Safety and efficacy of CDX-014, an antibody-drug conjugate directed against T cell immunoglobulin mucin-1 in advanced renal cell carcinoma

Summary

CDX-014 is an antibody-drug conjugate directed against TIM-1, a surface marker highly expressed in renal cell carcinoma (RCC) and ovarian carcinoma. This phase I, first-in-human trial was conducted to evaluate the safety and preliminary activity of CDX-014 in patients with advanced refractory RCC, following a dose-escalation and dose expansion design. CDX-014 was administered intravenously at doses ranging from 0.15 to 2.0 mg/kg every 2 or 3 weeks until progression or unacceptable toxicity. Sixteen patients received at least one dose of CDX-014. The maximum tolerated dose was not identified. Most frequent adverse grade 1 or 2 adverse events included nausea (38%), fatigue, alopecia, elevation of AST and decreased appetite (25% each). Adverse events of grade 3 or more included hyperglycemia (19%), urosepsis (6%), and one multi-organ failure (6%) responsible for one treatment-related death. Two patients discontinued therapy for adverse events including fatigue grade 2 and urosepsis grade 4. CDX-014 showed antitumor activity with one prolonged partial response and a clinical benefit rate (objective response or stable disease >6 months) of 31%. The two patients that exhibited the most marked tumor shrinkage had high TIM-1 expression on tumor tissue. Overall, CDX-014 exhibited a manageable toxicity profile and early signs of activity, supporting further evaluation of antibody-drug conjugates in patients with advanced RCC and potentially other TIM-1 expressing cancers. Trial registration https://clinicaltrials.gov/ct2/show/NCT02837991 NCT02837991; July 20, 2016.

Introduction

Outcomes of patients with renal cell carcinoma (RCC) have improved since the advent of antiangiogenic therapies and immune checkpoint inhibitors; with combinations of immunotherapies alone or with vascular-endothelial growth factor receptor (VEGFR) inhibitors approved in the first-line setting given phase 3 data showing superiority over the tyrosine kinase inhibitor (TKI) sunitinib [1-3]. Most patients with metastatic disease will eventually progress on therapy. Unfortunately, therapeutic options are limited after failure of VEGFR-targeted therapies and immune checkpoint inhibitors, and identification of new therapeutic targets is essential.

Antibody-drug conjugates allow for ability to target high concentrations of cytotoxic chemotherapy to tumor sites using a monoclonal antibody as a delivery vehicle. They are emerging as a new standard of care in other genitourinary malignancies with drugs such as enfortumab vedotin targeting nectin-4 universally expressed on urothelial carcinoma, recently approved by the FDA in December 2019 [4]. Similar target antigens in RCC are needed. T cell ImmunoglobulinMucin-1 (TIM-1), also known as Kidney Injury Molecule 1 (KIM-1), is a transmembrane glycoprotein that is upregulated in tubular cells following ischemia [5]. Expression of TIM-1 notably allows tubular cells to detect and phagocytize apoptotic bodies in the context of kidney injury [6]. Considering that such tubular cells are involved in specific dedifferentiation programs that favor proliferation in response to acute kidney injury, TIM-1 has also been considered as a marker of dedifferentiated tubular cells [7]. Recent insights demonstrated that TIM-1 was detectable in tumor tissue in more than 70% of patients with clear cell or papillary RCC, consistent with dedifferentiation and proliferation processes in tumor cells arising from the proximal tubule [8]. Conversely, TIM-1 expression levels were low in normal renal cortex (10%), as well as other solid solid tumors [9]. Therefore, TIM-1 has been considered as a selective biomarker for RCC diagnosis, as well as a potential putative target for an ADC therapeutic approach.

CDX-014 is an antibody-drug conjugate (ADC) consisting of a human immunoglobulin G1 directed against TIM-1, covalently linked with the antimitotic monomethyl auristatin E (MMAE) via standard maleimide conjugation [10]. CDX-014 binds to TIM-1 expressing cells and upon internalization, proteolysis of the valine-citrulline linker releases MMAE, leading to cytotoxic death through inhibition of tubulin polymerization [11]. In vitro and in vivo activity has been demonstrated in TIM-1 expressing tumor cell lines [11]. Herein we evaluate CDX-014 in a phase I trial in patients with advanced RCC who failed prior VEGFR-directed therapy.

Materials and methods

Patients

Patients with advanced clear cell or papillary RCC who experienced progression after at least two lines of systemic therapy, including at least one TKI were included. Additional main inclusion criteria included measurable disease by Response Criteria in Solid Tumors (RECIST) 1.1 criteria, a Karnofsky performance status ≥70%, and adequate renal, hepatic, and hematological laboratory values. Archival or pre-treatment formalin-fixed paraffin embedded (FFPE) tumor tissue were required for TIM-1 expression assessment. Patients with history of brain metastases were excluded unless previously treated and asymptomatic for at least 2 months. The full list of inclusion criteria is availiable at https://clinicaltrials.gov/ct2/show/NCT02837991.

Procedures

This trial was designed as a multicenter phase I, open-label, single arm study, and included a dose escalation phase followed by a dose expansion phase. Patients would be treated with one of the following doses of intravenous CDX-014: 0.15, 0.30, 0.60, 1.2, 1.8, 2.0, 2.2, and 2.4 mg/kg of body weight every 3 weeks. After 14 patients enrolled, protocol was amended to treat with escalating doses of 1.0, 1.2, 1.4, 1.6 and 1.8 mg/kg every 2 weeks. Treatment was administered until radiological progression, unacceptable toxicity, pregnancy, withdrawal of consent, decision of discontinuation by the physician, or death. Dose reductions were allowed following guidelines provided in the Data Supplement.

Toxicity was recorded at each visit according to the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI-CTCAE) v4.0. Disease assessments were performed every 6 weeks for 6 months, then every 8 weeks by computed tomography of the chest, abdomen, and pelvis, according to the RECIST 1.1 guidelines. Pharmacokinetic assessments were used to monitor exposure to free MMAE levels at each planned visit. TIM-1 expression levels were analyzed on archival or pre-treatment tumor tissue by immunohistochemistry and scored according to staining intensity and percentage of cells expressing TIM-1.

This trial was approved by the Institutional Review Boards of each participating institution and conducted in accordance with the principles of the Declaration of Helsinki and the International Conference on Harmonization.

Study objectives

The primary objective of the dose escalation phase was to determine the maximum tolerated dose (MTD) of CDX-014 in advanced refractory RCC. The primary objective of the dose expansion phase was to determine preliminary antitumor activity of CDX-014 at the MTD in this population as measured by the objective response rate (ORR).

Secondary objectives included tolerability, pharmacokinetics, and additional efficacy assessments: clinical benefit rate, duration of response, progression-free survival, and overall survival. Exploratory objectives included determination of TIM-1 expression in tissue and correlations with outcomes.

Clinical benefit rate was defined as the rate of complete response (CR), partial response (PR) and stable disease (SD) of at least 6 months. PFS was defined as time from therapy initiation to progression or death. Patients who are event-free were censored at the date of last evaluation for PFS. Overall survival (OS) was defined as time from therapy initiation to death or last contact for patients alive.

Statistical design

In the dose escalation phase, a minimum of 6 patients would be treated at the MTD prior to dose expansion. The first four dose levels of CDX-014 (0.15 to 1.2 mg/kg every 3 weeks) were to include a single patient each, and dose escalation would proceed in the absence of dose-limiting toxicity (Data Supplement). Further dose levels as would follow a standard 3 + 3 dose escalation scheme. A total of 56 patients were expected to be enrolled in the dose escalation phase. The dose expansion phase would include approximately 15 patients evaluable for response, allowing an estimation of the ORR with a 95% confidence interval (95%CI) width not greater than 0.52 based on the Clopper Pearson method.

All patients who receive at least one dose of CDX-014 were evaluable for safety. Patients also had at least one post-baseline disease assessment, or discontinued therapy due to death or clinical deterioration.

Descriptive statistics were used to report baseline characteristics of the population, response data, and safety analyses. Median and range were used to describe quantitative data, frequency, and proportion for qualitative data. Survival analyses were performed using Kaplan-Meier estimators and reported with 95% CIs. For exploratory biomarker analyses, patients were divided into two groups based on median TIM-1 expression on tumor tissue; Fisher and log-rank tests were respectively used to compare response rates and survival according to TIM-1 expression. All analyses were performed using NCSS Statistical Software (2019), NCSS, LLC. Kaysville, Utah.

Results

Patients

Sixteen patients were enrolled in the study and received at least one dose of CDX-014 between 7/20/2016 to 3/21/2018. Baseline characteristics of the patients are reported in Table 1. Patients were predominantly male (94%), with a median age of 67 years, and most (57%) had a Karnofsky performance status of 90% or 100%. Most patients had clear cell RCC (14/16, 88%), while 2/16 (13%) had papillary RCC, and 75% (12/16) belonged to intermediate or poor risk prognostic groups according to the IMDC criteria. The study population was heavily pretreated, with a median of 4 previous systemic therapies: VEGFR-directed therapies in all patients, immune checkpoint inhibitors in 94%, and mTOR inhibitors in 44%. Median followup was 15.2 months (95%CI 12.9–17.6). Twelve (75%) patients experienced progression or death at data cutoff on 11/26/2018, three did not experience progression and were still alive on therapy, and one was lost to follow-up.

Table 1.

Baseline characteristics of the patients

Characteristic	Patients (N = 16)
Median age, years (range)	67 (56–84)
Sex
Female	1 (6%)
Male	15 (94%)
Histology
Clear-cell	14 (88%)
Papillary	2 (13%)
Median number of previous therapies (range)	4 (2–9)
Type of previous therapies
Immune checkpoint inhibitor	15 (94%)
VEGF/VEGFR-directed therapy	16 (100%)
MTOR inhibitor	7 (44%)
Karnofsky performance status
70	2 (13%)
80	5 (31%)
90	7 (44%)
100	2 (13%)
IMDC risk groups
Favorable	4 (25%)
Intermediate	8 (50%)
Poor	4 (25%)

Risk groups are defined according to the following risk factors: time from diagnosis to initiation of systemic therapy <1 year, hemoglobin < lower limit of normal, neutrophils > upper limit of normal, platelets > upper limit of normal, corrected calcium > upper limit of normal, Karnofsky performance status <80%

IMDC International metastatic renal cell carcinoma database

Safety

Patients were enrolled at increasing dose levels up to 2.0 mg/kg every 3 weeks and 1.2 mg/kg every 2 weeks (Fig. 1). Patients received a median of 4 cycles (range 1–28) of CDX-014 (Supplementary Table 1). Discontinuation of CDX-014 development led to early termination of the trial. The MTD of CDX-014 has not been reached at assessed dose levels.

Fig. 1.

CONSORT diagram and Enrollment summary

There were two dose-limiting toxicities, both occurring at a dose level of 2.0 mg/kg every 3 weeks. One patient experienced grade 3 rash, grade 3 liver dysfunction, grade 4 neutropenia, and ultimately grade 5 fatal multi-organ failure, after receiving one dose of CDX-014. Of note, this grade 5 adverse event occurred in a context of unrelated grade 1 acute kidney injury prior to the event, probably due to dehydration and/or administration of iodine-based radiocontrast. A second patient experienced a dose limiting toxicity of grade 3 hyperglycemia in the context of known diabetes mellitus at this dose level, which resolved with appropriate supportive therapy. The study protocol was amended to allow further dose exploration despite this second dose-limiting toxicity.

Most common adverse events were nausea (6/16, 38%), fatigue, alopecia, elevation of AST, and decreased appetite (4/16, 25% each), all of which were grade 1 or 2. In addition to previously mentioned adverse events, other grade 3 or 4 toxicities included grade 3 hyperglycemia (3/16, 19%) in patients treated at 1.2, 1.8 and 2.0 mg/kg, as well as one occurrence of grade 4 urosepsis after 5 doses of CDX-014 at 1.8 mg/kg in a patient treated with history of multiple urinary tract infections (Table 2). Two patients discontinued therapy with CDX-014 for toxicity: one for grade 4 urosepsis as described above, and one for grade 2 fatigue.

Table 2.

Treatment-related adverse events

Adverse event	Grade 1–2	Grade 3–4	Total
Nausea	6 (38%)	0	6 (38%)
Fatigue	4 (25%)	0	4 (25%)
Alopecia	4 (25%)	NA	4 (25%)
AST Increased	4 (25%)	0	4 (25%)
Decreased Appetite	4 (25%)	0	4 (25%)
Rash	2 (13%)	1 (6%)	3 (19%)
Neuropathy	3 (19%)	0	3 (19%)
ALT Increased	3 (19%)	0	3 (19%)
Diarrhea	3 (19%)	0	3 (19%)
Hyperglycemia	0	3 (19%)	3 (19%)
Neutropenia	1 (6%)	1 (6%)	2 (13%)
Blood Creatinine Increased	2 (13%)	0	2 (13%)

Reported adverse events occurred in at least 10% of patients. Data presented as N (%). One patient died from multi-organ failure, potentially treatment-related, after one dose of CDX-014 at 2.0 mg/kg

Pharmacokinetics

Free MMAE levels in peripheral blood were assessed at cycle 1 (Supplementary Table 2) and following cycles. There was a trend towards correlation between CDX-014 dosing and free MMAE Cmax after cycle 1 (Spearmann r = 0.48, p = 0.08). Free MMAE Cmax after cycle 1 was highest in the study patient who experienced grade 5 multi-organ failure suggestive of poor renal clearance of the MMAE [12].

Activity

CDX-014 activity was ultimately assessed in all patients who received at least one dose of therapy. One patient (6%) treated at 0.3 mg/kg every 3 weeks exhibited PR as best overall response (Fig. 2), ongoing after 17 months on therapy; this patient had been on single agent PD-1 inhibition prior to initiating therapy with CDX-014. Five patients (31%) exhibited clinical benefit from CDX-014. PFS and OS were 2.7 months (95%CI 1.2–8.0) and 12.6 months (95%CI 5.7–12.6), respectively.

Fig. 2.

Best overall RECIST response to CDX-014. CDX-014 dose [mg/kg] and TIM-1 expression [% tumor cells positive] are displayed in parenthesis for each patient. All patients are Q3W schedule unless noted as Q2W

Expression of TIM-1 and response to therapy

Most patients (14/16, 88%) had detectable expression of TIM-1 on tumor cells from archival tissue. There was no statistical correlation between expression of TIM-1 and response, clinical benefit, or survival. However, both patients with the largest decrease in tumor burden had strong TIM-1 expression, including one patient with a prolonged PR and one patient with SD that experienced 29% tumor shrinkage (Fig. 3).

Fig. 3.

Response by “spider plots”

Discussion

CDX-014 demonstrated a manageable safety profile and modest evidence of antitumor activity in this highly pretreated RCC population, with 3/16 pts. (19%) remaining alive without progression at data cutoff. The observed toxicity profile is compatible with those of ADCs, including chemotherapy-like adverse events including fatigue, nausea, and cytopenia. Hyperglycemia was recorded in 3 patients (19%), and may represent a class-dependent adverse event though at this time, the mechanism is unclear.

There was one event of multi-organ failure of unknown origin deemed related to the study drug. TIM-1 is notably overexpressed not only in RCC tissue, but also in regenerating tubular cells in a context of acute kidney injury [13], and it has been suggested that measuring the level of TIM-1 in the urine can be a sensitive test for the early diagnosis of acute kidney injury [14]. TIM-1 expressing tubular epithelial cells phagocytose apoptotic cells to reduce inflammation and innate immune response, but its expression is also involved in the repair process where it promotes migration of renal tubular epithelial cells [15, 16]. The fatal adverse event occurred after one dose of treatment in a patient who experienced initial grade 1 acute kidney injury in the setting of recent CT scan with contrast. While it is unknown whether this patient had increased expression of TIM-1 in their normal kidney, increased TIM-1 may have led to increased binding of CDX-014 to non-neoplastic kidney tissue, begetting further kidney injury and ultimately multisystem organ failure and death similar to what is seen with a hyperacute rejection to an allogeneic renal transplant [17]. Unfortunately, autopsy was not able to be performed. The high level of free MMAE in this patient was also a likely contributing factor to the toxicity.

Grade 3 or 4 hyperglycemia was noted in 3 patients, and the mechanism for this remains unclear but it is seen with other antibody drug conjugates such as enfortumab vedotin which utilize the same cytotoxic payload MMAE [4, 18]. These toxicities highlight the need to find RCC-specific biomarkers, especially as ADCs exhibit prolonged half-life compared to small molecules, but shorter than unconjugated antibodies [19].

Further biomarker studies are needed to identify appropriate ADC targets in the context of RCC. TIM-1 plays a role outside of the kidney, notably in immune dysfunction diseases such as systemic lupus erythematosus, and it has been linked with Th2 immune response in asthma [18, 20]. As such, with the pivotal role immunotherapy now plays in the management of RCC, there is potential for significant interactions with preceding therapies.

Major advances have been made with ADCs in other genitourinary malignancies, primarily enfortumab-vedotin [4, 21] and sacituzumab-govitecan [22] in urothelial carcinoma, which demonstrated response rates of 55% and 31%, respectively, in highly refractory patients. Other ADCs have been investigated in RCC. The CD70-targeted ADC SGN-75 administered every 3 weeks showed prolonged disease control in up to 44% of patients with CD70-positive refractory RCC (N = 32) [23] though no further studies have been performed to date. The ENPP3-targeted ADC AGS-16C3F showed responses in 23% of patients with refractory RCC at 1.8 mg/kg (N = 13) [24]. This led to a phase 2 trial (NCT02639185) comparing AGS-16C3F to axitinib in patients with ccRCC or nccRCC expressing ENPP-3 which has completed accrual. Notably, unlike CDX-014, both SGN-75 and AGS-16C3F employ a non-cleavable linker between antibody and drug payload.

The success of ADCs in urothelial carcinoma is intriguing, but this is in the setting of a disease where cytotoxic chemotherapy remains the standard; combination of cisplatin with gemcitabine can achieve ORR approaching 50% [25]. While there is clearly a role for cytotoxic chemotherapy in the management of certain RCC histologies such as medullary carcinoma or collecting duct, there is minimal role in metastatic ccRCC. Chemotherapy plays a role in in aggressive subsets of RCC, including notably RCC with sarcomatoid dedifferentiation [26], where small studies have shown efficacy for doxorubicin and gemcitabine [27] or the combination of gemcitabine with sunitinib [28] in those patients, though of limited duration. Our data using a targeted approach, though limited, suggest delivering high doses of chemotherapy with an ADC targeting TIM-1 showed signs of anti-tumor activity.

Bringing new therapeutic avenues to RCC is essential, as options after failure of immune checkpoint inhibitors and VEGFR-directed TKIs are limited. CDX-014 is an ADC targeting TIM-1, a novel marker in RCC. The data was ultimately insufficient to make any judgement based on clinical activity. There are ongoing discussions with multiple companies to continue development of CDX- 014. In examining the relationship of high TIM-1 expression to response and response duration, there did appear to be a relationship between these parameters and high TIM-1 expression, however, a group of patients with similar high expression in their tumors experienced rapid progression on treatment (Figs. 2 and 3). This suggests that simply having high TIM-1 expression is insufficient for efficacy, with other factors such as cellular sensitivity to and intratumoral disposition of MMAE playing a potential role. Other ADCs are being explored, though it remains unknown how they will interact with approved therapy, including TKIs and immunotherapies, or if the chemotherapy delivered will have a beneficial effect given inherent resistance of ccRCC to cytotoxic chemotherapy. Furthermore, the dynamic nature and heterogeneity of target expression may affect ADC efficacy and would have to be specifically explored in RCC. Identification of specific targets could may also be relevant to the development of bispecific antibodies that could be used as immunotherapy strategies to direct immune cells towards tumor cells, aiming at overcoming mechanisms of resistance to conventional immune checkpoint inhibitors [29].

Conclusion

This first-in-human trial of CDX-014 demonstrated a novel approach to the management of refractory RCC. CDX-014 had a manageable safety profile and some patients experienced anti-tumor activity. Ongoing work is necessary to determine its clinical activity and optimal dosing strategy.