Full enrollment results from cohort 1 of the phase III randomized THOR trial of erdafitinib versus chemotherapy in FGFR-altered advanced urothelial carcinoma

Abstract

Background

Cohort 1 of the phase III THOR trial met the primary endpoint of significantly longer overall survival (OS) with erdafitinib versus chemotherapy in patients with metastatic urothelial carcinoma (mUC) and select FGFR3/2 alterations who have progressed after prior anti–PD-(L)1 treatment. We present efficacy and safety results for the full cohort 1 at the end-of-study data cut-off (final database lock: 7 June 2024).

Patients and methods

Adults with mUC with susceptible FGFR3/2 alterations who had progression after 1-2 prior treatments that included an anti–PD-(L)1 were randomly assigned 1 : 1 to receive erdafitinib or investigator’s choice of chemotherapy (docetaxel or vinflunine). The primary endpoint was OS and progression-free survival (PFS) and objective response rate (ORR) were key secondary endpoints.

Results

At the end of data collection, median follow-up was 21.0 months for erdafitinib and 18.0 months for chemotherapy for the 278 patients in the fully enrolled THOR cohort 1 (erdafitinib, n = 143; chemotherapy, n = 135). Median OS remained significantly longer with erdafitinib versus chemotherapy [12.1 versus 8.7 months, hazard ratio (HR) 0.65, 95% confidence interval (CI) 0.48-0.86, P = 0.0031]. Erdafitinib also showed sustained improvements over chemotherapy in PFS (median 5.4 versus 2.7 months, HR 0.59, 95% CI 0.45-0.77, P < 0.0001) and ORR (46.9% versus 12.6%, relative risk 3.74, 95% CI 2.31-6.04, P < 0.001). Grade 3-4 treatment-emergent adverse events occurred in 67.6% of patients in the erdafitinib group and 65.8% in the chemotherapy group. Erdafitinib was associated with fewer treatment-related serious adverse events (AEs) (12.7% versus 24.8%), treatment-related deaths (0.7% versus 6.0%), and treatment-related AEs that led to discontinuations (9.2% versus 14.5%) compared with chemotherapy.

Conclusions

At the end of data collection, erdafitinib continued to demonstrate survival and clinical response benefit versus chemotherapy, consistent with the primary analysis. No new safety signals were observed. These mature data support the use of erdafitinib in patients with advanced/mUC and susceptible FGFR3 alterations after PD-(L)1 inhibition.

Highlights

• •THOR was a phase III study of erdafitinib versus chemotherapy in FGFR-altered advanced/mUC after prior anti–PD-(L)1.
• •OS, PFS, and ORR with erdafitinib versus chemotherapy were maintained in the fully enrolled THOR cohort 1.
• •Median OS, PFS, and ORR were 12.1 and 5.4 months and 46.9% for erdafitinib and 8.7 and 2.7 months and 12.6% for chemotherapy.
• •Erdafitinib demonstrated manageable tolerability, with no new safety signals at the end of data collection.
• •Full cohort 1 results support erdafitinib use in locally advanced/mUC and susceptible FGFR3 alterations after anti–PD-(L)1.

Introduction

Urothelial carcinoma remains challenging to manage, particularly in advanced or metastatic stages, due to its high recurrence rates, molecular heterogeneity, and limited durable responses to systemic therapy.^1,2 The standard of care for first-line treatment of locally advanced or metastatic urothelial carcinoma (mUC) is the combination of the antibody–drug conjugate enfortumab vedotin plus the anti–PD-1 inhibitor pembrolizumab, with platinum-based chemotherapy as an option where enfortumab vedotin plus pembrolizumab is not available or not indicated.³ However, treatment options for subsequent-line therapy remain limited, particularly for patients who have previously received immunotherapy. More than half of patients with advanced or mUC are ineligible for cisplatin treatment and those who receive chemotherapy experience disease progression within a few months, highlighting an unmet need for new treatment options.^4,5

Up to 20% of patients with advanced or mUC and up to 36% of patients with upper tract urothelial carcinomas harbor alterations in the fibroblast growth factor receptor (FGFR) gene, which may act as oncogenic drivers.6, 7, 8 Erdafitinib is an oral, selective pan-FGFR tyrosine kinase inhibitor that has demonstrated clinical activity in patients with alterations in the FGFR pathway, including urothelial cancer.⁹ In the phase II, open-label, single-arm study (BLC2001) of patients with locally advanced or mUC with susceptible FGFR3/2 alterations who had disease progression after platinum-containing chemotherapy, erdafitinib treatment resulted in an objective response rate (ORR) of 40%, meeting the primary endpoint.^9,10 The results of this trial led to accelerated approval of erdafitinib in the United States for the treatment of adult patients with locally advanced or mUC with susceptible FGFR3/2 alterations who have progressed after platinum-containing chemotherapy.¹¹

THOR is a confirmatory, randomized, phase III trial designed to evaluate the efficacy and safety of erdafitinib versus chemotherapy (docetaxel or vinflunine) and the PD-1 inhibitor pembrolizumab in patients with locally advanced or mUC and selected FGFR aberrations who had progressed on or after 1 or 2 prior treatments.¹² The study comprised two cohorts. Cohort 1 examined erdafitinib versus standard-of-care chemotherapy (investigator’s choice of docetaxel or vinflunine) after at least one line of treatment including an anti–PD-1 or anti–PD-L1 agent. Cohort 2 assessed erdafitinib versus pembrolizumab in patients who had not received a prior anti–PD-1 or anti–PD-L1 agent.^12,13 Cohort 2 results have been reported previously and provide additional context for the broader benefit–risk profile of erdafitinib.¹³ In the context of the evolving treatment landscape, erdafitinib represents an important guideline-supported targeted treatment option for patients with locally advanced or mUC and susceptible FGFR3 alterations following first-line enfortumab vedotin plus pembrolizumab.¹⁴

At the primary analysis (clinical cut-off of 15 January 2023), 266 patients had been randomly assigned in cohort 1 (erdafitinib, n = 136; chemotherapy, n = 130) of the THOR trial. The primary analysis demonstrated significantly longer overall survival (OS) with erdafitinib versus chemotherapy among patients with mUC and selected FGFR alterations after prior anti–PD-1 or anti–PD-L1 treatment and met the primary endpoint.¹² At a median follow-up of 15.9 months, median OS was 12.1 months with erdafitinib versus 7.8 months with chemotherapy [hazard ratio (HR) 0.64, 95% confidence interval (CI) 0.47-0.88, P = 0.005]. Progression-free survival (PFS) and ORR also favored erdafitinib over chemotherapy and adverse events (AEs) with erdafitinib were manageable.

Based on these results, erdafitinib was granted full approval in the United States and was also approved in Europe for the treatment of adult patients with unresectable or mUC harboring susceptible FGFR3 alterations whose disease progressed during or after at least one line of therapy containing a PD-1 or PD-L1 inhibitor in the unresectable or metastatic setting.^11,15 Erdafitinib is also a recommended treatment option in clinical practice guidelines for patients with selected FGFR3 alterations who have previously been treated with an immune-checkpoint inhibitor.^14,16,17 This report presents efficacy and safety results for all randomly assigned patients in cohort 1 at the end of data collection, defined as the date of the last observation recorded in the clinical database, representing the most mature follow-up for erdafitinib reported to date in this setting.

Patients and methods

Study design and participants

THOR (NCT03390504; EudraCT: 2017-002932-18) is a randomized, phase III, open-label trial conducted in 121 clinical sites across 23 countries. The study methodology for cohort 1 has been previously described.¹² In brief, adults with metastatic or surgically unresectable urothelial cancer and select FGFR3/2 alterations, Eastern Cooperative Oncology Group (ECOG) performance status score of 0, 1, or 2, adequate organ function, and progression during or after 1 or 2 prior treatments that included an anti–PD-1 or anti–PD-L1 agent, were eligible. Molecular eligibility was confirmed by central laboratory or local historical tests; tumors must have at least one of the following FGFR3 mutations: R248C, S249C, G370C, or Y373C; or one of the following fusions (translocations): FGFR2–BICC1, FGFR2–CASP7, FGFR3–TACC3_V1, FGFR3–TACC3_V3, or FGFR3–BAIAP2L1.

This study was conducted in accordance with the Declaration of Helsinki and all applicable laws and regulations. The ethics committees and institutional review boards at the participating centers approved the protocol and all patients provided informed consent.

Interventions

Patients were randomly assigned 1 : 1 to receive erdafitinib or chemotherapy using an interactive web response system. Randomization was carried out using randomly permuted blocks and was stratified by region (North America versus Europe versus the rest of the world), ECOG performance status (0 or 1 versus 2), and disease distribution [presence versus absence of visceral (lung, liver, or bone) metastases]. Erdafitinib was administered orally in 21-day cycles at a starting dose of 8 mg once daily, with a pharmacodynamically guided increase to 9 mg on day 14. Patients randomly assigned to chemotherapy received either vinflunine or docetaxel once every 3 weeks, as determined by the investigator. Treatment continued until disease progression or intolerable toxicity.

Endpoints and assessments

The primary endpoint was OS. Secondary endpoints included PFS, ORR, duration of response, and safety. PFS and ORR were assessed per RECIST version 1.1 by the investigator. Tumor response and progression were assessed every 6 weeks for the first 6 months and every 12 weeks for the next 6 months and beyond. AEs were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.03. Ophthalmologic examination at baseline included an Amsler grid test, optical coherence tomography (OCT), and ophthalmologic evaluation. Amsler grid test was carried out at every cycle and repeat OCT was conducted as clinically indicated based on the Amsler grid test or clinical assessment.

Statistical analysis

The statistical methodology for the primary and key secondary endpoints was previously reported.¹² In cohort 1, ∼280 patients were to be enrolled to provide at least 85% power to detect an HR for death of 0.65 for erdafitinib relative to chemotherapy, with a two-sided type I error level of 0.05. One interim analysis was planned at an information fraction of ∼65%; the prespecified number of events required to trigger this event-driven analysis was reached before enrollment in cohort 1 was complete. This analysis occurred at a clinical cut-off of 15 January 2023, before enrollment in cohort 1 was complete and was considered the primary analysis.¹² The present analysis includes the fully enrolled cohort 1 at the end of data collection, after all patients had completed study follow-up or discontinued participation. Given the group-sequential design of the trial and that the testing of primary and key secondary efficacy endpoints was achieved at the primary analysis with strong type I error control, all P values associated with the end of data collection analyses presented here are nominal. All tests were conducted at a two-sided alpha level of 0.05. OS and PFS were estimated using the Kaplan–Meier method; comparisons between arms were carried out with a log-rank test and HRs and associated 95% CIs were calculated based on a Cox proportional hazards model. ORR and estimate of the relative benefit and associated 95% CIs were computed using the Cochran–Mantel–Haenszel method. Efficacy endpoints were analyzed in the intention-to-treat population, which included all randomly assigned patients. Safety was analyzed in the safety population, which included all patients who received at least one dose of study treatment. The data presented here are based on a data cut-off of 15 April 2024 (date of last observation recorded as part of the clinical database), with a final database lock date of 7 June 2024.

Results

Patients and treatment

Between 6 August 2018 and 13 February 2023, a total of 278 patients were randomly assigned to receive erdafitinib (n = 143) or chemotherapy (n = 135); this full cohort 1 included 12 patients enrolled after the clinical cut-off date of the primary analysis (erdafitinib, n = 7; chemotherapy, n = 5; Supplementary Figure S1, available at https://doi.org/10.1016/j.esmoop.2026.108583). Among these 12 patients, 11 had samples submitted for central molecular testing and all 11 patients had at least one selected FGFR alteration (Supplementary Table S1, available at https://doi.org/10.1016/j.esmoop.2026.108583).

Baseline demographics and disease characteristics were balanced between the treatment arms and consistent with those reported in the primary analysis (Supplementary Table S2, available at https://doi.org/10.1016/j.esmoop.2026.108583). Overall, 28.4% of patients in cohort 1 had received one line of prior systemic therapy and 71.2% had received two lines of prior systemic therapy. With the exception of three patients who had been incorrectly assigned, all patients had received prior treatment with anti–PD-1 or anti–PD-L1 therapy (Supplementary Table S3, available at https://doi.org/10.1016/j.esmoop.2026.108583). More than half of patients (56.6% in the erdafitinib group and 59.3% in the chemotherapy group) had received an anti–PD-(L)1 agent as a single agent for second-line therapy. Most patients (89.1% overall) had also received prior chemotherapy (90.9% in the erdafitinib group and 88.1% in the chemotherapy group).

At the end of data collection, all patients had discontinued study treatment, most commonly because of progressive disease [91 of 143 (64.1%) in the erdafitinib group and 69 of 135 (59.0%) in the chemotherapy group].

Efficacy

In the fully enrolled cohort 1 at the end of data collection, the median follow-up duration for OS was 21.0 months in the erdafitinib group and 18.0 months in the chemotherapy group, compared with 15.9 months at the primary analysis.¹² Among 278 randomly assigned patients, efficacy and safety findings were consistent with the primary analysis, confirming the survival and response benefits of erdafitinib. Median OS was 12.1 months (95% CI 10.3-16.4) in the erdafitinib group and 8.7 months (95% CI 6.7-10.7) in the chemotherapy group, corresponding to an HR of 0.65 (95% CI 0.48-0.86, P = 0.0031, Figure 1). The 6- and 12-month estimated survival rates were 83% (95% CI 76-89) and 51% (95% CI 42-59), respectively, in the erdafitinib group and 67% (95% CI 58-74) and 38% (95% CI 29-47) in the chemotherapy group. The six patients (4.4%) in the chemotherapy group who crossed over to erdafitinib during the long-term extension were not reassigned to the erdafitinib group for efficacy analysis. Their postcrossover data were not included in the OS endpoint and they were censored at their last clinical database visit.

Figure 1.

OS.

CI, confidence interval; OS, overall survival.

The median PFS was 5.4 months (95% CI 4.4-5.6) in the erdafitinib group and 2.7 months (95% CI 2.1-4.2) in the chemotherapy group, with an HR of 0.59 (95% CI 0.45-0.77, P < 0.0001, Figure 2). Following the primary analysis, five patients in the erdafitinib group and two patients in the chemotherapy group achieved an objective response (including a complete response in one patient in the erdafitinib group). At the end of data collection, ORR by investigator assessment was 46.9% in the erdafitinib group and 12.6% in the chemotherapy group, corresponding to a relative risk of 3.74 (95% CI 2.31-6.04, P < 0.001; Table 1), favoring erdafitinib. The disease control rate (DCR) was 83.9% (95% CI 76.9%-89.5%) with erdafitinib and 46.7% (95% CI 38.0%-55.4%) with chemotherapy (relative risk, 1.81; 95% CI 1.49-2.19; P < 0.001). Median duration of response was 4.9 months (95% CI 3.8-6.9) in the erdafitinib group and 5.6 months (95% CI 2.8-6.0) in the chemotherapy group. Analysis of confirmed tumor responses by investigator (based on ≥2 consecutive assessments) also showed higher ORR [39.2% versus 10.4% (relative risk 3.80, 95% CI 2.22-6.50)] and DCR [79.0% versus 43.0% (relative risk 1.85, 95% CI 1.50-2.28)] with erdafitinib versus chemotherapy (Supplementary Table S4, available at https://doi.org/10.1016/j.esmoop.2026.108583).

Figure 2.

PFS per RECIST v1.1 by investigator assessment.

CI, confidence interval; PFS, progression-free survival.

Table 1.

Tumor responses per RECIST v1.1 by investigator

Tumor response	Erdafitinib (n = 143)	Chemotherapy (n = 135)
ORR, % (95% CI)	46.9 (38.5-55.4)	12.6 (7.5-19.4)
Relative benefit (95% CI)a	3.74 (2.31-6.04); P < 0.001
Best overall response, n (%)
Complete response	10 (7.0)	1 (0.7)
Partial response	57 (39.9)	16 (11.9)
Stable diseaseb	53 (37.1)	46 (34.1)
Progressive disease	16 (11.2)	33 (24.4)
Not evaluable	7 (4.9)	39 (28.9)
DCR, % (95% CI)	83.9 (76.9-89.5)	46.7 (38.0-55.4)
Relative benefit (95% CI)a	1.81 (1.49-2.19); P < 0.001
Duration of response, median (95% CI), months	4.9 (3.8-6.9)	5.6 (2.8-6.0)

CI, confidence interval; DCR, disease control rate; ECOG, Eastern Cooperative Oncology Group; ORR, objective response rate.

^aEstimated using the Cochran–Mantel–Haenszel test with ECOG performance status (0 or 1 versus 2) as a stratification factor.

^bIncludes patients with no measurable disease at baseline and whose best response was noncomplete response/nonprogressive disease. Minimum duration requirement for stable disease was 6 weeks from the date of randomization.

A total of 54 patients (37.8%) in the erdafitinib group and 53 patients (39.3%) in the chemotherapy group received subsequent anticancer therapy (Supplementary Table S5, available at https://doi.org/10.1016/j.esmoop.2026.108583). The most common were antibody–drug conjugates (18.9% in the erdafitinib group and 11.9% in the chemotherapy group), chemotherapy (18.2% and 16.3%, respectively), and immunotherapy (7.0% and 6.7%, respectively).

Safety

The safety population included 142 patients in the erdafitinib group and 117 patients in the chemotherapy group who received at least one dose of study treatment. Median (range) duration of exposure was 4.9 (0-42) months with erdafitinib and 1.4 (0-31) months with chemotherapy. In the erdafitinib group, 110 of 142 (77.5%) patients uptitrated from 8 to 9 mg and 66 of 142 (46.5%) maintained a dose ≥8 mg. Overall, dose reductions due to treatment-emergent adverse events (TEAEs) were required in 74 of 142 (52.1%) patients in the erdafitinib group (defined as a reduction to a dose below 8 mg); 2 patients had a dose reduction of erdafitinib for other reasons (Supplementary Table S6, available at https://doi.org/10.1016/j.esmoop.2026.108583). In the chemotherapy group, TEAEs leading to dose reductions occurred in 30 of 117 patients (25.6%). TEAEs led to treatment discontinuation in 15.5% with erdafitinib and 18.8% with chemotherapy (Supplementary Table S7, available at https://doi.org/10.1016/j.esmoop.2026.108583), and the AEs were considered related to study treatment in 9.2% and 14.5%, respectively.

TEAEs occurred in all patients who received erdafitinib and in 97.4% of patients who received chemotherapy (Table 2). Most TEAEs occurred within the first 90 days after erdafitinib initiation. Late-onset TEAEs, defined as events first reported beyond 12 months, occurred in 24 of 142 patients (16.9%), with no new safety signals identified. The most common TEAEs of any grade were hyperphosphatemia (82.4%), diarrhea (63.4%), stomatitis (52.1%), and dry mouth (43.0%) in the erdafitinib group, and anemia (33.3%), constipation (27.4%), asthenia (24.8%), and nausea (23.9%) in the chemotherapy group (Supplementary Table S8, available at https://doi.org/10.1016/j.esmoop.2026.108583). Grade 3 or 4 TEAEs occurred in 67.6% in the erdafitinib group and 65.8% in the chemotherapy group; the most common were palmar-plantar erythrodysesthesia syndrome (9.9%), stomatitis (8.5%), anemia (7.7%), hyponatremia (7.0%), and onycholysis (6.3%) with erdafitinib and neutropenia (15.4%), leukopenia (10.3%), febrile neutropenia (8.5%), and anemia (7.7%) with chemotherapy. Treatment-related AEs were reported in 98.6% of patients in the erdafitinib group and 87.2% in the chemotherapy group (Supplementary Table S9, available at https://doi.org/10.1016/j.esmoop.2026.108583).

Table 2.

Safety summary

AEs—no. (%)	Erdafitinib (n = 142)	Chemotherapy (n = 117)
TEAEs	142 (100)	114 (97.4)
Treatment-related	140 (98.6)	102 (87.2)
TEAEs leading to dose interruption	106 (74.6)	37 (31.6)
Treatment-related	99 (69.7)	24 (20.5)
TEAEs leading to dose reductiona	74 (52.1)	30 (25.6)
Treatment-related	72 (50.7)	26 (22.2)
TEAEs leading to treatment discontinuation	22 (15.5)	22 (18.8)
Treatment-related	13 (9.2)	17 (14.5)
Grade 3-4 TEAEs	96 (67.6)	77 (65.8)
Treatment-related	66 (46.5)	55 (47.0)
Serious TEAEs	63 (44.4)	50 (42.7)
Treatment-related	18 (12.7)	29 (24.8)
TEAEs leading to death	6 (4.2)	8 (6.8)
Treatment-related	1 (0.7)	7 (6.0)

AEs, adverse events; TEAEs, treatment-emergent adverse events.

^aDose reductions in the erdafitinib group are defined as any reduction to below the 8 mg dose (de-escalation from 9 to 8 mg are not included).

Treatment-emergent serious AEs occurred in 44.4% of patients in the erdafitinib group and 42.7% in the chemotherapy group (Supplementary Table S10, available at https://doi.org/10.1016/j.esmoop.2026.108583). The most common treatment-emergent serious AEs observed in ≥2% of patients in either treatment group were urinary tract infection (4.9% in the erdafitinib group versus 2.6% in the chemotherapy group), hematuria (3.5% versus 0.9%), febrile neutropenia (0% versus 6.8%), and febrile bone marrow aplasia (0% versus 3.4%). Treatment-related serious AEs occurred in 12.7% in the erdafitinib group and 24.8% in the chemotherapy group. In total, TEAEs led to death in six patients (4.2%, including one treatment-related death) in the erdafitinib group and in eight patients (6.8%, including seven treatment-related deaths) in the chemotherapy group. Since the primary analysis, only one additional death due to TEAEs was observed in the chemotherapy group.

Grade 3-4 TEAEs of interest based on the known safety profile of erdafitinib included nail disorders (12.0%), skin disorders (12.0%), central serous retinopathy (2.1%), and other eye disorders (2.1%) (Supplementary Table S11, available at https://doi.org/10.1016/j.esmoop.2026.108583). Since the primary analysis, no additional patients in the erdafitinib group experienced any events of central serous retinopathy of grades 3-4.

Discussion

In the fully enrolled THOR cohort 1 at the end of data collection, erdafitinib maintained a 35% reduction in the risk of death compared with chemotherapy in patients with locally advanced or mUC harboring FGFR3 alterations after previous treatment with anti–PD-(L)1 agents, consistent with the primary analysis.¹² At data cut-off, no patients remained on study treatment, supporting the maturity of follow-up. Erdafitinib also continued to demonstrate significantly longer OS and PFS and higher ORR compared with chemotherapy. The safety profile of erdafitinib remained consistent with the primary analysis and the known safety profile of erdafitinib, with no new safety signals. Although treatment-related AEs were more frequent with erdafitinib than with chemotherapy, the incidence of grade ≥3 TEAEs was similar between treatment groups, whereas incidences of discontinuations due to treatment-related AEs, treatment-related serious AEs, and treatment-related deaths were less frequent with erdafitinib than chemotherapy.

Based on the positive results of the primary analysis, erdafitinib received approval from the European Medicines Agency for the treatment of adult patients with unresectable or mUC, harboring susceptible FGFR3 genetic alterations whose disease progressed during or following at least one line of therapy containing a PD-1 or PD-L1 inhibitor in the locally advanced unresectable or metastatic setting.¹⁵ Erdafitinib was also granted full approval by the United States Food and Drug Administration for adult patients with locally advanced or mUC with susceptible FGFR3 genetic alterations whose disease has progressed on or after at least one line of prior systemic therapy (not recommended for patients who are eligible for and have not received prior PD-1 or PD-L1 inhibitor therapy).¹¹ The efficacy and safety findings from the fully enrolled cohort 1 of THOR support the use of erdafitinib in the subsequent-line, post-immune checkpoint inhibitor setting for patients with advanced or mUC and susceptible FGFR3 alterations.

The therapeutic landscape for advanced urothelial carcinoma has evolved rapidly since the initial THOR analysis, with the combination of enfortumab vedotin plus pembrolizumab recently established as the first-line standard of care for cisplatin-eligible and -ineligible patients, irrespective of PD-L1 expression.³ In this new paradigm, most patients will have received both an immune-checkpoint inhibitor and an antibody–drug conjugate by the time of disease progression. Consequently, the number of active agents available for subsequent therapy is limited. In this setting, erdafitinib remains one of the few options with demonstrated survival benefit and a distinct mechanism of action, providing an important alternative to chemotherapy for patients whose tumors harbor susceptible FGFR3 alterations. The updated results from cohort 1 of the THOR trial reinforce the durable efficacy and manageable safety of erdafitinib, supporting its use after progression on immunotherapy in patients with mUC and selected FGFR3 alterations. These data also highlight the critical importance of routine molecular testing to identify FGFR3 alterations early in the disease course, enabling timely integration of erdafitinib in personalized treatment sequencing. This is in line with clinical practice guidelines, which recommend molecular/genomic testing to be conducted early, ideally at the time of diagnosis of advanced urothelial cancer to facilitate clinical decision-making.^14,16,17

Although most FGFR testing was carried out on primary tumor samples, testing of either primary or metastatic tumors was permitted. This may introduce discordance because FGFR3 alteration test results may differ between archival primary lesions and metastatic lesions due to spatial heterogeneity and temporal genomic evolution, potentially affecting the identification of patients most likely to benefit from FGFR-directed therapy.^18,19

In the fully enrolled THOR cohort 1, the significant improvements in OS, PFS, and ORR observed with erdafitinib compared with chemotherapy in the primary analysis were maintained at the end of data collection, with no new safety signals identified. These data support the use of erdafitinib in patients with advanced or mUC harboring susceptible FGFR3 alterations following progression on PD-(L)1 inhibitors and underscore the importance of routine molecular profiling to identify actionable alterations including FGFR3 that may inform treatment selection.