SUMMARY

THOR-2/BLC-2003 (NCT04172675) was a phase 2, randomized, open-label, multicenter, multi-cohort study that evaluated the efficacy and safety of BALVERSA compared with investigator’s choice of therapy in patients with recurrent high-risk non-muscle-invasive bladder cancer (HR-NMIBC) and fibroblast growth factor receptor (FGFR) mutations or fusions who received prior Bacillus Calmette-Guérin (BCG) therapy. The primary endpoint was recurrence-free survival (RFS).^1,2

Cohort 1 (n=73) included patients with recurrent BCG-treated, papillary-only HR-NMIBC and were randomized 2:1 to receive either BALVERSA or chemotherapy (mitomycin C or gemcitabine). Cohort 2 (n=16) included patients with BCG-unresponsive HR-NMIBC carcinoma in situ (CIS) who refused or were ineligible for cystectomy. Cohort 3 (n=18) included patients with histologically confirmed, recurrent, low-grade, intermediate-risk non-muscle-invasive bladder cancer (IR-NMIBC) with a <5% progression risk at 2 years and a >50% recurrence risk. Patients in cohorts 1-3 received BALVERSA 6 mg once daily in 28day cycles.³

• Catto et al (2024)¹  reported interim efficacy and safety outcomes from THOR-2 cohort 1. Median follow-up for RFS was 13.4 months for both treatment groups. Median RFS was not reached (NR) for BALVERSA (95% confidence interval [CI], 16.9 months-not estimable [NE]) and was 11.6 months (95% CI, 6.4-20.1) for chemotherapy, hazard ratio (HR) 0.28 (95% CI, 0.1-0.6). Adverse events (AEs) based on the known safety profile of BALVERSA included: nail disorders (78%), hyperphosphatemia (74%), eye disorders (excluding central serous retinopathy, 59%), skin disorders (51%), dry mouth (47%), mucositis (41%), and central serous retinopathy (39%). The most common any-grade AEs due to chemotherapy were UTIs (17%) and hematuria (17%).
• Daneshmand et al (2025)³ reported final analysis results of all 3 cohorts of the study.
  • For cohort 1, the median follow-up for RFS was 18.5 months and 16.6 months in the BALVERSA and chemotherapy treatment groups, respectively. Median RFS was NR (95% CI, 28.6 months-NE) for patients receiving BALVERSA and 11.6 months (95% CI, 5.3-NE) for patients receiving chemotherapy (HR, 0.28; 95% CI, 0.13-0.61).
  • For cohort 2, the median follow-up was 18.6 months. The complete response (CR) rate was 94% (15/16 evaluable patients; 95% CI, 70-100) at 8 weeks and 81% (13/16 evaluable patients; 95% CI, 54-96) at 32 weeks.
  • For cohort 3, the median follow-up was 16.6 months. The CR rate was 89% at the 3-month assessment (95% CI, 65-99); partial response (PR) was observed in 1 patient (6%). The median time to response was 1.2 months, and 1 patient (6%) had high-grade recurrence. In 17 responders, the median duration of response (DOR) was NR (95% CI, 13.4 months-NE), with 9 responses ongoing, 4 censored, and 4 ending with recurrence or progression.
  • Safety results were consistent with the known safety profile for BALVERSA. Hyperphosphatemia was the most common TRAE reported among the pooled BALVERSA cohorts (63/83 patients; 76%).

CLINICAL STUDY

• Assessments included cystoscopy for disease response, bladder mapping for new lesions, urine cytology for local assessment, and computed tomography (CT)/magnetic resonance imaging (MRI) urogram.³
  • For cohorts 1 and 2, cystoscopy was performed at screening, on cycle 3 day 1 (3 months), and then every 12 weeks (±1 week) for up to 2 years of treatment or until high-risk disease recurrence or progression.
  • For cohort 3, cystoscopy was performed at screening, then on cycle 2 day 1 (1 month ± 2 days), cycle 3 day 1, and cycle 4 day 1 (±1 week) or until CR, whichever occurred first.
• The study was terminated early due to poor accrual in cohort 1, which resulted in smaller sample sizes across all cohorts. Reasons included the COVID-19 pandemic, intermittent shortage of BCG leading to patients not receiving adequate BCG, limited tissue availability in the NMIBC population resulting in molecular testing challenges, and urologist/patient concerns about systemic toxicities.³

Catto et al (2024)¹ presented interim efficacy and safety outcomes from THOR-2 cohort 1 (n=73) for patients with recurrent, BCG-treated, papillary-only HR-NMIBC (high-grade Ta/T1) and select FGFR alterations who refused or were ineligible for radical cystectomy.

Patient Characteristics

• The demographics and baseline characteristics of patients in cohort 1 are included in Table: Cohort 1 Select Demographics and Baseline Characteristics.

Efficacy

• Median follow-up for RFS was 13.4 months in both BALVERSA and chemotherapy treatment groups.
• Efficacy outcomes from cohort 1 are summarized in Table: Cohort 1 RFS.
• At clinical cutoff, 25 RFS events had occurred (BALVERSA, n=11; chemotherapy, n=14).
• Median RFS was NR for patients receiving BALVERSA (95% CI, 16.9 months-NE) and 11.6 months (95% CI, 6.4-20.1) for patients receiving chemotherapy (HR, 0.28 [95% CI, 0.1-0.6]).
• The 6-month and 12-month RFS rates (95% CI) were 96% (83.7-98.9) and 77% (60‑87.4) for BALVERSA vs 73% (50.1-87.1) and 41% (18.9-61.7) for chemotherapy, respectively.
• RFS benefit for BALVERSA was generally consistent across subgroups of prior BCG therapy (experienced vs unresponsive) and tumor stage (Ta vs T1).
• Nine (38%) patients crossed over from chemotherapy to receive BALVERSA.

Safety

• The median duration of exposure was 9 months (range, 1-23.1) with BALVERSA and 6.4 months (range, 1.1-14.2) with chemotherapy.
• The most frequent grade ≥3 TRAEs were:
  • Stomatitis (10%), nail dystrophy (4%), and glossitis (4%) in the BALVERSA group.
  • Elevated alanine aminotransferase (4%) in the chemotherapy group.
• Serious AEs occurred in 11 (22%) and 3 (13%) patients in the BALVERSA and chemotherapy groups, respectively.
• No treatment-emergent adverse events (TEAEs) leading to death were reported.
  • After treatment discontinuation, 3 deaths occurred in the BALVERSA group (disease progression, n=2; secondary malignancy, n=1). All 3 were deemed unrelated to treatment.
• Any-cause AEs led to treatment discontinuation in 14 (29%) patients in the BALVERSA group (the most frequent AE leading to discontinuation was stomatitis [6%]) and no patients in the chemotherapy group.
• AEs based on the known safety profile of BALVERSA included:
  • Nail disorders (78%), hyperphosphatemia (74%), eye disorders (excluding central serous retinopathy, 59%), skin disorders (51%), dry mouth (47%), mucositis (41%), and central serous retinopathy (39%).
    • Most central serous retinopathy AEs were grade 1/2; 2 patients in the BALVERSA group reported grade 3 events (at clinical cutoff, events had resolved in 1 patient and were resolving in the other patient).
    • Central serous retinopathy AEs (central serous retinopathy, detachment of macular retinal pigment epithelium, maculopathy, and macular detachment) led to treatment discontinuation in 6 patients.
• Any-cause AEs are listed by preferred term and worst toxicity grade that was reported in >15% of the patients in either treatment group in Table: Cohort 1 TEAEs. For patients who crossed over from chemotherapy to BALVERSA treatment, this table summarizes AEs before crossover.

Daneshmand et al (2025)^3,4 presented the final efficacy and safety outcomes for cohort 1 (n=73), cohort 2 (n=16), and cohort 3 (n=18) of the THOR-2 study.

Patient Characteristics

Cohorts 1, 2, and 3

• The demographics and baseline characteristics of patients in the 3 cohorts are included in Table: Select Demographics and Baseline Characteristics.

Efficacy

Cohort 1

• Median follow-up for RFS was 18.5 months and 16.6 months in the BALVERSA and chemotherapy treatment groups, respectively.
• Efficacy outcomes from cohort 1, including subgroup analysis based on prior BCG therapy (experienced vs unresponsive) and tumor stage (Ta vs T1), are summarized in Table: Cohort 1 RFS.
• Median RFS was NR (95% CI, 28.6 months-NE) for patients receiving BALVERSA and 11.6 months (95% CI, 5.3-NE) for patients receiving chemotherapy (HR, 0.28; 95% CI, 0.13-0.61).

Cohort 2

• The median study follow-up was 18.6 months (n=16).
• The CR rate in disease-evaluable patients at 8 weeks was 94% (95% CI, 70-100; 15/16 patients).
• The CR rate in disease-evaluable patients at 32 weeks was 81% (95% CI, 54-96; 13/16 patients).
• The median DOR in 15 responders was 23.3 months (95% CI, 10.0-NE), with 9 responses ongoing at data cutoff, 5 with recurrence or progression after initial response, and 1 censored for consent withdrawal.

Cohort 3

• The median study follow-up was 16.6 months (n=18).
• The CR rate was 89% (95% CI, 65-99) at the 3-month assessment. One (6%) patient had PR, and 1 patient had a high-grade NMIBC.
  • The median time to response was 1.2 months (IQR, 1.0-1.8).
• In 17 responders, the median DOR was NR (95% CI, 13.4 months-NE), with 9 responses ongoing, 4 who experienced recurrence or progression, and 4 censored.

Safety

Cohorts 1, 2, and 3

• The median duration of exposure in cohort 1 was 12.8 months (IQR, 5.4-20.3) with BALVERSA and 6.2 months (IQR, 5.9-14.4) with chemotherapy.
• The median duration of exposure in cohort 2 and cohort 3 was 15.0 months (IQR, 5.4-18.7) and 12.0 months (IQR, 4.2-18.4), respectively.
• Safety results were consistent with the known safety profile for BALVERSA.
• The overall safety results for the THOR-2 study are reported in Table: Overall Safety Summary (All Cohorts).

• Hyperphosphatemia was the most common TRAE reported among the pooled BALVERSA cohorts (63/83 patients; 76%).
• Any-cause TRAEs are listed by preferred term and worst toxicity grade that was reported in ≥10% of the patients of all cohorts in Table: TRAEs (All Cohorts)

• FGFR inhibitor class-specific AEs in BALVERSA-treated patients were reported in 49 patients (100%) in cohort 1, 15 (94%) patients in cohort 2, and 18 (100%) patients in cohort 3. TEAEs are listed by preferred term and worst toxicity grade that was reported in >10% of the patients in any of the cohorts in Table: FGFR Inhibitor Class-Specific TEAEs (All Cohorts)

FGFR Inhibitor Class-Specific TEAEs (All Cohorts)⁴

• Chronic serous retinopathy (CSR) TEAEs were reported in 20 (41%), 4 (25%), and 3 (17%) patients in cohorts 1, 2, and 3, respectively, with grades 1-2 being the most reported maximum toxicity (18/20 [90%], 4/4 [100%], and 3/3 [100%], respectively).
  • CSR events resolved in 12/20 (60%) patients in cohort 1. For the 8 patients with ongoing unresolved CSR events, 3 were associated with treatment withdrawal and 5 were not resolved at data cutoff.
  • CSR events resolved in all patients in cohort 2 (4/4 [100%]).
  • Two CSR events remained unresolved at study end in cohort 3, 1 of which was due to treatment withdrawal.
• Visual acuity was measured in all patients treated with BALVERSA (n=83).
  • In the pooled cohorts, 22/56 patients reported worsening of visual acuity when comparing with baseline to worst post-baseline visual acuity data.
  • In the pooled cohorts with at least one visual acuity record after the occurrence of worst post-baseline visual acuity, 16/42 patients reported improvement since their last assessment.

Pharmacokinetics

Cohorts 1, 2, and 3

• BALVERSA predose plasma concentrations at cycle 2 day 1 for all cohorts are summarized in Table: BALVERSA Plasma Concentrations (All Cohorts)

BALVERSA Plasma Concentrations (All Cohorts)⁴

Literature Search

A literature search of MEDLINE^®, Embase^®, BIOSIS Previews^®, DERWENT Drug File (and/or other resources, including internal/external databases) was conducted on 03 November 2025.