Study design and participants
This was a secondary analysis of data from 4 clinical trials. Details of the study designs and populations have previously been published
Study 1 (LVHJ) was a randomized, double-blind, placebo-controlled, parallel-design multinational study to evaluate the efficacy and safety of daily tadalafil 5 mg once daily for 12 weeks. Study participants were men with LUTS secondary to BPH (BPH-LUTS) residing in Argentina, Germany, Italy, Mexico, or the United States (Clinicaltrials.gov: NCT00827242
Study 2 (LVHR) was a randomized, double-blind, 3-group, placebo-controlled, parallel-design, multinational study to evaluate the efficacy and safety of tadalafil 2.5- and 5-mg once-daily dosing for 12 weeks for the treatment of erectile dysfunction (ED) and BPH-LUTS. Study participants were men with both ED and BPH-LUTS residing in Canada, France, Germany, Greece, Italy, Mexico, Portugal, Russia, and the United States (Clinicaltrials.gov: NCT00855582
Study 3 (LVHB) was a randomized, double-blind, 4-group, placebo and tamsulosin 0.2 mg controlled, parallel design, multinational study to evaluate the efficacy and safety of tadalafil 2.5 mg and 5 mg once-a-day dosing for 12 weeks in men with BPH-LUTS residing in Japan, Korea, and Taiwan (Clinicaltrials.gov: NCT00861757
Study 4 (LVHT) was a randomized, double-blind, 3-group, placebo-controlled, parallel-design, pilot study to evaluate the efficacy and safety of tadalafil 5 mg and tamsulosin 0.2 mg once-a-day dosing for 12 weeks in men with BPH-LUTS residing in Korea (Clinicaltrials.gov: NCT00540124
Men who were at least 45 years of age, with moderate to severe LUTS due to BPH and evidence of bladder obstruction, were eligible to participate in all 4 studies. ED was an entry criterion only in Study 2 (LVHR). The 4 studies were in compliance with the Helsinki Declaration.
In each study, participants were screened at V1. In all 4 studies, there was a 2- (Study 3, LVHB) or 4-week screening/washout period (V1-V2), a 4-week placebo lead-in period (V2-V3), and a 12-week double-blinded treatment period (V3 to last visit). At V2, participants in all studies were required to have an IPSS ≥13 and a uroflowmetry measure of urinary peak flow rate (Qmax) ≥4 to ≤15 ml/second on a voided volume of 125 mL to continue in the study. One study (Study 3, LVHB) also required a minimum total prostate volume of 20 mL. V3 (randomization) initiated the double-blind active-treatment placebo-controlled 12-week period. Patient characteristics were assessed at V2.
Postvoid residual volume (PVR) and prostate-specific antigen (PSA) were assessed in all studies, but not consistently carried out at the V2. The change in peak urine flow rate (Qmax) from randomization (V3) to endpoint was a secondary objective in all the studies. Changes in Qmax included in this validation were based on assessment at V2 and at endpoint.
The change in IPSS from randomization (V3) to endpoint was the primary objective in all the studies; changes in IPSS-QoL, BII and Qmax from V3 to endpoint were secondary objectives. Changes in these patient-reported outcomes included in this validation were based on assessment at V2 and at endpoint.
The PGI-I also was a secondary objective, with 1 assessment at endpoint. PGI-S assessed participants’ perception of symptom severity at V2 and was an exploratory parameter. The Korean pilot study (Study 4) did not assess BII at V2 and the BII scores could therefore not be used from this study to correlate with PGI-S.
The IPSS is a self-administered 7-item urinary symptom severity scale about symptoms occurring over the past month (Additional file
1; International Prostate Symptom Score). The urinary symptoms can be categorized into voiding, storage and post-micturition symptoms. Item scores are summed for a total IPSS score that ranges from 0 to 35 with a higher score indicating more severe symptoms.
The IPSS-QoL is a single question: “If you were to spend the rest of your life with your urinary condition just the way it is now, how would you feel about that?” with responses scored as follows: “delighted” (0), “pleased” (1), “mostly satisfied” (2), “mixed about equally satisfied and dissatisfied” (3), “mostly dissatisfied” (4), “unhappy” (5), and “terrible” (6) (Additional file
1; International Prostate Symptom Score).
The BII is a self-administered validated questionnaire with 4 questions about urinary problems during the past month regarding physical discomfort, worry about health, how bothersome symptoms are, and whether the symptoms are interfering with doing usual activities (Additional file
2; BPH Impact Index). The first 3 questions are scored from 0 to 3, while the fourth is scored from 0 to 4. Item scores are summed for a BII total score that ranges from 0 to 13, with higher scores indicating more bother or problems associated with urinary symptoms within the previous month.
PGI-I is a 1-item questionnaire designed to assess the patient’s impression of changes in urinary symptoms. The PGI-I asks the patient to best describe how his urinary symptoms are now, compared with how they were before he began taking medication in the study. (“Check the one number that
best describes how your urinary symptoms are now, compared with how they were before you began taking medication in this study”). The patient enters his answer on a 7-point scale scored as: (1) “very much better,” (2) “much better,” (3) “a little better,” (4) “no change,” (5) “a little worse,” (6) “much worse,” or (7) “very much worse.” Examples of urinary symptoms were provided based on the symptoms in the IPSS questionnaire
 with the addition of terminal dribble and accidental urinary leakage
: “Urinary symptoms include difficulties in postponing urination, having to push or strain to begin urination, a weak urinary stream, stopping and restarting urination several times when attempting to urinate, prolonged urination with the end of urine flow slowing to a trickle (terminal dribble), the feeling that you haven’t emptied your bladder after you have finished urinating, having to urinate again less than 2 hours after the last time you finished urinating, accidental urinary leakage, or having to get up at night to urinate”. The PGI-I was administered once at the end-of-study visit (Week 12 or the last visit).
PGI-S is a 1-item questionnaire designed to assess patient’s impression of disease severity. The PGI-S item asks the respondent to best describe how his urinary symptoms are now (“Check the one number that
best describes how your urinary symptoms are now”) on a 4-point scale scored as: “normal” (1), “mild” (2), “moderate” (3), or “severe” (4) . Examples of urinary symptoms similar to the PGI-I were provided. The PGI-S was administered once at the beginning of the placebo lead-in period (V2).
All patient-reported measures were originally developed in English for the United States. Therefore, linguistic validations that included forward and backward translations, review/reconciliation, and cognitive debriefing with targeted patients were conducted for primary languages spoken in the countries participating in the 4 trials. In addition, for most linguistic validations, harmonization, a meeting in which all translations and the original are compared to ensure cross-cultural equivalence of concepts and the use of colloquial language, was performed.
The study results for tadalafil efficacy versus placebo are reported elsewhere
In the current paper, because the aim of the analysis was not to make conclusions about tadalafil efficacy, but rather to establish construct validity of the PGI-S and the PGI-I, tadalafil and placebo data from the 4 clinical studies were pooled (N=1694) and analyzed. Furthermore, an evaluation of tadalafil efficacy versus placebo would be based on changes from randomization (V3) to endpoint, while the validation of the PGI-I per its question ( see above) was based on changes from treatment onset at V2 (initiating with single-blind placebo therapy) to endpoint.
For a single-item global assessment to adequately (i.e., validly) assess a patient’s overall appraisal of their condition of change in their condition, its single score would be expected to capture not only symptom burden but also impact of that symptom burden on the patient’s life. Therefore, we hypothesized that the PGI-S and PGI-I scores would be significantly and meaningfully associated with the IPSS, IPSS-QoL, and BII scores. In addition, we explored the relationship with Qmax, the only clinical measure consistently assessed at V2 in the 4 clinical studies.
The associations between PGI-S response at V2 and other patient-reported and clinical measures at V2 (IPSS, IPSS-QoL, BII, and Qmax) were evaluated using 2 types of analyses. First, Spearman rank correlation coefficients were calculated between PGI-S and the other measures. Spearman correlation coefficient is a nonparametric analysis which assesses how well the relationship between 2 variables can be described using a monotonic function
; for example, whether patients’ perception of greater severity, as measured by the PGI-S, is associated with greater symptom burden as measured by IPSS. Correlation coefficients range from −1.0 to 1.0. Using a criterion suggested by Guilford and Fruchter
, a significant correlation coefficient ≤−0.30 or ≥0.30 between the PGI-S and the V2 assessment of other patient-reported or clinical measures (i.e., supportive of the validity of the PGI-S).
The second type of analysis, unadjusted one-way analysis of variance (ANOVA), was performed to evaluate the differences of these measures among the 4 PGI-S categories and pair-wise comparisons were made. Identical ANOVA analyses, adjusted for covariates (e.g., age, prior alpha blocker use, country, baseline LUTs severity) were performed to confirm results. For all analyses, the a priori hypotheses were that patients perceiving more severe disease measured by PGI-S would have higher (worse) IPSS, IPSS-QoL, and BII scores and lower (worse) peak urine flow rate at V2.
The associations between PGI-I response at last visit and change in IPSS, IPSS-QoL, BII, and Qmax values from V2 to the last visit were evaluated using the same 2 types of analysis as performed with the PGI-S. For all analyses, the a priori hypotheses were that patients who reported change in their urinary symptoms would have better or worse IPSS, IPSS-QoL, and BII scores or Qmax values consistent with the direction of their perceived change in symptoms as measured by the PGI-I.
To further provide support for the validity of the PGI-S and PGI-I, the correlation analyses were repeated across 6 racial, ethnic, and severity subgroups: Asians versus Caucasians, Hispanic Caucasians versus non-Hispanic Caucasians, and LUTS severity at V2 (per IPSS classification of moderate versus severe). Assessment of race and ethnic subgroups was restricted within the relevant regions: Asians were only from Asian studies; Caucasians, including both Hispanic and non-Hispanic Caucasians, were only from non-Asian studies). It should be noted that these analyses were conducted to support the validity of the PGI-S and PGI-I and not to explore whether ethnic/racial groups differ in perception of severity or response to treatment. For all analyses, participants were included regardless of what type of treatment they received.