Abbreviations & Acronyms BPS bladder pain syndrome CI confidence interval DMSO dimethyl sulfoxide FAS full analysis set GRA global response assessment IC interstitial cystitis ICPI O’Leary-Sant Interstitial Cystitis Problem Index ICSI O’Leary-Sant Interstitial Cystitis Symptom Index LS least squares Max maximum Min minimum MMRM mixed model for repeated measures NRS numeric rating scale ROC receiver operating characteristic SD standard deviation Introduction

IC/BPS is a chronic bladder condition characterized by bladder pain, frequency, and urgency in the absence of other well-defined pathologies.1-3 Its etiology is not well understood, nor is there an internationally accepted definition.4 The Japanese Guideline for Diagnosis and Treatment of IC 2007,1 and Clinical Guidelines for IC and Hypersensitive Bladder Syndrome 20095 defined IC as a bladder pathology proven endoscopically, such as by the presence of Hunner lesions and/or mucosal bleeding after hydrodistension. However, recent reports suggest that bladder glomerulations with hydrodistension-induced mucosal bleeding are not a specific finding of IC/BPS.6, 7 IC with Hunner lesions, the incidence of which is reportedly up to 57% in IC/BPS patients, has been reported to be clinically and pathologically different from other types of IC/BPS, and is categorized as a separate disease from non-Hunner-type IC/BPS.8, 9 Accordingly, in the updated Japanese10 and East Asian IC guidelines from 2019,11 IC is defined as a cystoscopy-identified Hunner lesion phenotype only.

The clinical benefits of intravesical KRP-116D, 50% DMSO solution versus placebo were demonstrated in a randomized placebo-controlled double-blind parallel phase III study,12 and KRP-116D was approved for the treatment of IC patients with Hunner lesions in Japan in 2021. This was despite the fact that the KRP-116D clinical trial included both IC patients with Hunner lesions and those with hydrodistension-induced glomerulations according to the Japanese IC guideline from 2009. In addition, previous studies that reported the therapeutic efficacy of intravesical DMSO for relieving IC/BPS symptoms also included the bladder-centric phenotype of IC/BPS patients identified by Hunner lesions (Hunner-type IC) or hydrodistension-induced glomerulations (non-Hunner-type IC).13, 14 Therefore, the effect of DMSO treatment has not yet been confirmed in a selected Hunner-type IC patient population. Furthermore, a wide variety of endpoints, including ICSI, ICPI, number of micturitions per 24 h, voided volume per micturition, maximum voided volume per micturition, and an NRS for bladder pain, were used for evaluating the outcomes of clinical studies in IC/BPS. It remains to be clarified which endpoint(s) and/or how much change in an endpoint are more meaningful for predicting the minimal clinically important change.

Thus, the aims of this post-hoc analysis of the Japanese phase III trial of KRP-116D were as follows: (i) to evaluate the efficacy of KRP-116D in patients with Hunner-type IC; (ii) to evaluate the correlations between various endpoints and GRA; and (iii) to determine the cut-off values for a minimal clinically important change in those endpoints that are closely correlated with GRA improvements.

Methods Study design

This was a post hoc analysis of the randomized placebo-controlled double-blind study of KRP-116D (trial registration: JapicCTI-173566). The study was conducted at 24 sites in Japan from May 2017 to July 2018, and the study design and results have been described previously.12 Briefly, the study population comprised Japanese patients who exhibited the bladder-centric phenotype of IC/BPS with Hunner lesions or hydrodistension-induced glomerulations identified by cystoscopy and bladder-derived pain with an ICSI of ≥9. Patients were randomly allocated to receive 50 mL of either KRP-116D (n = 49) or placebo (n = 47). The study drug was intravesically administered every 2 weeks, six times over 12 weeks.

Outcomes

Changes at Week 12 from baseline in ICSI, ICPI, NRS for bladder pain, number of micturitions/24 h, voided volume/micturition, and maximum voided volume/micturition were evaluated. Patients also completed a GRA questionnaire at Week 12, which is a seven-point symmetric scale: markedly improved, moderately improved, slightly improved, no change, slightly worse, moderately worse, and markedly worse. Patients whose condition was moderately improved or markedly improved were defined as responders. The Japanese-language versions of ICSI and ICPI, which were linguistically validated in the Japanese Guideline for Diagnosis and Treatment of IC 2007, were used.

Statistical analysis

Post hoc analyses were performed in the FAS population, which included all enrolled patients after excluding those who did not exhibit the bladder-centric phenotype or who had no efficacy data records after the start of treatment. Summary statistics of the change in efficacy variables at Week 12 from baseline were compared between the KRP-116D and placebo groups in Hunner-type IC or non-Hunner-type IC patients. The LS mean and two-sided 95% CI of changes in the efficacy variables at Week 12 from baseline in each patient group were calculated using a MMRM. The difference in the LS mean and the two-sided 95% CI between the groups was estimated using the MMRM, and the LS mean difference test (two-sided 5% significance level) was conducted to verify the difference between the KRP-116D and placebo groups in separate groups of patients with Hunner-type or non-Hunner-type IC. Thereafter, correlation coefficients between efficacy variables at Week 12 and GRA were analyzed in all Hunner-type IC patients pooled together regardless of their treatment status, either KRP-116D or placebo, using Pearson’s method. For efficacy variables closely correlated with GRA, the area under the ROC curve and the cut-off value for judgment were calculated in all Hunner-type IC patients including those treated with KRP-116D or placebo, together, as the value that maximizes the Youden index (sensitivity + specificity − 1).15

Results Patients

A total of 97 patients were confirmed as having the bladder-centric phenotype of IC/BPS identified by Hunner lesions or other pathologies such as hydrodistension-induced glomerulations during cystoscopy. Among 96 patients in the FAS population, 86.5% (n = 83) had Hunner-type IC (42 in the KRP-116D group and 41 in the placebo group), and 13.5% (n = 13) had non-Hunner-type IC (seven in the KRP-116D group and six in the placebo group; Table 1). Baseline demographics and disease characteristics were well balanced between the KRP-116D and placebo groups of patients with Hunner-type or non-Hunner-type IC.

Table 1. Summary of demographic and other baseline characteristics

Hunner-type IC (n = 83)

Non-Hunner-type IC (n = 13)

KRP-116D (n = 42)

Placebo (n = 41)

KRP-116D (n = 7)

Placebo (n = 6)

Sex, n (%) Male 5 (11.9) 2 (4.9) 1 (14.3) 1 (16.7) Female 37 (88.1) 39 (95.1) 6 (85.7) 5 (83.7) Age <65 years, n (%) 17 (40.5) 16 (39.0) 4 (57.1) 3 (50.0) ≥65 years, n (%) 25 (59.5) 25 (61.0) 3 (42.9) 3 (50.0) Mean ± SD, years 64.9 ± 13.8 65.5 ± 13.0 55.6 ± 14.9 57.3 ± 16.1 Min, median, max, years 31, 68.5, 87 36, 67.0, 87 30, 61.0, 70 36, 61.0, 77 Weight <50 kg, n (%) 17 (40.5) 18 (43.9) 3 (42.9) 1 (16.7) ≥50 kg, n (%) 25 (59.5) 23 (56.1) 4 (57.1) 5 (83.3) Mean ± SD, kg 55.40 ± 10.93 52.52 ± 12.11 58.24 ± 16.99 55.33 ± 6.28 Min, median, max, kg 38.6, 53.50, 80.5 29.6, 52.00, 86.8 42.3, 53.60, 90.0 46.5, 54.95, 63.8 Body mass index <25.0 kg/m2, n (%) 32 (76.2) 33 (80.5) 5 (71.4) 6 (100.0) ≥25.0 kg/m2, n (%) 10 (23.8) 8 (19.5) 2 (28.6) 0 (0.0) Mean ± SD, kg/m2 23.14 ± 3.88 22.17 ± 4.41 22.70 ± 4.06 21.72 ± 2.14 Min, median, max, kg/m2 18.1, 22.82, 34.9 14.5, 21.39, 36.0 17.6, 23.26, 27.9 19.0, 21.73, 24.1 Type of IC, n (%) Hunner-type 42 (100.0) 41 (100.0) 0 (0.0) 0 (0.0) Non-Hunner-type 0 (0.0) 0 (0.0) 7 (100.0) 6 (100.0) History of hydrodistension, n (%) No 6 (14.3) 8 (19.5) 0 (0.0) 1 (16.7) Yes 36 (85.7) 33 (80.5) 7 (100.0) 5 (83.3) History of intravesical DMSO, n (%) No 36 (85.7) 36 (87.8) 7 (100.0) 6 (100.0) Yes 6 (14.3) 5 (12.2) 0 (0.0) 0 (0.0) Use of drugs for primary disease at screening, n (%) No 7 (16.7) 15 (36.6) 2 (28.6) 1 (16.7) Yes 35 (83.3) 26 (63.4) 5 (71.4) 5 (83.3) ICSI score 9≤, <13, n (%) 14 (33.3) 15 (36.6) 3 (42.9) 2 (33.3) 13≤, <17, n (%) 18 (42.9) 16 (39.0) 2 (28.6) 3 (50.0) 17≤, ≤20, n (%) 10 (23.8) 10 (24.4) 2 (28.6) 1 (16.7) Mean ± SD 13.9 ± 3.2 13.8 ± 3.2 13.7 ± 3.1 13.2 ± 2.7 Min, median, max 9, 14.0, 20 9, 13.0, 20 11, 13.0, 19 9, 13.0, 17 ICPI score Mean ± SD 12.0 ± 3.1 11.9 ± 2.8 11.9 ± 4.2 11.3 ± 4.2 Min, median, max 4, 13.0, 16 6, 12.0, 16 7, 12.0, 16 6, 11.5, 16 Number of micturitions per 24 h Mean ± SD 15.67 ± 7.85 15.43 ± 5.92 20.86 ± 16.13 11.50 ± 1.95 Min, median, max 8.0, 14.50, 50.0 8.0, 13.50, 37.0 9.0, 16.50, 56.5 9.0, 11.75, 14.0 Voided volume per micturition, mL Mean ± SD 113.8 ± 52.9 110.5 ± 58.0 83.3 ± 63.0 138.6 ± 61.2 Min, median, max 27, 109.9, 257 31, 103.6, 305 38, 66.2, 217 80, 125.9, 248 Maximum voided volume per micturition <150 mL, n (%) 15 (35.7) 18 (43.9) 4 (57.1) 1 (16.7) ≥150 mL, n (%) 27 (64.3) 23 (56.1) 3 (42.9) 5 (83.3) Mean ± SD, mL 183.5 ± 101.3 172.4 ± 100.9 183.9 ± 129.7 265.8 ± 94.5 Min, median, max 50, 187.5, 600 45, 160.0, 500 90, 135.0, 465 125, 257.5, 415 NRS for bladder pain <7, n (%) 25 (59.5) 26 (63.4) 5 (71.4) 4 (66.7) ≥7, n (%) 17 (40.5) 15 (36.6) 2 (28.6) 2 (33.3) Mean ± SD 6.64 ± 1.46 6.54 ± 1.56 5.62 ± 1.21 6.28 ± 1.10 Min, median, max 4.0, 6.33, 10.0 4.0, 6.67, 10.0 4.3, 5.00, 7.0 5.0, 6.17, 8.0 Efficacy in Hunner-type IC patients

The change in the LS mean ICSI at Week 12 from baseline in Hunner-type IC patients was −5.6 in the KRP-116D group and −3.4 in the placebo group. The estimated difference between the KRP-116D and placebo groups was −2.2 (95% CI −3.8, −0.6; P = 0.0080 [Table 2]). Similar efficacies were shown in LS mean changes from baseline in ICPI and NRS for bladder pain at Week 12 and estimated differences between KRP-116D and placebo. The LS mean change at Week 12 from baseline, and estimated differences between KRP-116D and placebo in voiding variables including number of micturitions/24 h, voided volume/micturition (mL), and maximum voided volume/micturition (mL) were also statistically significant. The proportions of responders on GRA at Week 12 in the KRP-116D and placebo groups were 53.7% (22/41) and 27.5% (11/40), respectively. The difference in the proportion of responders between the KRP-116D and placebo groups was 26.2% (95% CI 4.8, 44.4; P = 0.0236).

Table 2. Changes in efficacy variables from baseline to Week 12 Hunner-type IC Non-Hunner-type IC

KRP-116D (n = 42)

Placebo (n = 41)

KRP-116D (n = 7)

Placebo (n = 6)

ICSI score No. of subjects at Week 12 41 39 7 6 LS mean change at Week 12 from baseline −5.6 −3.4 −2.9 −3.8 (95% CI) (−6.7, −4.4) (−4.5, −2.2) (−5.4, −0.3) (−6.5, −1.1) Estimated difference −2.2 0.9 (95% CI for estimated difference) (−3.8, −0.6) (−2.8, 4.7) P value 0.0080 0.6008 ICPI score No. of subjects at Week 12 41 39 7 6 LS mean change at Week 12 from baseline −5.3 −2.2 −2.2 −3.6 (95% CI) (−6.4, −4.2) (−3.3, −1.1) (−5.4, 1.0) (−7.0, −0.1) Estimated difference −3.1 1.3 (95% CI for estimated difference) (−4.7, −1.5) (−3.4, 6.0) P value 0.0002 0.5467 Number of micturitions per 24 h No. of subjects at Week 12 41 39 7 6 LS mean change at Week 12 from baseline −3.86 −1.27 −3.06 −5.76 (95% CI) (−5.01, −2.72) (−2.44, −0.10) (−4.99, −1.14) (−7.85, −3.67) Estimated difference −2.59 2.70 (95% CI for estimated difference) (−4.23, −0.96) (−0.26, 5.66) P value 0.0023 0.0688 Voided volume per micturition (mL) No. of subjects at Week 12 41 39 7 6 LS mean change at Week 12 from baseline 45.2 11.9 −5.1 35.3 (95% CI) (32.6, 57.9) (−0.9, 24.7) (−26.6, 16.4) (11.9, 58.7) Estimated difference 33.4 −40.4 (95% CI for estimated difference) (15.4, 51.4) (−73.3, −7.5) P value 0.0004 0.0203 Maximum voided volume per micturition (mL) No. of subjects at Week 12 41 39 7 6 LS mean change at Week 12 from baseline 52.3 7.1 −9.5 24.0 (95% CI) (36.0, 68.5) (−9.3, 23.5) (−54.9, 35.9) (−25.2, 73.1) Estimated difference 45.2 −33.4 (95% CI for estimated difference) (22.1, 68.2) (−101.6, 34.8) P value 0.0002 0.3048 NRS for bladder pain No. of subjects at Week 12 41 39 7 6 LS mean change at Week 12 from baseline −3.28 −2.05 −0.54 −2.26 (95% CI) (−3.99, −2.57) (−2.77, −1.34) (−1.91, 0.82) (−3.73, −0.78) Estimated difference −1.23 1.72 (95% CI of estimated difference) (−2.23, −0.22) (−0.31, 3.74) P value 0.0173 0.0896 GRA No. of subjects at Week 12 41 40 7 6 Responder 22 (53.7) 11 (27.5) 3 (42.9) 3 (50.0) Non-responder 19 (46.3) 29 (72.5) 4 (57.1) 3 (50.0) Difference 26.2 −7.1 (95% CI) (4.8, 44.4) (−48.5, 37.6) P value 0.0236 1.0000 Efficacy in non-Hunner-type IC patients

In non-Hunner-type IC patients, no statistical difference was found between the KRP-116D and placebo groups in LS mean changes at Week 12 from baseline in ICSI, ICPI, NRS for bladder pain, or voiding variables including number of micturitions/24 h and maximum voided volume/micturition (Table 2). The proportions of responders on GRA at Week 12 in the KRP-116D and placebo groups were 42.9% (3/7) and 50.0% (3/6), respectively. In addition, the statistical power of the t-test was not higher than 0.438 for any of the study endpoints due to the small number of non-Hunner-type IC patients in this study (KRP-116D, n = 7; placebo, n = 6 [data not shown]).

Correlations between efficacy variables and GRA, and ROC curves in patients with Hunner-type IC

At Week 12 after intravesical treatment with KRP-116D or placebo, Pearson correlation coefficients for GRA with ICSI, ICPI, and NRS for bladder pain were 0.676, 0.626, and 0.603, respectively; however, the Pearson correlation coefficients for voiding variables such as micturitions/24 h, voided volume/micturition (mL), and maximum voided volume/micturition (mL) were lower at 0.252, −0459, and −0.475, respectively (Fig. 1). Next, ROC curves for ICSI, ICPI and NRS for bladder pain were constructed (Fig. 2). The areas under the curve for ICSI, ICPI and NRS for bladder pain were 0.8789, 0.8350, and 0.8532, respectively. The cut-off values for clinically meaningful differences in ICSI, ICPI, and NRS for bladder pain were −5 (sensitivity 81.3%, specificity 83.3%), −3 (sensitivity 84.4%, specificity 70.8%), and −2.3 (sensitivity 84.4%, specificity 77.1%), respectively.

Correlations between GRA and (a) ICSI, (b) ICPI, (c) NRS for bladder pain, (d) number of micturitions/24 h, (e) voided volume/micturition (mL), and (f) maximum voided volume/micturition (mL) at Week 12 after intravesical treatment with KRP-116D or placebo. Each figure shows boxplot. r, correlation coefficient.

ROC curves for (a) ICSI, (b) ICPI, (c) NRS for bladder pain, (d) number of micturitions/24 h, (e) voided volume/micturition (mL), and (f) maximum voided volume/micturition (mL) at Week 12 after intravesical treatment with KRP-116D or placebo. The areas under the curves for ICSI, ICPI, NRS for bladder pain, number of micturitions/24 h, voided volume/micturition (mL), and maximum voided volume/micturition (mL) were 0.8789, 0.8350, 0.8532, 0.6947, 0.7484, and 0.7793, respectively.

Discussion

In the randomized placebo-controlled double-blind phase III study of KRP-116D, it was shown that there were clinical benefits for patients with the bladder-centric phenotype of IC/BPS including both Hunner- and non-Hunner-type IC, which were defined according to the 2009 Japanese IC guideline1 available at the time when the study was designed. In the updated 2019 guidelines,10 a new definition of IC/BPS was introduced, in which only Hunner-type IC is defined as IC while other types are categorized as BPS. Thus, in this post hoc analysis of the phase III study of KRP-116D, we performed separate evaluations for the effectiveness of KRP-116D in patients with Hunner-type IC, who represented 86.5% of enrolled patients, and those with non-Hunner-type IC with hydrodistension-induced glomerulations. As a result, better clinical outcomes were demonstrated in Hunner-type IC patients than in non-Hunner-type IC patients for subjective endpoints including ICSI, ICPI, and NRS for bladder pain, as well as for the objective endpoints including number of micturitions/24 h, voided volume/micturition, maximum voided volume/micturition. The proportion of responders to KRP-116D on GRA was 53.7%, compared to 27.5% with placebo in patients with Hunner-type IC. In contrast, KRP-116D was not found to be effective in patients with non-Hunner-type IC. This is in line with the findings of a previous study by Tomoe16 showing that the Hunner-type IC patient is a good responder to DMSO compared to the non-Hunner-type IC patient. However, it is also possible in our KRP-116D trial that the number of non-Hunner-type IC patients (n = 7 and n = 6 in the respective groups) was too small to provide adequate power for a meaningful statistical comparison between the placebo and KRP-116D groups. Previous studies have shown that non-Hunner-type IC/BPS patients have limited inflammatory cha