EV is a novel therapy for UC refractory to prior chemotherapy and ICIs, which targets nectin-4 expressed in UC [1,2,3]. Nectin-4 is also expressed homogenously in normal skin keratinocytes and appendages. TRSRs are one of EV’s most frequent AEs [1, 3,4,5], yet their association with EV survival and treatment response remains uncertain. We compared survival outcomes and treatment response to EV between those with and without TRSRs. The present study demonstrated that patients with TRSRs showed significantly longer CSS and PFS and superior ORR and suggested that susceptibility to EV is shared in skin and UC within an individual patient. Further, TRSRs may serve as a surrogate marker of treatment efficacy.

A Japanese subgroup analysis of EV 301 and real-world data showed median OS and PFS of 10–15 and 4–7 months, respectively [9, 16, 17]. ORR is reportedly 34.4%–57.7% [9, 16], similar to our study (although our cohort showed slightly higher ORR). Several Japanese studies have reported 29.4–55.6% TRSR occurrences [9, 16, 18]. Similarly, approximately half of our patients experienced TRSRs.

TRSRs typically occur relatively soon after EV initiation [1, 19, 20]. Eighty percent of TRSRs in the present cohort happened within the first three cycles, comparable to the previous studies. A study from a single center reported the association between treatment efficacy and TRSRs [21]. The study also showed that the patients with TRSRs showed a significantly superior response (ORR 57.7% vs. 24.0%) [21]. However, the researchers did not examine survival benefits. This was the pioneering study demonstrating an association between TRSRs and survival outcomes.

The mechanism of TRSRs to EV has not been fully elucidated; however, there are two proposed mechanisms. One is the direct delivery of MMAE, which causes damage to normal tissue, such as the epidermis, due to its expression of nectin-4 [4]. Another mechanism could be an allergy-related skin reaction, with type IV delayed hypersensitivity considered its pathology. This is believed to be linked to severe skin reactions such as Stevens-Johnson syndrome and toxic epidermal necrolysis [22]. MMAE’s direct effects are considered to be the leading cause of TRSRs.

In addition, the pathophysiology of EV-related TRSRs has been assessed in several previous studies. Nectin-4, the target of EV, is typically expressed in the human skin [6, 23]. Accordingly, the skin is likely to be affected by the pharmacological effect of EV. Furthermore, biopsy specimens of skin collected from the TRSR group demonstrated prominent ring and starburst mitoses and dyskeratosis within the epidermis, which reflect mitotic activity arrest [24]. These findings were similar to those of taxane-induced toxic erythema [25]. The MMAE of EV and taxanes is the microtubule inhibitor, which promotes the antitumor effect through mitosis inhibition and consequent apoptosis [3, 25]. Therefore, this pharmacological mechanism of EV may be related to TRSRs as well as the antitumor effect.

Similarly, TRSRs due to other antitumor agents have been previously reported as clinical indicators of treatment efficacy. For example, skin reactions related to epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (EGFR-TKI) for non-small cell lung cancer indicated treatment efficacy [26]. Similar to nectin-4, EGFR is abundantly expressed in the keratinocytes in the epidermis and drives cell cycle activation and proliferation by activating downstream pathways [27, 28]. Accordingly, EGFR-TKI affects keratinocytes and causes TRSRs by inducing growth arrest and apoptosis, which also promote antitumoral effects [28]. Thus, TRSRs due to EGFR-TKI would be induced by the common pharmacological effect that acts in tumors, so the TRSRs could be a clinical indicator of therapeutic effect. In this perspective, EVs and EGFR-TKIs are similar in that their inherent pharmacological effects appear to arise from a common mechanism for AEs and antitumor effects. Therefore, this precedent would support the association between TRSRs and treatment response in EV.

Regarding the association between other TRAEs of EV and treatment efficacy, several studies have reported that peripheral neuropathy (PN) was a favorable factor in survival outcomes [29, 30]. EV-related PN would be MMAE-mediated inhibition of microtubule-dependent axonal transport [31]. However, PN has a relatively late onset, with a median of 15 weeks [30]. Considering this, PN would be a dose-dependent AE, which could affect only responders to EV. Therefore, PN might be a consequent manifestation of sustained therapeutic effect rather than a predictive factor for treatment efficacy.

One possible mechanism explaining the association between treatment response and TRSRs in an individual is genetic polymorphism. ATP binding cassette (ABC) transporter, which is encoded by the multidrug resistance (MDR) genes, promotes the efflux of MMAE [32]. Of these transporters, many single nucleotide variants (SNVs) have been described, and some SNVs have reportedly affected prognosis and AEs in various cancer types [32].

For instance, the specific genotype of rs1045642 variant in MDR1 was reportedly a favorable prognostic indicator for colorectal cancer treated with oxaliplatin [33]. Regarding ADCs, MDR1 upregulation was related to poor prognosis in patients with Hodgkin lymphoma who received brentuximab vedotin, an ADC for CD30 incorporated MMAE; however, it was not referred to as whether specific SNVs of ABC transporter regarding MMAE [34]. Thus far, whether the specific variant of MDR for MMAE exists remains uncertain, and further investigation is needed.

Furthermore, we evaluated how dose reduction/discontinuation of EV affects treatment efficacy. In this study, 37% of the patients (25/67) experienced dose reduction/discontinuation due to TRAEs. Among these patients, 68% (17/25) had TRSRs. Nevertheless, the dose reduction/discontinuation group demonstrated superior survival outcomes. A Japanese real-world study did not find a relationship between dose intensity and survival outcomes [35]. Rather, several studies that have shown that even lower dose intensity due to PN revealed favorable survival outcomes and response [29, 30]. These results show that dose intensity may not influence survival outcomes.

There are several limitations to the present study. First, it was retrospective and included a small number of patients. Second, there might be unrecognized confounding factors for survival outcomes, although we performed comprehensive analyses. Third, it was a multicenter study, and a central review regarding radiological evaluation was not performed. Finally, TRAEs were based on physicians’ reports, so there might be slight variances regarding grade evaluation. Considering these limitations, further investigation with a larger sample size would be needed to confirm the propriety of our results and verify our hypothesis.