaD 5.4%, tadalafil PRN 4.0%, sildenafil PRN 8.8%; P=0.055). There were no significant treatment group differences regarding the overall frequency of AEs, serious AEs, or AEs leading to discontinuation.