Abstract
Background
Enfortumab vedotin (EV), an innovative antibody–drug conjugate targeting Nectin-4, has emerged as a breakthrough therapy for locally advanced or metastatic urothelial carcinoma (la/mUC). However, EV-related cutaneous toxicity (EVRCT) remains a significant concern because of Nectin-4 expression in skin tissue. This study aimed to understand the incidence, risk factors, and clinical implications of EVRCT, focusing on relationships with treatment efficacy, using one of the largest real-world datasets from a Japanese la/mUC patient cohort.
Materials and methods
Data from 207 la/mUC patients treated with EV, mainly as a third-line therapy between 2020 and 2023, were analyzed. Multivariate logistic regression and propensity score matching (PSM) were used to assess the risk factors and impact of EVRCTs on patient overall survival (OS) and progression-free survival (PFS).
Results
EVRCTs were observed in 59.9% of patients, with 83% occurring within the first 3 months, defined as early-phase EVRCTs (epEVRCTs). Multivariate analysis identified better Eastern Cooperative Oncology Group performance status (ECOG PS = 0), higher hemoglobin levels (≥11 g/dl), and the standard initial dose (1.25 mg/kg) as significant risk factors for epEVRCTs. Patients with epEVRCTs demonstrated significantly improved PFS and OS compared with those without. Post-PSM analysis confirmed longer OS for patients with epEVRCTs, particularly those with mild (grade 1) toxicities, suggesting that these reactions may be significantly linked to favorable treatment outcomes.
Conclusions
Our data suggest that epEVRCTs are common and correlate with better clinical outcomes in la/mUC patients treated with EV, underscoring the importance of proactive EVRCT management to optimize therapeutic benefits.
Key words: EV-related cutaneous toxicity, incidence, risk factor, clinical outcome
Highlights
-
•
EVRCTs occurred in 59.9% of patients, mostly in 3 months.
-
•
Better ECOG PS, high hemoglobin, and standard EV dose raise risk of early EVRCTs.
-
•
Early EVRCT patients had better PFS and OS than those without EVRCTs.
-
•
Mild grade 1 EVRCTs were linked to the best OS rates.
-
•
Early EVRCTs may reflect effective EV therapy, suggesting close monitoring.
Introduction
Enfortumab vedotin (EV), an innovative antibody–drug conjugate targeting Nectin-4, has garnered significant attention as a breakthrough therapy for patients with locally advanced or metastatic urothelial carcinoma (la/mUC) since its approval in 2021.1 Initially used as a third-line treatment for patients who did not respond to chemotherapy and immune checkpoint inhibitors, EV applications have expanded into first-line therapy when used in combination with pembrolizumab, as evidenced by promising results of the EV302 trial.2
A notable adverse event with this treatment is EV-related cutaneous toxicity (EVRCT), a consequence of Nectin-4 expression in skin tissues.3 EVRCT encompasses a spectrum of dermatological reactions, including various rashes, pruritus, and, in rare instances, severe conditions such as Stevens–Johnson syndrome.4 These toxicities typically manifest early in the treatment cycle3,5 Low-grade toxicity is generally manageable with supportive care measures like antihistamines, topical steroids, and moisturizers. However, grade 3 toxicities require more intensive interventions, such as treatment interruption, pulse oral steroid administration, and dermatologist consultation. If a grade 3 event diminishes to grade 1, then re-administration at a reduced dose is potentially feasible. However, grade 4 toxicities warrant permanent discontinuation of EV treatment.
Despite the clinical significance of EVRCT, comprehensive data on its frequency and onset in real-world settings remain sparse. Additionally, the specific risk factors contributing to its development are not well understood.6,7 Research by Vlachou et al. suggested correlations between EVRCT incidence and patient body weight, which can influence EV dosage.6 However, these findings require validation across diverse populations, particularly in Asian cohorts with varying body sizes. Moreover, the study indicated a potential link between EVRCT occurrence and EV treatment efficacy, necessitating a careful analysis to address the immortal time bias that arises with prolonged treatment duration. Vlachou et al. recently published a follow-up report indicating that EVRCT is associated with overall survival (OS), utilizing a time-dependent Cox model to address the potential for immortal time bias.8 This finding highlights the importance of validation in external cohorts.
In this study, we aimed to clarify the relationship between EVRCT incidence and patient characteristics using extensive real-world data from Japan, striving to present an as-unbiased-as-possible analysis of EVRCT incidence and its impact on patient outcomes. By incorporating landmark analysis to mitigate immortal time bias, our findings contribute meaningful insights to optimize patient management and therapeutic strategies involving EV.
Materials and Methods
Study design
Clinical data were collected from the Japan Urological Oncology Group nationwide cohort of la/mUC patients. Briefly, this cohort comprised 207 patients who were diagnosed with la/mUC between January 2020 and December 2021 and received EV treatment by December 2023 at 34 participating institutions (flow chart in Supplementary Figure S1, available at https://doi.org/10.1016/j.esmorw.2024.100094). All participating institutions were either university hospitals or large community-based hospitals that treated >10 la/mUC cases per year. Data were entered by physicians at each institution via electronic case report form. This dataset included clinical background information, sequential treatment details, laboratory data, details of EVRCTs and intervention methods, and treatment outcomes of the patients. Survival data were collected in December 2023. Early-phase EVRCTs (epEVRCTs) were defined as skin adverse events that occurred within 3 months of the start of EV administration. The EVRCT severity was graded according to the Common Terminology Criteria for Adverse Events version 5.0. There were no restrictions on the choice of treatment, which was in accordance with the policies at each institution. This study was approved by the institutional review board at the main study institution (approval number R4223) and by the local institutional review board at each participating institution. This study conformed to the provisions of the Declaration of Helsinki.
The following parameters were evaluated in this study: patient age, sex, primary cancer site, variant histology, site of metastasis at initiation of EV treatment, prior treatment, Eastern Cooperative Oncology Group performance status (ECOG PS), hemoglobin (Hb) level, neutrophil-to-lymphocyte ratio (NLR), creatinine clearance (CCr), body weight, initial dose of EV, number of EV cycles, and use of prophylactic moisturizers. CCr was calculated using the Cockcroft–Gault formula, as follows: CCr = {[(140 − age) × weight]/[72 × serum creatinine]} × 0.85 (if female).
Statistical analysis
All statistical analyses were carried out using JMP Pro, version 15.1.0 (SAS Institute Inc., Cary, NC) and EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). OS and progression-free survival (PFS) were calculated from the date of EV initiation and were estimated by Kaplan–Meier analysis using the log-rank test. Multivariate logistic regression analysis was carried out to identify the risk factors in patients who develop epEVRCTs. The cut-off values for Hb levels and the NLR were in accordance with those in a previous report.9
Landmark analysis was conducted, with patients who survived >3 months included in the evaluation to minimize the immortal time bias. We used a propensity score matching (PSM) method to achieve the between-group comparability of the patients with and without epEVRCTs. PSM was carried out using the logistic regression model. Covariates, including age, sex, primary tumor site, variant histology, presence of liver metastasis, Hb, ECOG PS (0, 1, or >2), and initial dose of EV, were converted into propensity scores. As a sensitivity analysis, PFS and OS with and without epEVRCTs were compared using multivariate Cox proportional hazards model, with EVRCT as a time-dependent variable.
All statistical tests were two-sided, with P < 0.05 being considered statistically significant.
Results
EVRCT characteristics
In our cohort of 207 patients, 124 (59.9%) experienced EVRCTs, with the severity of these events distributed as follows: grade 1 in 20.7% of all patients, grade 2 in 31.8%, grade 3 in 6.2%, and grade 5 in 1.4% (Table 1). The most common manifestations included pruritus (43.9%), eczema (23.1%), maculopapular rash (15.9%), dry skin (14.4%), bullous dermatitis (6.6%), and severe conditions such as Stevens–Johnson syndrome/toxic epidermal necrolysis (1.4%). The median EVRCT onset time was 14 days, ranging from the day of the first dose to 730 days thereafter. Notably, 103 patients (49.7%) experienced epEVRCTs within the first 3 months of starting EV treatment, which accounted for 83% of all EVRCTs (Figure 1A). In particular, the incidence of grade 2 or higher EVRCTs was concentrated in the first month after treatment initiation. This early-onset trend was observed regardless of the EVRCT type (Figure 1B). These toxicities occurred mainly in the lower limbs (30.9%), upper limbs (28.5%), and trunk (26.6%) within 2 weeks of EV initiation, but were less common on the peripheral side, including the palms (3.3%) and feet (4.0%) (Figure 1C, Supplementary Figure S2, available at https://doi.org/10.1016/j.esmorw.2024.100094).
Table 1.
Characteristics of EV-related cutaneous toxicity (n = 207)
| Any skin reaction, n (%) | 124 (59.9) |
| Preferred term, n (%) | |
| Pruritus | 91 (43.9) |
| Eczema | 48 (23.1) |
| Rash maculopapular | 33 (15.9) |
| Dry skin | 30 (14.4) |
| Bullous dermatitis | 14 (6.7) |
| SJS/TEN | 3 (1.4) |
| Maximum grade, n (%) | |
| None | 82 (39.6) |
| G1 | 43 (20.7) |
| G2 | 66 (31.8) |
| G3 | 13 (6.2) |
| G5 | 3 (1.4) |
| Site, n (%) | |
| Face and neck | 23 (11.1) |
| Upper limb | 59 (28.5) |
| Axilla | 18 (8.6) |
| Chest and abdomen | 55 (26.6) |
| Groin | 16 (7.7) |
| Back | 36 (17.3) |
| Lower limb | 64 (30.9) |
| Palmar | 7 (3.3) |
| Foot | 9 (4.3) |
| Median time to development, days (range) | 14 (0-730) |
| Management | |
| Topical corticosteroids | 75 (36.2) |
| Antihistamines | 62 (29.9) |
| Oral corticosteroids | 16 (7.7) |
| Dose interruption | 37 (17.9) |
| Dose reduction | 25 (12.0) |
| Treatment withdrawal | 8 (3.8) |
EV, enfortumab vedotin; G, grade; SJS, Stevens–Johnson syndrome; TEN, toxic epidermal necrolysis.
Figure 1.
Enfortumab vedotin (EV)-related cutaneous toxicity (EVRCT) characteristics. (A and B) Kaplan–Meier plots displaying the cumulative incidence of EVRCT development for the entire cohort of 207 patients by maximum grade (A) and EVRCT type (B). (C) EVRCT onset time and occurrence by site. For each site, the overall size of the pie chart represents the frequency and the time of onset breakdown is represented by the area of the pie chart. The area of the legend pie chart in the lower right represents the time of onset breakdown for the whole body. SJS, Stevens–Johnson syndrome; TEN, toxic epidermal necrolysis.
EVRCT management involved the use of topical corticosteroids in 36.2% of cases and systemic corticosteroids in 7.7%. Additionally, 17.9% of patients required an interruption in EV treatment, 12.0% had a dose reduction, and 3.8% discontinued the treatment entirely.
Characteristics of patients with and without EVRCTs
Table 2 summarizes the characteristics of patients with and without EVRCTs. Those who developed EVRCTs generally had a better ECOG PS (ECOG PS = 0), lower incidence of low Hb levels, were more likely to be initially treated with the standard 1.25 mg/kg dose, and received a greater number of treatment cycles. Prophylactic topical application of moisturizers did not correlate with EVRCT frequency. Multivariate analysis identified that an ECOG PS of 0, Hb ≥11 g/dl, and an initial dose of 1.25 mg/kg were significant risk factors for epEVRCTs (Table 3).
Table 2.
Characteristics of patients with and without EV-related cutaneous toxicity (n = 207)
| Characteristics | Entire | Patients without cutaneous toxicity (n = 83) | Patients with cutaneous toxicity (n = 124) | P value |
|---|---|---|---|---|
| Age, years, average (range) | 71.0 (41-87) | 71.2 (47-87) | 70.8 (41-86) | 0.9971a |
| Sex, n (%) | 0.6352b | |||
| Female | 151 (72.9) | 24 (28.9) | 32 (25.8) | |
| Male | 56 (27.1) | 59 (71.1) | 92 (74.2) | |
| Primary site, n (%) | 0.4712c | |||
| Bladder | 110 (53.1) | 44 (53.0) | 66 (53.2) | |
| Upper urinary tract | 96 (46.4) | 38 (45.8) | 58 (46.8) | |
| Variant histology, n (%) | 0.8505b | |||
| Yes | 34 (16.4) | 13 (15.7) | 21 (16.9) | |
| No | 173 (83.6) | 70 (84.3) | 103 (83.1) | |
| Metastatic sites, n (%) | ||||
| Lung | 80 (38.6) | 29 (34.9) | 51 (41.1) | 0.3862b |
| Liver | 49 (23.7) | 20 (24.1) | 29 (23.4) | 1.000b |
| Bone | 55 (26.6) | 22 (26.5) | 33 (26.6) | 1.000b |
| Immediate prior treatment, n (%) | 0.3345d | |||
| Pembrolizumab | 123 (59.4) | 49 (59.0) | 74 (59.7) | |
| Avelumab | 42 (20.3) | 13 (15.7) | 29 (23.4) | |
| Chemotherapy | 38 (18.4) | 19 (22.9) | 19 (15.3) | |
| Others | 3 (1.5) | 1 (0.5) | 2 (1.0) | |
| Unknown | 1 (0.5) | 1 (0.5) | 0 (0) | |
| ECOG PS on CID1, n (%) | 0.002c | |||
| 0 | 96 (46.4) | 27 (32.5) | 69 (55.6) | |
| 1 | 77 (37.2) | 42 (50.6) | 35 (28.2) | |
| 2 | 34 (16.4) | 14 (16.9) | 20 (16.1) | |
| Hb <11 ng/dl on C101 | 111 (53.6) | 54 (65.1) | 57 (45.9) | 0.0073c |
| NLR >3 on CID1, n (%) | 130 (62.8) | 52 (62.6) | 78 (62.9) | 0.7645b |
| Creatinine clearance <30 on CID1, n (%) | 26 (12.6) | 9 (10.8) | 17 (13.7) | 0.6699b |
| Weight on CID1 (kg), median (IOR) | 60.5 (53.0-68.0) | 60 (52-67.4) | 61.2 (54.9-68.9) | 0.2617a |
| Initial dose on CID1 | 0.0014c | |||
| 1.25 mg/kg | 175 (84.5) | 62 (74.7) | 113 (91.1) | |
| 1.00 mg/kg | 27 (13.0) | 20 (24.1) | 7 (5.6) | |
| 0.75 mg/kg | 4 (1.9) | 1 (12) | 3 (2.4) | |
| Unknown | 1 (0.5) | 0 (0) | 1 (0.8) | |
| Number of EV cycles, median (IQR) | 5 (3-9) | 4 (2-7) | 6 (3-11) | 0.0047a |
| Use of prophylactic moisturizers, n (%) | 69 (33.3) | 24 (28.9) | 45 (36.3) | 0.2951b |
| Best of response, n (%) | 0.0004c | |||
| OR | 11 (5.3) | 2 (24) | 9 (7.2) | |
| PR | 73 (35.3) | 18 (21.7) | 55 (44.3) | |
| SD | 59 (28.5) | 25 (30.1) | 34 (27.4) | |
| PD | 53 (25.6) | 33 (39.7) | 20 (16.1) | |
| Unknown | 11 (5.3) | 5 (6.0) | 6 (4.8) | |
| Progression-free survival, median (95% CI) | 6.3 (5.1-7.5) | 4.2 (3.0-4.9) | 8.4 (6.5-10.0) | <0.0001c |
| Overall survival, median (95% CI) | 11.8 (10.3-15) | 7.2 (5.8-8.8) | 19.8 (12.2-225) | <0.0001c |
Bold values denote statistical significance at the P < 0.05 level.
C1D1, cycle 1 day 1; CR, complete response; ECOG PS, Eastern Cooperative Oncology Group performance status; EV, enfortumab vedotin; Hb, hemoglobin; NLR, neutrophil-to-lymphocyte ratio; PD, progressive disease; PR, partial response; SD, stable disease.
Wilcoxon rank sum test.
Fisher’s exact test.
Pearson’s chi-square test.
Log-rank test.
Table 3.
Multivariate analysis of the factors associated with developing early-phase EV-related cutaneous toxicity
| Odds | Lower 95% CI | Higher 95% CI | P value | |
|---|---|---|---|---|
| Age | ||||
| <75 years | Ref. | |||
| ≥75 years | 1.7841 | 0.9533 | 3.3309 | 0.0691 |
| Sex | ||||
| Male | Ref. | |||
| Female | 1.0215 | 0.5125 | 2.0362 | 0.9517 |
| Primary site | ||||
| Bladder/urethra | Ref. | |||
| Upper tract | 0.8083 | 0.4366 | 1.4963 | 0.4982 |
| Variant histology | ||||
| Absent | Ref. | |||
| Present | 1.0992 | 0.4844 | 2.4939 | 0.8210 |
| Immediate prior treatment | ||||
| Chemotherapy | Ref. | |||
| ICIs | 1.3983 | 0.6564 | 2.9786 | 0.3849 |
| ECOG PS | ||||
| 0 | Ref. | |||
| 1 | 0.4571 | 0.2375 | 0.8795 | 0.0191 |
| ≥2 | 1.1416 | 0.4578 | 2.8463 | 0.7763 |
| Liver metastasis | ||||
| No | Ref. | |||
| Yes | 0.8964 | 0.4374 | 1.8371 | 0.7653 |
| Hb | ||||
| ≥11 g/dl | Ref. | |||
| <11 g/dl | 0.4821 | 0.2635 | 0.8819 | 0.0179 |
| Initial dose | ||||
| 1.25 mg/kg | Ref. | |||
| ≤10 mg/kg | 0.2032 | 0.0779 | 0.5301 | 0.0011 |
Bold values denote statistical significance at the P < 0.05 level.
ECOG PS, Eastern Cooperative Oncology Group performance status; EV, enfortumab vedotin; Hb, hemoglobin; ICI, immune checkpoint inhibitor.
Occurrence of epEVRCTs and efficacy
Patients who developed epEVRCTs demonstrated higher complete response (CR) and combined complete and partial response (CR + PR) rates compared with those who did not develop these toxicities (Table 2). Moreover, patients with epEVRCTs had significantly longer PFS and OS (PFS: 8.4 versus 4.2 months, P < 0.0001, Figure 2A; OS: 19.8 versus 7.2 months, P < 0.0001, Figure 2B).
Figure 2.
Occurrence of early-phase EV-related cutaneous toxicities (EVRCTs) and efficacy. (A and B) Kaplan–Meier (KM) plot analysis showing progression-free survival (PFS) (A) and overall survival (OS) (B) for 207 patients with and without early-phase EVRCTs. (C and D) KM plot analysis showing PFS (C) and OS (D) for 194 patients with and without early-phase EVRCTs who survived longer than 3 months after landmark analysis. (E and F) KM plot analysis showing PFS (E) and OS (F) for 136 patients with and without early-phase EVRCTs who survived longer than 3 months after propensity score matching. Pt, patient.
We identified 98 patients with epEVRCTs and 96 without who survived longer than 3 months for landmark analysis. In this subgroup, the median PFS value was 8.4 months for patients with epEVRCTs compared with 4.9 months for those without (P = 0.0057; Figure 2C), while the median OS values for these patients were 19.8 months and 10.3 months, respectively (P = 0.0071; Figure 2D).
Further analysis, stratified by the occurrence of epEVRCTs and using PSM, included 68 patients from each group. The baseline characteristics were well balanced between the groups (Supplementary Table S1, available at https://doi.org/10.1016/j.esmorw.2024.100094). Post-PSM analysis revealed a median PFS value of 8.2 months for patients with epEVRCTs compared with 5.8 months for those without (P = 0.245; Figure 2E). The median OS value was 18.2 months in the epEVRCT group compared with 11.5 months in the non-EVRCT group (P = 0.0298; Figure 2F). In a multivariate analysis using a time-dependent Cox proportional hazards model for PFS and OS, epEVRCT was significantly associated with OS [hazard ratio (HR) 0.6457, 95% confidence interval (CI) 0.4407-0.9461, P = 0.0248; Supplementary Table S2, available at https://doi.org/10.1016/j.esmorw.2024.100094], along with Hb and ECOG PS.
Severity of epEVRCTs and efficacy
After analyzing the severity of epEVRCTs, we observed that patients with grade 1 EVRCTs had significantly better outcomes. The median PFS values were 4.9 months for patients without epEVRCTs, 9.7 months for grade 1 (HR 0.475, 95% CI 0.274-0.783), 7.9 months for grade 2 (HR 0.642, 95% CI 0.418-0.968), and 4.9 months for grade 3 (HR 1.198, 95% CI 0.418-2.703) (P = 0.0198 by log-rank test for the four groups; Figure 3A and C). Similarly, the median OS values were 10.3 months for patients without epEVRCTs, 23.7 months for grade 1 (HR 0.314, 95% CI 0.145-0.598), 12.1 months for grade 2 (HR 0.800, 95% CI 0.512-1.223), and 13.2 months for grade 3 (HR 0.747, 95% CI 0.288-1.593) (P = 0.0045 by log-rank test for the four groups; Figure 3B and C).
Figure 3.
Severity of early-phase EV-related cutaneous toxicities (EVRCTs) and efficacy. (A and B) Kaplan–Meier (KM) plot analysis showing progression-free survival (PFS) and overall survival (OS) for patients without EVRCTs and those with grade 1, 2, or 3 early-phase EVRCTs who survived longer than 3 months. (C) Forest plots representing the hazard ratios (HRs) of patients with each EVRCT grade versus patients without EVRCTs for PFS and OS who survived longer than 3 months. (D and E) KM plot analysis showing PFS and OS for patients without EVRCTs and those with grade 1, 2, or 3 early-phase EVRCTs who survived longer than 3 months after propensity score matching. (F) Forest plots representing the HRs of patients with each EVRCT grade versus patients without EVRCTs for PFS and OS who survived longer than 3 months after propensity score matching. CI, confidence interval; Pt, patient.
In the PSM cohort, OS was significantly longer for patients with grade 1 epEVRCTs. The median OS value was 11.5 months for patients without epEVRCTs, not reached for grade 1 (HR 0.279, 95% CI 0.145-0.598), 12.1 months for grade 2 (HR 0.858, 95% CI 0.488-1.453), and 13.2 months for grade 3 (HR 0.796, 95% CI 0.274-1.838) (P = 0.0217 by log-rank test for the four groups; Figure 3E and F). However, no significant difference in PFS was observed (5.8 months versus 8.9 months versus 8.2 months versus 4.8 months, respectively, P = 0.2722 by log-rank test for the four groups; Figure 3D and F). These findings highlight the relationship between the severity of epEVRCTs and therapeutic outcomes in patients undergoing EV treatment.
Discussion
Our study provides a comprehensive analysis of EVRCTs in patients with la/mUC treated with EV. Utilizing one of the largest real-world datasets available, we have elucidated the incidence, risk factors, and clinical implications of EVRCTs. We also specifically focused on early-phase toxicities and their impact on patient outcomes.
Our findings indicate that EVRCTs are a common adverse event, affecting nearly 60% of patients undergoing EV treatment. The median onset time of 14 days underscores the need for vigilant monitoring during the early stages of therapy. The skin reactions ranged from pruritus and eczema to severe conditions like Stevens–Johnson syndrome, necessitating a proactive management approach. Topical corticosteroids were the primary treatment approach in over one-third of cases, while systemic corticosteroids were reserved for more severe reactions. Treatment interruptions, dose reductions, and discontinuations highlight the significant clinical burden of these toxicities on treatment efficacy.
Our analysis identified several significant risk factors for the development of epEVRCTs. Patients with better PS (ECOG PS = 0), higher Hb levels (≥11 g/dl), and those who received the standard initial dose of 1.25 mg/kg were more likely to develop these toxicities. These findings suggest that patients treated with a more robust regimen, who likely receive higher doses and more cycles of EV, are at an increased risk. EVRCTs have been reported to more likely occur in intertrigo areas,3 with physical irritation caused by high activity and frequent movement also contributing to EVRCT risk. Our results were also consistent with reports of dose effects on EVRCT occurrence that EVRCT occurs more frequently with higher dosage due to body weight.6 These insights can inform clinical decisions, particularly for tailoring EV dosing and monitoring strategies to mitigate the risk of severe skin toxicities.
A notable finding of our study is the positive correlation between epEVRCTs and improved clinical outcomes. Patients who developed epEVRCTs had significantly higher response rates (CR and CR + PR) and longer PFS and OS compared with those who did not develop these toxicities. This association persisted even after adjusting for immortal time bias and employing PSM to balance the baseline characteristics between groups. Our findings align with the report by Vlachou et al. showing a correlation between EVRCT occurrence and OS, using a time-dependent Cox model to reduce immortal time bias.8 The enhanced survival outcomes in patients with epEVRCTs may be attributed to a variety of factors. One hypothesis is that the occurrence of EVRCTs could be an indicator of a more robust immune response, potentially reflecting a more effective antitumor activity of EV. This theory aligns with observations in other cancer therapies, where immune-related adverse events correlate with better treatment responses.10
Further stratification of epEVRCTs by severity revealed that patients who had experienced grade 1 toxicities achieved the most favorable outcomes, particularly for OS. This finding is important if mild-to-moderate skin reactions may serve as a biomarker for treatment efficacy, although the underlying biological mechanism is unclear. Notably, mild-to-moderate skin reactions do not compromise patient safety or necessitate any treatment alteration, which may explain why patients who experienced grade 2 or higher skin reactions did not show improvement in oncological outcomes. Another possible explanation is that grade 1 EVRCTs may simply indicate administration of the optimal amount of EV for an individual patient, with no EVRCT suggesting suboptimal EV dosing. It was reported that EV serum exposure was associated with efficacy and safety outcomes from three EV monotherapy studies (EV-101, EV-201, and EV-301). Since EV serum exposure cannot be monitored in routine practice, the degree of EVRCTs may be a biomarker for whether the ideal blood concentration range has been reached.
Future research should aim to elucidate the underlying mechanisms linking EVRCTs to treatment efficacy. Understanding whether these toxicities are merely a marker of a stronger systemic response or play an active role in enhancing antitumor activity could provide crucial details for the development of strategies to therapeutically harness this phenomenon. Additionally, exploring preventative measures or treatments to manage EVRCTs without compromising efficacy could further improve patient outcomes. The use of moisturizers, which has been recommended by experts,3 was not an effective preventive approach in this study. Future research is needed to establish a method to prevent EVRCTs without compromising treatment effectiveness.
Despite the significant insights provided by our study, several limitations must be acknowledged to properly contextualize our findings. The first limitation is the retrospective study design, which introduces potential biases, such as selection and recall biases. Data collected from medical records may vary in accuracy and completeness, impacting the reliability of the findings. Secondly, the study has a limited population diversity. The study’s cohort is predominantly Japanese, which limits the generalizability of the results to other ethnic groups. Differences in genetic backgrounds and health care practices across populations could influence the incidence and management of EVRCT. A third limitation is the potential for confounding variables. Despite using multivariate analysis and PSM, unmeasured confounding variables may still affect the study’s results. Factors like genetic predispositions, comorbidities, and concurrent treatments were not fully controlled, potentially influencing the observed associations between EVRCTs and treatment outcomes.
In conclusion, our study provides real-world data of epEVRCTs as a common toxicity in la/mUC patients treated with EV. Furthermore, we have demonstrated a positive correlation between mild skin toxicity and favorable antitumor efficacy. The underlying mechanisms remain an important focus for future investigations. These findings highlight the need for vigilant monitoring and proactive management of skin toxicities to optimize the therapeutic benefits of EV.
Acknowledgements
We thank J. Iacona, PhD, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.
Funding
This work was supported by Astellas Pharma Inc. (no grant number). The funder played no role in the design or performance of this study.
Disclosure
YK: advisory board from Astellas; speaking honoraria from Astellas, MSD, Merck Biopharma; research grant from Astellas. Y. Matsushita: speaking honoraria from Astellas. HI: speaking honoraria from Astellas. T. Sazuka: speaking honoraria from MSD. NT: speaking honoraria from MSD. M. Kato: advisory board from Astellas, MSD, Merck Biopharma; speaking honoraria from Astellas, MSD, Merck Biopharma. ST: speaking honoraria from Astellas, MSD, Merck Biopharma. TM: speaking honoraria from Astellas. JW: support for travel expenses by Astellas, MSD. HN: speaking honoraria from Astellas, MSD, Merck Biopharma; research grant from Astellas. HK: advisory board from Astellas, MSD; speaking honoraria from Astellas, MSD, Merck Biopharma; research grant from MSD. T. Kobayashi: advisory board from Astellas, MSD, Merck Biopharma; speaking honoraria from Astellas, MSD, Merck Biopharma; research grant from Astellas. All other authors have declared no conflicts of interest.
Supplementary Data
References
- 1.Powles T., Rosenberg J.E., Sonpavde G.P., et al. Enfortumab vedotin in previously treated advanced urothelial carcinoma. N Engl J Med. 2021;384(12):1125–1135. doi: 10.1056/NEJMoa2035807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Powles T., Valderrama B.P., Gupta S., et al. Enfortumab vedotin and pembrolizumab in untreated advanced urothelial cancer. N Engl J Med. 2024;390(10):875–888. doi: 10.1056/NEJMoa2312117. [DOI] [PubMed] [Google Scholar]
- 3.Lacouture M.E., Patel A.B., Rosenberg J.E., O’Donnell P.H. Management of dermatologic events associated with the Nectin-4-directed antibody-drug conjugate enfortumab vedotin. Oncologist. 2022;27(3):e223–e232. doi: 10.1093/oncolo/oyac001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Nguyen M.N., Reyes M., Jones S.C. Postmarketing cases of enfortumab vedotin-associated skin reactions reported as Stevens–Johnson syndrome or toxic epidermal necrolysis. JAMA Dermatol. 2021;157(10):1237–1239. doi: 10.1001/jamadermatol.2021.3450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Malik R., Xiang D.H., Riew G.J., et al. Characterization of adverse cutaneous effects in the setting of enfortumab vedotin for metastatic urothelial carcinoma: a retrospective review. J Am Acad Dermatol. 2024;91:753–755. doi: 10.1016/j.jaad.2024.06.037. [DOI] [PubMed] [Google Scholar]
- 6.Vlachou E., Matoso A., McConkey D., et al. Enfortumab vedotin-related cutaneous toxicity and radiographic response in patients with urothelial cancer: a single-center experience and review of the literature. Eur Urol Open Sci. 2023;49:100–103. doi: 10.1016/j.euros.2023.01.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Molina G.E., Schwartz B., Srinivas S., Shah S., Zaba L.C. In patients with advanced urothelial carcinoma, immune checkpoint inhibition prior to enfortumab vedotin is associated with high-grade skin toxicity. Eur Urol. 2023;83(4):377–378. doi: 10.1016/j.eururo.2022.12.009. [DOI] [PubMed] [Google Scholar]
- 8.Vlachou E., Johnson B.A., McConkey D., et al. Enfortumab vedotin-related cutaneous toxicity correlates with overall survival in patients with urothelial cancer: a retrospective experience. Front Oncol. 2024;14 doi: 10.3389/fonc.2024.1377842. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kobayashi T., Ito K., Kojima T., et al. Risk stratification for the prognosis of patients with chemoresistant urothelial cancer treated with pembrolizumab. Cancer Sci. 2021;112(2):760–773. doi: 10.1111/cas.14762. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Otsuka H., Kita Y., Ito K., et al. Immune-related adverse events in urothelial cancer patients: adjustment for immortal time bias. Cancer Sci. 2022;113(11):3912–3921. doi: 10.1111/cas.15539. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.