Introduction:

Patients with high-risk non-muscle–invasive bladder cancer (HR-NMIBC) presenting as carcinoma in situ (CIS) have high risk of recurrence/progression. Treatment options for patients with bacillus Calmette-Guérin (BCG)-unresponsive CIS who refuse/are ineligible for radical cystectomy include pembrolizumab, intravesical chemotherapy, or nadofaragene firadenovec. Limited information exists regarding incidence and oncogenic role of FGFR alterations in CIS. Erdafitinib is an oral selective pan-FGFR tyrosine kinase inhibitor FDA-approved to treat locally advanced/metastatic urothelial cancer in adults with FGFR3/2 alterations who have progressed during or following ≥1 line of platinum-containing chemotherapy. THOR-2 (NCT04172675) is a phase 2 study of oral erdafitinib in patients with HR-NMIBC. We report updated data of the exploratory cohort of patients with BCG-unresponsive CIS harboring FGFR alterations with/without papillary disease. At study onset, it was unknown whether FGFR alterations were present in CIS. This represents the first prospective study of an FGFR inhibitor in this population.

Methods:

Patients were aged ≥18 years with histologically confirmed BCG-unresponsive HR-NMIBC presenting as CIS with FGFR3/2 alterations with or without a papillary tumor and either refused or were ineligible for cystectomy. BCG-unresponsive was defined as having persistent/recurrent CIS alone or with recurrent Ta/T1 or recurrent high-grade Ta/T1 disease ≤6 months after adequate BCG therapy, or high-grade T1 at the first disease assessment after BCG therapy. Patients received  oral erdafitinib at 6 mg once daily without uptitration in 28-day cycles (dose selected to improve tolerability). Complete response (CR) was defined as having a negative cystoscopy and negative (including atypical) urine cytology. If no CR was observed within 3 months, erdafitinib was discontinued. The exploratory end points were CR rate at 8 weeks and 32 weeks. Safety was a key secondary end point.

Results:

At data cutoff (27-Jun-2023; median study follow-up 14.8 months), 16 patients (median age: 68.5 years [range, 35-83]; tumor stage: 100% CIS) received treatment for median duration 6.7 months (range, 1.9-18.6). CR rate in disease-evaluable patients was 93.8% and 72.7% at 8 and 32 weeks, respectively (Table). Median duration of response was not reached, with 12 responses ongoing (Figure). Common treatment-emergent adverse events (TEAEs) were mostly grade 1-2: dry mouth (56.3%; n=9), hyperphosphatemia (56.3%; n=9), diarrhea (56.3%; n=9), stomatitis, dysgeusia, and dry skin (43.8%; n=7 each). One (6.3%) patient reported grade 2 treatment-related retinal detachment which led to treatment discontinuation and 1 (6.3%) had grade 1 treatment-related subretinal fluid; both resolved. Six (37.5%) patients had grade ≥3 treatment-related TEAEs (onychomadesis, acute kidney injury, alanine aminotransferase increased, chronic kidney disease, dry mouth, hypotension, nail disorders, stomatitis, and sepsis). One (6.3%) patient discontinued treatment due to a treatment-related TEAE. No treatment-related deaths occurred.

Conclusion:

Data from this trial provide first evidence of occurrence of FGFR alterations in CIS and demonstrate the promising and durable efficacy of an FGFR inhibitor such as erdafitinib in patients with BCG-unresponsive CIS with FGFR alterations. Safety data were consistent with the known safety profile of erdafitinib.

Funding: Janssen Research & Development