Phase II Trial of Enfortumab Vedotin in Patients With Previously Treated Advanced Head and Neck Cancer

Abstract

PURPOSE

Despite advances in immunotherapy, unresectable recurrent/metastatic head and neck cancer (HNC) carries a poor prognosis, and effective treatments are needed. As nectin-4 is widely expressed in HNC, enfortumab vedotin (EV), a nectin-4–directed antibody-drug conjugate, was explored in HNC in EV-202 (ClinicalTrials.gov identifier: NCT04225117).

METHODS

This open-label, multicohort, phase II study evaluated intravenous EV 1.25 mg/kg on days 1, 8, and 15 of each 28-day cycle. In the HNC cohort, eligible patients had recurrent/metastatic HNC and had received platinum-based therapy for locally advanced/metastatic disease and a PD-1/PD-L1 inhibitor. The primary end point was investigator-assessed confirmed objective response rate (ORR) per RECIST version 1.1. Secondary end points were investigator-assessed duration of response (DOR), disease control rate (DCR), and progression-free survival (PFS); overall survival (OS); and safety.

RESULTS

The primary analysis included 46 patients; all received EV (median follow-up, 9.3 months). Most patients (52.2%) had ≥3 previous lines of systemic therapy in the metastatic setting. Confirmed ORR was 23.9%, DCR was 56.5%, and median DOR was not reached (median DOR was 9.4 months at a later data cutoff [median follow-up, 11.3 months]). Median PFS and OS were 3.9 and 6.0 months, respectively. Treatment-related adverse events (TRAEs) occurring in >20% of patients were alopecia (28.3%), fatigue (26.1%), and peripheral sensory neuropathy (23.9%). Sixteen patients (34.8%) experienced grade ≥3 TRAEs; anemia and decreased neutrophil count occurred in ≥1 patient (both n = 2; 4.3%).

CONCLUSION

EV demonstrated antitumor activity in heavily pretreated HNC. Safety was consistent with the known safety profile of EV; no new safety signals were identified. These data support further evaluation of EV for advanced HNC not amenable to definitive local therapy.

INTRODUCTION

Globally, head and neck cancer (HNC) is the sixth most commonly diagnosed malignancy, with an estimated 946,000 new cases and 482,000 related deaths in 2022.^1,2 Head and neck squamous cell carcinoma (HNSCC) arising from mucosal epithelial cells of the oral cavity, pharynx, and larynx comprises >90% of HNC cases.^3,4 Approximately 50% of individuals diagnosed with locally advanced HNSCC experience recurrent disease, and up to 10% present with metastatic disease.^3-6

CONTEXT

• Key Objective

• To assess the antitumor activity and safety of enfortumab vedotin (EV) in patients with recurrent or metastatic head and neck cancer (HNC) whose disease progressed after receiving platinum-based chemotherapy and PD-1/PD-L1 inhibitor treatment.

• Knowledge Generated

• EV demonstrated a confirmed objective response rate of 23.9% in patients with HNC who received EV. The safety profile of EV in patients with HNC was consistent with the well-established safety profile observed with its use in patients with advanced urothelial carcinoma.

• Relevance (J.W. Friedberg)

• Favorable safety and efficacy of EV in this trial provides rationale for further clinical development in patients with relapsed and refractory HNC.*

• *Relevance section written by JCO Editor-in-Chief Jonathan W. Friedberg, MD.

Currently, unresectable recurrent or metastatic (R/M) HNC is treated with palliative intent.^4,7 On the basis of multiple studies, the PD-1 inhibitor pembrolizumab with/without platinum-based chemotherapy (PBC) is central to first-line treatment of R/M HNSCC.^4,7 For patients who previously received PBC and a PD-1/PD-L1 inhibitor, no standard treatment has been established.⁴ Despite the advent of immunotherapy and advances in treatment, median overall survival (OS) is approximately 10-13 months in patients with R/M HNSCC and ≤6 months in patients whose disease progressed within 6 months after PBC.^3,8,9 Thus, improved treatments are needed for patients with R/M HNC.

Enfortumab vedotin (EV) is an antibody-drug conjugate (ADC) comprising a fully human monoclonal antibody directed to the cell-adhesion molecule nectin-4 attached to the microtubule-disrupting agent monomethyl auristatin E (MMAE) by a protease-cleavable linker.^10,11 Nectin-4 is expressed in a wide variety of cancers, including most bladder cancers and HNCs.^10,12 Notably, data from bladder cancer trials, including the EV-201 clinical trial of EV monotherapy in urothelial carcinoma (UC), suggest that the degree of nectin-4 expression is not associated with response.^13-16 Preclinical evidence suggests EV selectively binds to and is internalized by nectin-4–expressing cells, resulting in intracellular release of MMAE and apoptotic cell death.^10,11 Furthermore, studies have suggested that nectin-4 can promote cancer cell proliferation, migration, angiogenesis, and epithelial-mesenchymal transition in other tumor types/malignancies.^17-19 However, the mechanistic role of nectin-4 in HNSCC tumorigenesis and progression is not well understood.

In the pivotal EV-301 trial in patients with locally advanced or metastatic (la/m) UC who previously received PBC and experienced progression/relapse during/after PD-1/PD-L1 inhibitor treatment, EV reduced risk of death by 30% versus standard chemotherapy.¹¹ EV is approved in multiple countries globally for adults with previously treated la/mUC.^20,21 In addition, in the EV-302 trial, EV in combination with pembrolizumab demonstrated superior OS and progression-free survival (PFS) compared with PBC in patients with previously untreated la/mUC; this combination is approved in multiple countries for the treatment of adults with la/mUC.^20,21

Given the antitumor activity of EV in la/mUC, EV is being investigated in the multicohort, phase II trial EV-202 (ClinicalTrials.gov identifier: NCT04225117) for treatment of other nectin-4–expressing solid tumors with known sensitivity to microtubule-disrupting agents, including HNC. We report antitumor activity and safety of EV in adults with R/M HNC previously treated with PBC and PD-1/PD-L1 inhibitor therapy.

METHODS

Trial Design

EV-202 is an open-label, multicohort, phase II study evaluating antitumor efficacy and safety of EV in patients with select la/m solid tumors (Data Supplement, Fig S1, online only). Patients were enrolled into cohorts by tumor type. The trial used a two-stage Bayesian optimal design for phase II wherein an interim analysis of antitumor activity was conducted for each cohort when 20 patients treated with EV were evaluable for investigator-assessed response. The minimum number of responders required to proceed to stage II was based on reference objective response rate (ORR) and target ORR for each cohort. EV-202 aimed to include 40 evaluable patients in the final analysis for the HNC cohort, with a minimum of seven responders (17.5%) to declare promising antitumor activity.

Study Patients

Adults with histologically/cytologically confirmed HNC arising from the oral cavity, oropharynx (both human papillomavirus [HPV]-related and HPV-unrelated), hypopharynx, and larynx were eligible for the HNC cohort, which enrolled patients with R/M disease. Patients were required to have measurable disease by RECIST version 1.1 and an Eastern Cooperative Oncology Group performance status of 0 or 1. Additional inclusion criteria included R/M disease not amenable to curative-intent treatment and evidence of radiographic progression on/after the last treatment. Eligible patients had experienced tumor progression/relapse or treatment discontinuation because of toxicity after a platinum-based standard-of-care regimen for R/M disease, with ≤2 previous lines of cytotoxic anticancer therapy in the R/M setting. Patients with three or more previous lines of systemic therapy in the R/M setting were eligible for the trial as long as ≤2 of these previous lines included cytotoxic therapy. Patients with locally advanced disease who previously received curative-intent treatment with platinum-based standard of care in the adjuvant or neoadjuvant setting or with concomitant radiation therapy (ie, definitive therapy) were eligible if their tumors progressed/relapsed within 6 months of completion. Patients must have received PD-1/PD-L1 inhibitor therapy unless contraindicated. Nectin-4 expression was not an inclusion criterion but was retrospectively assessed. Exclusion criteria included preexisting grade ≥2 sensory or motor neuropathy, active CNS metastases, ongoing clinically significant toxicity (grade ≥2, except alopecia) associated with previous treatment, uncontrolled diabetes mellitus (hemoglobin A1c [HbA1c] ≥8% or HbA1c between 7% and <8% with associated diabetes symptoms [polyuria or polydipsia] that are not otherwise explained) within 3 months before the first dose of study treatment, and previous treatment with an MMAE-based ADC.

Intervention

Patients received intravenous EV 1.25 mg/kg (maximum dose: 125 mg) on days 1, 8, and 15 of each 28-day cycle until disease progression or treatment discontinuation because of unacceptable toxicity, initiation of a new anticancer therapy, continuous dose interruption >6 months, or meeting another discontinuation criterion. Dose modifications and interruptions for treatment-associated adverse events (AEs) were permitted per protocol.

End Points and Assessments

The primary end point was investigator-assessed confirmed ORR (complete response [CR] + partial response [PR]) per RECIST version 1.1. Secondary end points were investigator-assessed duration of response (DOR), disease control rate (DCR; CR + PR + stable disease [SD]), and PFS per RECIST version 1.1; OS; and safety. Exploratory end points included antitumor activity per blinded independent central review, exploratory biomarkers possibly correlating with treatment outcome, selected pharmacokinetic (PK) parameters for EV and MMAE, incidence of antitherapeutic antibodies to EV, and patient-reported outcomes (PROs).

Efficacy was assessed by computed tomography/magnetic resonance imaging at screening/baseline and every 8 weeks (±1 week) from the first dose of study treatment through post-treatment follow-up until radiologically confirmed disease progression, initiation of new anticancer therapy, death, consent withdrawal, loss to follow-up, or study closure, whichever came first. After 1 year on study treatment, response assessment frequency was reduced to every 12 weeks (±1 week). If response was assessed as CR or PR per investigator, confirmatory imaging was required 4 weeks (±1 week) after the first response. Confirmed best overall response was derived per RECIST version 1.1. If a patient did not have a confirmed CR or PR, but had ≥one tumor assessment of CR, PR, or SD ≥49 days after the first dose, they were considered to have a best overall response of SD.

AEs were graded per the National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03. Prespecified AEs of special interest for EV were analyzed and reported as composite terms (hyperglycemia, peripheral neuropathy, ocular events [dry eye, corneal disorders, and blurred vision], skin reactions, and infusion-related reactions).

Although not required per protocol, HPV status was reported as part of each patient's baseline disease history when available.

Methods for immunohistochemistry, immunogenicity and PK analyses, and PROs are reported in the Data Supplement.

Statistical Analysis

The data cutoff date was April 11, 2022. Data were summarized descriptively for continuous data and with frequencies/percentages for categorical data. For time-to-event end points (DOR, PFS, and OS), median survival was estimated using the Kaplan-Meier method with a two-sided 95% CI. For ORR and DCR, two-sided 95% CIs were calculated using the Clopper-Pearson method. Mean and standard deviation of EV and MMAE concentrations at selected time points are presented. PROs were summarized at each visit. Results from the EQ-5D-5L questionnaire were reported as number and percentage of patients in each level (no, slight, moderate, severe, and extreme problems) of the five dimensions. EQ-5D-5L visual analog scale scores and global pain were summarized descriptively.

OS, PFS, safety, and PROs were analyzed in all patients receiving any study drug. ORR, DOR, and DCR were analyzed in all patients with measurable disease at baseline per investigator assessment who had ≥two postbaseline response assessments or were no longer in follow-up for response at the time of analysis. The PK analysis population included patients with ≥one blood sample (with known times of sampling and dosing on the day of sampling) assayed for EV and MMAE concentrations.

Trial Oversight

Study protocol/amendments were approved by institutional review boards or independent ethics committees and conducted in accordance with the Declaration of Helsinki and Good Clinical Practice Guidelines of the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use. Written informed consent was obtained from each patient before any study-related screening procedures were performed.

RESULTS

Patient Demographics and Disposition

Sixty-two patients were screened for the HNC cohort; 46 were enrolled, treated with EV, and included in the safety and efficacy analyses (Fig 1). Median follow-up was 9.3 months (95% CI, 7.4 to 12.9). Patients received EV for a median of 2.8 months (range, 0.8-19.5), and six patients (13.0%) remained on treatment at the data cutoff date.

FIG 1.

Patient flow diagram as of April 11, 2022. Only the primary reason for discontinuation was collected. ^aPatients could be excluded for more than one reason. ^bPatient also met an exclusion criterion (active CNS metastases). ^cOne patient also did not meet the inclusion criteria (la/m disease not amenable to curative-intent treatment). ^dExclusion criteria included the following cardiovascular conditions: unstable angina, myocardial infarction, or cardiac symptoms (including congestive heart failure) consistent with New York Heart Association Class III-IV within 6 months before the first dose of study drug. AE, adverse event; ECOG PS, Eastern Cooperative Oncology Group performance status; la/m, locally advanced or metastatic; PD, progressive disease.

Median age was 65 years (range, 33-81), and 40 patients (87.0%) were male (Table 1). Almost all patients (97.8%) had HNC of squamous histology; one patient (2.2%) had adenocarcinoma. Most primary tumors were in the oral cavity (15 patients; 32.6%) or pharynx (14 patients; 30.4%). All patients had previous treatment with a PD-1/PD-L1 inhibitor and PBC, 34 (73.9%) had previous treatment with a taxane, 26 (56.5%) were previously treated with cetuximab, and 24 (52.2%) had received ≥three previous lines of systemic therapy in the metastatic setting.

TABLE 1.

Patient Demographics and Disease Characteristics

Characteristic	N = 46
Age, years, median (range)	65 (33-81)
Sex, No. (%)
Male	40 (87.0)
Female	6 (13.0)
Race, No. (%)
White	27 (58.7)
Asian	15 (32.6)
Black or African American	1 (2.2)
Not reported	3 (6.5)
Region, No. (%)
North America	30 (65.2)
Asia	16 (34.8)
BMI, kg/m2,a No. (%)
<18.5	3 (6.8)
≥18.5 to <25.0	34 (77.3)
≥25.0	7 (15.9)
ECOG PS, No. (%)
0	16 (34.8)
1	30 (65.2)
Tobacco history, No. (%)
Never used	15 (32.6)
Former	30 (65.2)
Current	1 (2.2)
Primary tumor location, No. (%)
Oral cavity	15 (32.6)
Pharynx	14 (30.4)
Larynx	10 (21.7)
Other	7 (15.2)
Human papillomavirus status,b No. (%)
Negative	6 (13.0)
Positive	20 (43.5)
Unknown	20 (43.5)
Previous lines of systemic therapy,c No. (%)
1	7 (15.2)
2	8 (17.4)
≥3	31 (67.4)
Previous lines of systemic therapy in  metastatic setting, No. (%)
1	7 (15.2)
2	9 (19.6)
≥3	24 (52.2)
Time since initial diagnosis,d months,  median (range)	28.7 (8.4-161.8)
Type of previous systemic therapy, No. (%)
PD-1/PD-L1 inhibitor	46 (100)
PBC	46 (100)
Taxane	34 (73.9)
Cetuximab	26 (56.5)
Targeted therapy	1 (2.2)
Best response to previous PBC, No. (%)
CR	15 (32.6)
PR	9 (19.6)
SD	8 (17.4)
PD	8 (17.4)
Not evaluable	2 (4.3)
Not applicable	2 (4.3)
Unknown	2 (4.3)
Best response to previous PD-1/PD-L1 inhibitor  therapy, No. (%)
CR	2 (4.3)
PR	5 (10.9)
SD	16 (34.8)
Non-CR/non-PD	1 (2.2)
PD	20 (43.5)
Not evaluable	1 (2.2)
Unknown	1 (2.2)
PD on previous systemic therapy, No. (%)
PD-1/PD-L1 inhibitor	20 (43.5)
PBC	8 (17.4)
Taxane	8 (17.4)
Cetuximab	10 (21.7)
Targeted therapy	1 (2.2)
Nectin-4 immunohistochemistry H-score  (tissue),e,f median (range)	180 (20-300)
PD-L1 immunohistochemistry (tissue),f,g No. (%)
Low (CPS <1)	6 (15.4)
High (CPS ≥1)	33 (84.6)

NOTE. Some proportions were calculated on the basis of nonmissing data.

Abbreviations: CPS, combined positive score; CR, complete response; ECOG PS, Eastern Cooperative Oncology Group performance status; PBC, platinum-based chemotherapy; PD, progressive disease; PR, partial response; SD, stable disease.

^a Data missing for two patients (n = 44).

^b Among tumors in any anatomic location (not limited to oropharyngeal).

^c Includes previous systemic therapy in the locally advanced or metastatic setting or previous platinum-based therapy received in the neoadjuvant/adjuvant setting if disease progression occurred ≤6 months after therapy completion.

^d Time from initial diagnosis of primary cancer to date of first dose.

^e n = 43 patients with evaluable tumor tissue. H-score range, 0-300.

^f Assessed using validated immunohistochemical assays (nectin-4, M22-321b41.1; PD-L1, 22C3 antibody clones).

^g n = 39 patients with evaluable tumor tissue.

Tumor HPV status was reported for 26 patients (56.5%) across multiple anatomic subtypes; 20 tumors (43.5%) were HPV-positive and six (13.0%) were HPV-negative (Table 1). Among three patients diagnosed with oropharyngeal squamous cell carcinoma, three tumors (100%) were HPV-positive. Sixteen patients in the cohort had diagnoses of pharyngeal squamous cell carcinoma or tongue squamous cell carcinoma without specification of whether they were oropharyngeal. Tissue was available for baseline nectin-4 testing for 43 of the 46 patients; all 43 tumors had detectable nectin-4 (H-score >0). Median nectin-4 H-score was 180 (range, 20-300). Among 39 patients for whom tissue was available for baseline PD-L1 expression, six tumors (15.4%) were PD-L1–low and 33 (84.6%) were PD-L1–high.

Antitumor Activity

Because the number of patients (four responders) with a confirmed response (CR or PR) in the HNC cohort exceeded the prespecified minimum number of responders (two responders) at the stage I interim analysis, the study proceeded to stage II. Among all 46 response-evaluable patients, investigator-assessed confirmed ORR was 23.9% (95% CI, 12.6 to 38.8; Table 2). Investigator-assessed confirmed DCR was 56.5% (95% CI, 41.1 to 71.1), with a best overall response of CR for one patient (2.2%), PR for 10 patients (21.7%), and SD for 15 patients (32.6%). Postbaseline tumor assessments were conducted in 39 of 46 patients; reasons for study discontinuation before completion of disease evaluation for the remaining seven patients are provided in Figure 1. Of the 39 patients, 23 (59.0%) had tumor reduction from baseline, 14 (35.9%) had reductions of 30% or more from baseline (Fig 2A), and three (7.7%) had only one postbaseline tumor assessment. Because the assessments for these three patients were completed before the protocol-specified minimum time of 49 days after the first dose, their response was not evaluable per RECIST version 1.1. Time to response was rapid (Fig 2B); among 11 patients with a confirmed response (CR or PR), median time to response was 1.7 months (range, 1.5 to 1.9). Median DOR was not reached (95% CI, 2.5 months to not reached; Fig 3A). As of the data cutoff, six patients had ongoing progression-free responses beyond 6 months. Among all 46 patients, median PFS was 3.9 months (95% CI, 2.8 to 4.7) and median OS was 6.0 months (95% CI, 4.4 to 10.7; Figs 3B and 3C). Response assessments per blinded independent central review were generally consistent with investigator assessments (Data Supplement, Table S1). As of a later data cutoff date of July 15, 2022 (median follow-up, 11.3 months), median DOR was 9.4 months (95% CI, 2.5 to not reached; Fig 3D).

TABLE 2.

Summary of Response by Investigator Assessment

Parameter/Variable	Patients (N = 46)
Confirmed ORR,a No. (%)	11 (23.9)
95% CIb	12.6 to 38.8
Confirmed DCR,c No. (%)	26 (56.5)
95% CIb	41.1 to 71.1
Best overall response, No. (%)
Confirmed CR	1 (2.2)
Confirmed PR	10 (21.7)
SD	15 (32.6)
PD	10 (21.7)
Not evaluabled	10 (21.7)

Abbreviations: CR, complete response; DCR, disease control rate; ORR, objective response rate; PD, progressive disease; PR, partial response; SD, stable disease.

^a Patients whose best overall response was confirmed CR or PR according to RECIST version 1.1.

^b Using exact method on the basis of binomial distribution (Clopper-Pearson).

^c Patients with best overall response of confirmed CR, confirmed PR, or SD (≥7 weeks).

^d Seven patients were considered not evaluable because of study discontinuation before postbaseline response assessment was performed, and three patients were considered not evaluable because of having only one postbaseline tumor assessment, which was completed before the protocol-specified minimum time of 49 days after the first dose.

FIG 2.

Efficacy per investigator assessment. (A) Best change from baseline in size of target lesion. (B) Outcomes in patients with confirmed responses (includes CR and PR). ^aA total of 39 patients with postbaseline tumor assessments were included; three patients were considered NE because their only postbaseline scan was performed before the protocol-specified minimum time of 49 days after the first dose. CR, complete response; NE, not evaluable; PD, progressive disease; PR, partial response; SD, stable disease.

FIG 3.

Kaplan-Meier plots of (A) DOR per investigator assessment, (B) PFS per investigator assessment, (C) OS, and (D) DOR per investigator assessment as of the July 15, 2022, data cutoff date. CR, complete response; DOR, duration of response; NR, not reached; OS, overall survival; PFS, progression-free survival; PR, partial response.

ORRs were consistent across prespecified subgroups (Table 3). The investigator-assessed confirmed ORR and median OS for patients in Asia (Japan; n = 16) versus North America (n = 30) were 37.5% versus 16.7% and 10.7 months versus 5.2 months, respectively. Among the ad hoc subgroups, the ORR and median OS were 38.5% and 11.4 months, respectively, for the subgroup of patients (n = 13) who had received ≤2 previous lines of systemic therapy and one previous line of cytotoxic therapy in the R/M setting (or in the adjuvant/neoadjuvant setting if progressive disease occurred within the following 6 months; Table 3). The distribution of and median nectin-4 expression were similar between responders and nonresponders (Data Supplement, Fig S2).

TABLE 3.

Subgroup Analysis of Confirmed ORR by Investigator Assessment and OS

Parameter	Value	No.	ORR,a No. (%)	95% CIb	Median OS,c Months	95% CI
Prespecified analyses
All		46	11 (23.9)	12.6 to 38.8	6.0	4.4 to 10.7
Age group 1, years	<65	22	6 (27.3)	10.7 to 50.2	6.0	3.4 to 10.7
	≥65	24	5 (20.8)	7.1 to 42.2	9.3	4.3 to NE
Age group 2, years	<75	43	10 (23.3)	11.8 to 38.6	6.0	4.3 to 10.7
	≥75	3	1 (33.3)	0.8 to 90.6	5.7	3.3 to NE
Sex	Female	6	1 (16.7)	0.4 to 64.1	4.1	0.8 to NE
	Male	40	10 (25.0)	12.7 to 41.2	6.0	4.4 to 10.7
Race	White	27	5 (18.5)	6.3 to 38.1	5.1	3.4 to 10.3
	Non-White	16	5 (31.3)	11.0 to 58.7	9.3	4.0 to NE
Region	North America	30	5 (16.7)	5.6 to 34.7	5.2	3.8 to 10.3
	Japan	16	6 (37.5)	15.2 to 64.6	10.7	4.0 to NE
ECOG PS	0	16	7 (43.8)	19.8 to 70.1	NE	8.7 to NE
	≥1	30	4 (13.3)	3.8 to 30.7	4.6	3.8 to 6.0
Hemoglobin at baseline, g/dL	<10	7	0	0 to 41.0	4.1	0.8 to NE
	≥10	39	11 (28.2)	15.0 to 44.9	8.7	5.1 to 10.7
Previous lines of therapy	<Three lines	15	5 (33.3)	11.8 to 61.6	9.3	3.3 to NE
	≥Three lines	31	6 (19.4)	7.5 to 37.5	5.7	4.0 to 10.7
Previous lines of therapy in metastatic setting	<Three lines	16	4 (25.0)	7.3 to 52.4	8.7	3.3 to NE
	≥Three lines	24	5 (20.8)	7.1 to 42.2	5.7	4.0 to 10.7
Previous use of taxane	Yes	34	6 (17.6)	6.8 to 34.5	5.1	4.0 to 10.3
	No	12	5 (41.7)	15.2 to 72.3	NE	3.3 to NE
Best response to previous systemic therapy	Respondera	25	8 (32.0)	15.0 to 53.5	6.0	4.0 to 10.7
	Nonrespondera	20	3 (15.0)	3.2 to 37.9	6.0	3.8 to 14.2
Best response to most recent therapy	Respondera	9	3 (33.3)	7.5 to 70.1	8.7	0.8 to NE
	Nonrespondera	33	8 (24.2)	11.1 to 42.3	6.0	4.4 to 14.2
Best response to previous PD-1/PD-L1	Respondera	7	3 (42.9)	9.9 to 81.6	NE	2.3 to NE
	Nonrespondera	37	8 (21.6)	9.8 to 38.2	5.7	4.1 to 10.3
Best response to previous platinum	Respondera	24	8 (33.3)	15.6 to 55.3	5.5	3.9 to 10.7
	Nonrespondera	16	3 (18.8)	4.1 to 45.7	10.3	4.1 to 14.2
Best response to previous taxane	Respondera	12	3 (25.0)	5.5 to 57.2	5.5	1.3 to 10.7
	Nonrespondera	19	3 (15.8)	3.4 to 39.6	5.7	2.8 to 14.2
Ad hoc analyses
≤2 previous lines of systemic therapy and one previous line of cytotoxic therapy	Yes	13	5 (38.5)	13.9 to 68.4	11.4	2.5 to NE
	No	33	6 (18.2)	7.0 to 35.5	6.0	4.3 to 9.3
Previous lines of cytotoxic therapy	1	21	7 (33.3)	14.6 to 57.0	11.4	4.1 to NE
	>1	25	4 (16.0)	4.5 to 36.1	5.7	3.9 to 8.7
Previous lines of systemic therapy	≤2	15	5 (33.3)	11.8 to 61.6	9.3	3.3 to NE
	>2	31	6 (19.4)	7.5 to 37.5	5.7	4.0 to 10.3

Abbreviations: ECOG PS, Eastern Cooperative Oncology Group performance status; NE, not evaluable; ORR, objective response rate; OS, overall survival.

^a Responder includes patients who had complete response or partial response as their best response. Nonresponder includes patients who had stable disease or progressive disease as their best response.

^b CI was estimated using exact method on the basis of binomial distribution (Clopper-Pearson).

^c On the basis of Kaplan-Meier estimate.

Safety

Forty-one patients (89.1%) experienced treatment-related AEs (TRAEs; Table 4). The most common TRAEs of any grade (occurring in >20% of patients) were alopecia (13 patients; 28.3%), fatigue (12 patients; 26.1%), and peripheral sensory neuropathy (11 patients; 23.9%). Sixteen patients (34.8%) experienced grade ≥3 TRAEs, the most common (occurring in ≥one patient) being anemia and decreased neutrophil count (both n = 2; 4.3%). TRAEs led to dose reduction in nine patients (19.6%), dose interruption in 18 patients (39.1%), and withdrawal of treatment in six patients (13.0%; Data Supplement, Table S2). Two patients (4.3%) experienced fatal TRAEs (one an unexplained death at home, the other because of a colonic perforation 32 days after starting EV in a patient with constipation and opioid use for pain). Reported TRAEs of special interest were skin reactions in 21 patients (45.7%; 2.2% grade ≥3), peripheral neuropathy in 15 patients (32.6%; 4.3% grade ≥3), ocular disorders in three patients (6.5%; 0% grade ≥3; all dry eye), and hyperglycemia in two patients (4.3%; 0% grade ≥3; Data Supplement, Table S3).

TABLE 4.

Treatment-Related AEs in the Safety Population

AE	Patients (N = 46), No. (%)
	Any Grade	Grade ≥3
Overall	41 (89.1)	16 (34.8)
Alopecia	13 (28.3)	NAa
Fatigue	12 (26.1)	1 (2.2)
Peripheral sensory neuropathy	11 (23.9)	1 (2.2)
Dysgeusia	9 (19.6)	NAa
Maculopapular rash	8 (17.4)	0
Decreased appetite	7 (15.2)	1 (2.2)
Diarrhea	7 (15.2)	0
Anemia	6 (13.0)	2 (4.3)
Decreased neutrophil count	3 (6.5)	2 (4.3)

NOTE. All patients enrolled and who received study treatment were included in the safety population. Any-grade AE occurring in ≥15% of patients or grade ≥3 AE occurring in >1 patient. AEs are reported on the basis of preferred term.

Abbreviations: AE, adverse event; NA, not available.

^a The Common Terminology Criteria for Adverse Events version 4.03 does not include grade ≥3 events for this AE.

Additional Results

Immunogenicity, PK data, and PROs are reported in the Data Supplement.

DISCUSSION

Despite recent advances in cancer therapy, including immunotherapy, effective treatment options remain limited for patients with R/M HNC. No standard of care exists for patients such as those in the HNC cohort of EV-202 whose tumors have progressed on PBC and PD-1/PD-L1 inhibitors.⁴ Outside of clinical trials, treatment generally consists of single agents not previously received by the patient (eg, taxane, methotrexate, and cetuximab) on the basis of tumor characteristics, but data suggest a lack of corresponding improvement in survival.^4,22 Thus, effective treatments for patients with R/M HNC for whom PBC and PD-1/PD-L1 inhibitors are not providing adequate benefit or are not well tolerated are critically needed.

Confirmed ORR in patients with HNC receiving EV in this study was 23.9%, exceeding the protocol-defined 17.5% threshold to declare promising antitumor activity. The ORR of 23.9% is higher than the response rates reported in studies conducted in similar HNC patient populations. In the CheckMate 141 and KEYNOTE-040 trials, response rates for comparator therapy groups (treated with methotrexate, docetaxel, or cetuximab) were 5.8% and 10.1%, respectively.^8,23,24 Notably, 67.4% of patients in the HNC cohort of EV-202 had received ≥three lines of systemic therapy, whereas 19.9% of patients in CheckMate 141 and 1.2% of patients in KEYNOTE-040 had received ≥three previous lines of therapy.^8,23,24 Median DOR in this cohort was 9.4 months as of an updated data cutoff date (July 15, 2022). The prespecified subgroup analysis (Table 3) indicates that patients in Japan may have improved outcomes with EV compared with patients in North America. However, treatments used for patients in Japan, such as taxanes, may not have been balanced, which may have contributed to differences in outcome. Moreover, although thought-provoking, the limited size of this phase II trial, small size of each subgroup, and lack of balanced baseline characteristics between the groups limit our ability to make conclusions. Nectin-4 expression was detected in all 43 patients with adequate tissue available, and levels were similar between responders and nonresponders.

The safety profile in this cohort was consistent with the well-established safety/tolerability profile of EV monotherapy in la/mUC.^11,13,25 Grade ≥3 TRAEs were reported in 34.8% of patients with HNC in this study and 51.4% of the 296 patients treated with EV in the phase III trial EV-301 in patients with previously treated la/mUC.¹¹ Rates of all-grade TRAEs of special interest (including skin reactions, peripheral neuropathy, and hyperglycemia) were also similar between the HNC cohort of EV-202 and EV-treated patients in EV-301.¹¹ No new safety signals were identified. Similar to the EV trials in la/mUC, AEs were managed with dose interruptions and dose reductions. Education of patients, caregivers, and health care practitioners on early recognition of AEs associated with EV is essential. In addition, proactive monitoring and management of the known AEs associated with EV, including dose modification to manage AEs, are important aspects of caring for patients being treated with EV. Together, these results suggest substantial antitumor activity with well-characterized safety, meriting further investigation of EV in R/M HNC.

Preclinical evidence has demonstrated that EV induces immunogenic cell death and immune cell activation and that combining EV with a PD-1 inhibitor enhances antitumor activity.²⁶ Moreover, in patients with previously untreated la/mUC in the EV-302 trial, the combination of EV plus pembrolizumab led to improved OS, PFS, and ORR compared with PBC treatment.²⁰ These findings support the evaluation of EV plus pembrolizumab in patients with HNC, and an additional cohort of EV-202 investigating EV in combination with pembrolizumab as first-line therapy in patients with R/M HNSCC with a PD-L1 combined positive score of ≥1 is ongoing (ClinicalTrials.gov identifier: NCT04225117).¹³

A limitation of this study is the small sample size, which may affect interpretation of efficacy analyses (particularly across subgroups) and reduce the ability to detect AEs or overemphasize potentially rare AEs. An adequate analysis of efficacy on the basis of HPV status among patients with oropharyngeal squamous cell carcinoma was not possible because many patients were reported to have pharyngeal or tongue cancer without specification of whether the tumor was oropharyngeal or nonoropharyngeal. This is important, as HPV positivity is increasingly recognized as a prognostic biomarker associated with improved survival in patients with oropharyngeal cancer and increased sensitivity to cytotoxic chemotherapies, potentially including EV.^4,27 Another limitation is that this was a single-arm study; therefore, results are currently compared with historical controls, and randomized trials are needed. Furthermore, historical control data are largely based on clinical trials enrolling patients who had never received PD-1/PD-L1 inhibitor therapy, while patients in the EV-202 HNC cohort had all received PD-1/PD-L1 inhibitor therapy. Retrospective data sets involving patients following PD-1/PD-L1 inhibitor therapy and receiving currently available therapy are limited in number and quality.

In summary, treatment with single-agent EV demonstrated clinically meaningful antitumor activity with a well-characterized safety profile in patients with R/M HNC who previously received platinum-based and PD-1/PD-L1 inhibitor therapy. These findings support further evaluation of EV in HNC.

DATA SHARING STATEMENT

Researchers may request access to anonymized participant-level data, trial-level data, and protocols from Astellas-sponsored clinical trials at www.clinicalstudydatarequest.com. For the Astellas criteria on data sharing, see https://clinicalstudydatarequest.com/Study-Sponsors/Study-Sponsors-Astellas.aspx.

AUTHOR CONTRIBUTIONS

Conception and design: Paul L. Swiecicki, Justine Yang Bruce, Michael Mihm, Seema Gorla

Provision of study materials or patients: Paul L. Swiecicki, Emrullah Yilmaz, Ari Joseph Rosenberg, Takao Fujisawa, Justine Yang Bruce, Yongyun Zhao, Jessica L. Geiger

Collection and assembly of data: All authors

Data analysis and interpretation: All authors

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Phase II Trial of Enfortumab Vedotin in Patients With Previously Treated Advanced Head and Neck Cancer

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jco/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).