A PHASE 2 STUDY OF ONTUXIZUMAB, A MONOCLONAL ANTIBODY TARGETING ENDOSIALIN, IN METASTATIC MELANOMA

Abstract

Objectives:

Ontuxizumab (MORAB-004) is a first-in-class monoclonal antibody that interferes with endosialin function, which is important in tumor stromal cell function, angiogenesis, and tumor growth. This Phase 2 study evaluated the 24-week progression-free survival (PFS) value, pharmacokinetics, and tolerability of 2 doses of ontuxizumab in patients with metastatic melanoma.

Patients and Methods:

Patients with metastatic melanoma and disease progression after receiving at least 1 prior systemic treatment were randomized to receive ontuxizumab (2 or 4 mg/kg) weekly, without dose change, until disease progression.

Results:

Seventy-six patients received at least 1 dose of ontuxizumab (40 received 2 mg/kg, 36 received 4 mg/kg). The primary endpoint, 24-week PFS value, was 11.4% (95% Confidence Interval [CI]: 5.3%−19.9%) for all patients (13.5% for 2 mg/kg and 8.9% for 4 mg/kg). The median PFS for all patients was 8.3 weeks (95% CI: 8.1–12.3 weeks). One patient receiving 4 mg/kg had a partial response, as measured by Response Evaluation Criteria in Solid Tumors v1.1. Twenty-seven of 66 response evaluable patients (40.9%) had stable disease. The median overall survival was 31.0 weeks (95% CI: 28.3–44.0 weeks). The most common adverse events overall were headache (55.3%), fatigue (48.7%), chills (42.1%), and nausea (36.8%), mostly grade 1 or 2.

Conclusions:

Ontuxizumab at both doses was well tolerated. The 24-week PFS value was 11.4% among all ontuxizumab-treated patients. The overall response rate was 3.1% at the 4 mg/kg dose, with clinical benefit achieved in 42.4% of response evaluable patients. Efficacy of single-agent ontuxizumab at these doses in melanoma was low.

INTRODUCTION

The recent approval of agents that target intracellular signaling pathways, such as BRAF-inhibitors (vemurafenib, dabrafenib) and MEK-inhibitors (trametinib, cobimetinib), as well as checkpoint blockade inhibitors directed at cell surface receptors such as PD1 (nivolumab, pembrolizumab) or CTLA-4 (ipilimumab) have resulted in markedly improved response rate, progression-free survival (PFS), and overall survival in patients with metastatic melanoma [1–10]. In addition, various combinations of these drugs have extended the efficacy observed with single therapy [11, 12]. Despite these achievements, melanoma cures are far from universal and novel treatment strategies for patients with this disease are needed.

One novel therapeutic strategy is the interference of endosialin-mediated activity given the key role of endosialin in tumor growth and angiogenesis in various tumor types. Endosialin, also known as tumor endothelial marker-1 (TEM-1) or CD248 [13, 14], is a glycoprotein present on the cell surface of tumor-associated pericytes involved in tumor vasculature, as well as on the surface of certain cancer cells, particularly those of mesenchymal origin such as sarcoma and melanoma. Endosialin is broadly expressed in cells localized to the tumor stroma and vasculature in many carcinomas and sarcomas, while relatively absent in normal tissues [13, 15–20]. A role for blockade of endosialin-mediated activity in metastatic melanoma was postulated as a potential treatment because of the presence of endosialin in melanoma cells [15–19, 21]. Huber et al found endosialin expressed in 14 of 15 cutaneous melanoma metastases, including tumor cells, pericytes, and tumor associated fibroblasts [19, 20].

Endosialin is believed to influence tumor growth by modulating communication between tumor cells and stromal cells and pathways involved in organization and initiation of new tumor vessel formation [22]. Specifically, pericytes are believed to stimulate tumor vascularization via endosialin-mediated signaling pathways, especially platelet-derived growth factor receptor-beta (PDGFR-β), resulting in endothelial cell differentiation into new vascular networks [23]. In preclinical studies using knockout experiments, elimination of endosialin prevented blood vessel formation in angiogenesis assays [24] and suppressed pericyte proliferation [25], tumor blood vessel development, and tumor growth.

Ontuxizumab (MORAB-004) is a first-in-class humanized immunoglobulin G subtype 1-kappa (IgG1/κ) monoclonal antibody specific for endosialin. Results of a Phase 1 dose escalation study in patients with advanced solid tumors demonstrated that ontuxizumab could be safely administered at weekly doses as high as 12 mg/kg [26]. A favorable safety profile was observed, with evidence of disease stabilization.

The present study is the first to assess the efficacy of ontuxizumab in patients with metastatic melanoma. The purpose of this study was to evaluate efficacy and safety of ontuxizumab administered at two dose levels in patients with metastatic melanoma.

METHODS

Patients:

Patients with disease progression after receiving at least 1 prior systemic treatment (chemotherapy, targeted therapy, immunotherapy, or combinations thereof) for metastatic melanoma were eligible to enroll. Additional eligibility criteria included at least 18 years of age, measurable disease (as defined by Response Evaluation Criteria in Solid Tumors [RECIST], v 1.1) [27], a life expectancy of at least 3 months, and an Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1. At least 21 days between the last systemic anticancer treatment and the first infusion of ontuxizumab must have passed. Patients were required to have adequate hematologic and coagulation parameters and no other active malignancy or evidence of brain metastases. At screening, disease stage and histologic type were obtained. All patients provided informed consent before initiation of treatment. This study was reviewed and approved by the local ethics committee before initiation of any procedures.

Study design and treatment:

This was a Phase 2, multicenter, open-label, dose-selection, proof-of-concept study to evaluate safety, efficacy, and pharmacokinetics of ontuxizumab in patients with metastatic melanoma. Patients were randomly assigned to receive treatment with either 2 mg/kg or 4 mg/kg ontuxizumab, administered intravenously on Days 1, 8, 15, and 22 of a 28-day cycle. This treatment cycle could be repeated, without dose escalation or reduction, until disease progression, unacceptable toxicity, or discontinuation for any reason (Fig. 1).

Fig. 1.

The study design with the number of patients screened, randomized, followed up, and analyzed

Efficacy assessment:

Target and non-target lesions were measured at baseline using computed tomography (CT) or magnetic resonance imaging (MRI) of the chest, abdomen, and pelvis. Lesions were re-measured every 8 weeks by RECIST, v 1.1 criteria by the investigative site. After discontinuation of treatment, patients were followed for documentation of additional anticancer therapies and survival.

The primary efficacy endpoint was PFS value at 24 weeks, as a means to compare response to similar patients treated with ipilimumab [9]. Secondary efficacy endpoints included PFS value at 16 and 52 weeks, overall survival, overall response rate, and time to response.

Safety:

Secondary objectives included evaluation of safety and tolerability. Safety assessments were made throughout the study and included adverse events that were treatment-emergent (from first dose to 45 days after the last dose of study drug), drug hypersensitivity events, clinical laboratory values, physical examinations, electrocardiographic data, and development of anti-drug antibody (ADA). Adverse events were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE), version 4.03. Adverse events that were Grade 4 or 5 were reported as serious adverse events. The causality of adverse events was graded as not related or related to the study drug as determined by the investigator.

Statistics:

Progression-free survival was defined as the time in weeks from the date of randomization to the date of the first observation of progression, according to RECIST, v 1.1, or date of death. Progression-free survival was summarized descriptively and graphically using Kaplan-Meier estimates. The primary endpoint of 24-week PFS value was estimated using the Kaplan-Meier method. Statistical testing of all efficacy endpoints was conducted using a 2-sided 5% level of significance with no adjustment for multiple endpoints.

A PFS response value of clinical interest was targeted as a value of ≥35% at 24 weeks based on a study with ipilimumab; however 69.7% of the patient population in our study failed prior ipilimumab therapy making the comparison more difficult to interpret [9]. The primary comparison of interest was the comparison of the 24-week PFS value in all patients with the pre-specified PFS value of 35%. The sample was chosen so that the 2 dose groups combined could yield PFS responses that differentiated between an uninteresting response (≤35% PFS value at 24 weeks) and a clinically meaningful response (≥50% PFS value at 24 weeks). To test this hypothesis, 72 patients were required based on a binomial exact test with 1-sided α=0.05 and ß=0.2. A 2-sided 95% confidence interval (CI) for the estimated value was constructed using the log-log transformation [28].

The secondary comparison of interest was the comparison of the PFS value between the 2 dose groups to assess for any dose-response relationship. Comparisons between the 2 dose groups utilized a stratified log-rank test with M stage (M1a/M1b vs M1c) as the stratification variable. Patients who were alive with no disease progression had their PFS time censored at the date of their last tumor assessment. The 95% CIs for median PFS were constructed [29].

Overall survival, defined as the time in weeks from the date of randomization to the date of death regardless of cause, was summarized descriptively and graphically using Kaplan-Meier estimates overall and for each treatment group. The median overall survival CIs were constructed as for PFS. Patients alive at the time of the analysis had the survival time censored at the date of the last study follow-up.

Overall response rate was defined as the proportion of patients with a complete response or a partial response based on RECIST v1.1. The relationship between baseline biomarker levels (tumor endosialin and PDGFR-β and serum endosialin) and clinical response parameters (best overall response, PFS, and overall survival) were analyzed using a logistic regression model and a proportional hazard model (Cox regression). Safety data were summarized descriptively by dose group and for all treated patients combined.

Statistical analyses were performed using SAS® version 9.1 or higher. All study analyses in this report include all data, with a cutoff date of 02 Dec 2013.

Pharmacokinetics:

Serum samples for pharmacokinetic analysis were obtained on Days 1, 8, 15, and 22 before and after the end of the study drug infusion during Cycle 1 and once on Days 3 and 4 of Cycle 1. During Cycle 1, an additional sample was collected 2 hours after the infusion on Days 1 and 22. After Cycle 1, samples were collected before and after study drug administration on Days 1 and 15 only. An additional sample was collected at the final visit.

Ontuxizumab concentrations were measured using an endosialin-antigen based electrochemiluminescent immunoassay to capture and quantify the serum concentration of free/partially complexed ontuxizumab [26].

Biomarker studies:

Tissue endosialin expression was assessed on tumor biopsy samples (paired sampling) obtained either at screening or from archival blocks and between Days 22 and 28 of Cycle 1 in a subset of 30 patients. PDGFR-β tumor tissue expression was assessed at screening only.

Formalin-fixed-paraffin-embedded slides were prepared and assessed for endosialin and PDGFR-β expression by immunohistochemistry using the previously described rat anti-human endosialin 9G5 monoclonal antibody [30] or clone 28E1 (Cell Signaling, Danvers, MA, USA) for PDGFR-β [20, 26] by Quintiles Laboratories Americas (Marietta, GA and Westmont, IL). Expression was reported as +1, +2, or +3 as assessed by a single board certified pathologist.

Baseline serum samples were collected for evaluation of soluble endosialin using electrochemiluminescent assays performed by Frontage Laboratories, Inc. (Exton, PA) using endosialin-specific monoclonal antibodies [30].

Anti-drug antibodies:

Serum samples for ADA analysis were collected before administration of study drug on Day 1 and Day 15 of Cycle 1 and all subsequent cycles and at the final visit. ADA assays were performed using an immunoassay bridging format based on the formation of ontuxizumab and ADA complexes detected in a semi-quantitative manner via electrochemiluminescence [26].

Patients with evaluable ADA data and at least 1 sample obtained after study drug administration were included in the analysis. Patients were considered positive for ADA if at least 1 ADA positive sample occurred at any time after study drug administration or, for baseline ADA positive patients, at least a 4-fold increase in ADA titer occurred after study drug administration compared to baseline values.

RESULTS

Disposition:

A total of 76 patients from the United States, Australia, Germany, and the United Kingdom were randomized and received at least 1 dose of ontuxizumab (40 in the 2 mg/kg dose group and 36 in the 4 mg/kg dose group). All 76 patients were included in the safety analysis population and the pharmacokinetic analysis population. One patient in the 2 mg/kg group had an active malignancy (Merkel cell carcinoma) requiring treatment within the past 5 years, therefore did not meet key eligibility requirements and was excluded from the primary efficacy population. Patients included in the overall response rate evaluation were required to have a baseline evaluation and at least 1 on-study radiologic evaluation as assessed by RECIST, v1.1 (overall response rate population). Sixty-six of the 75 (86.8%) patients in the primary efficacy population were included in the overall response rate population. The study was initiated 16 May 2011 and the data cutoff date was 02 Dec 2013.

The most common reason for treatment discontinuation was progressive disease (72.5% in the 2 mg/kg group and 72.2% in the 4 mg/kg group).

Demography and baseline characteristics:

The majority of patients were male (64.5%), all but 3 patients were white (96.1%), and the mean age was 63 years (Table 1) The majority of patients (68.4%) were Stage M1c at study entry. Of note, there were imbalances between the 2 treatment groups for baseline characteristics. The 2 mg/kg treatment group had more patients with an ECOG performance status of 1 (55.0% vs 30.6%) enrolled compared with the 4 mg/kg group, a negative prognostic factor. However, the 4 mg/kg group had more males (75% vs 55%), a higher percentage of patients with Stage M1c (77.8% vs 60.0%), a younger mean age (61.2 years vs 65.1 years), and more patients with a lactate dehydrogenase (LDH) value greater than the upper limit of normal (41.7% vs 27.5%) enrolled than the 2 mg/kg group; all negative prognostic factors.

Table 1:

Demographics and Baseline Characteristics

Parameter	Ontuxizumab 2.0 mg/kg (N=40)	Ontuxizumab 4.0 mg/kg (N=36)	Total (N=76)
Mean age (years)	65.1	61.2	63.3
Range	27–91	35–83	27–91
Gender, n (%)
Male	22 (55.0)	27 (75.0)	49 (64.5)
Race, n (%)
White	39 (97.5)	34 (94.4)	73 (96.1)
Other	1 (2.5)	2 (5.6)	3 (3.9)
M stage at study entry, n (%)
M1a/M1b	16 (40.0)	8 (22.2)	24 (31.6)
M1c	24 (60.0)	28 (77.8)	52 (68.4)
Lactate dehydrogenase, n (%)
Normal	17 (42.5)	15 (41.7)	32 (42.1)
Above upper level of normal	11 (27.5)	15 (41.7)	26 (34.2)
2x above upper level of normal	12 (30.0)	6 (16.7)	18 (23.7)
ECOG performance status, n (%)
ECOG 0	18 (45.0)	25 (69.4)	43 (56.6)
ECOG 1	22 (55.0)	11 (30.6)	33 (43.4)
BRAF positive status, n/n (%)a	5/21 (24)	1/16 (6)	6/37 (16)
NRAS positive status, n/n (%)a	2/8 (25)	1/6 (17)	3/14 (21)

^aMutation status was not reported or available for all patients. BRAF=a member of the RAF kinase family of serine/threonine-specific protein kinases, NRAS=neuroblastoma v-ras oncogene homolog.

All patients had prior melanoma-related surgery and at least 1 prior treatment for melanoma (Table 2), the most common being ipilimumab (69.7% overall). Prior melanoma treatments ranged from 1 to 12 with a median of 2.

Table 2:

Frequency of Key Prior Melanoma Treatments

	Number (%)
	Ontuxizumab		All Patients (N=76)
	2 mg/kg (N=40)	4 mg/kg (N=36)
Treatment
Any prior melanoma treatment	40 (100)	36 (100)	76 (100.0)
Ipilimumab	25 (62.5)	28 (77.8)	53 (69.7)
Vemurafinib	3 (7.5)	3 (8.3)	6 (7.9)
Dabrafenib	3 (7.5)	0	3 (3.9)
Trametinib	2 (5.0)	2 (5.6)	4 (5.3)

Efficacy:

The primary efficacy endpoint, 24-week PFS value, was 11.4% (CI: 5.3%–19.9%) for all patients (13.5% for the 2 mg/kg group and 8.9% in the 4 mg/kg group), below the protocol-stated value of interest (35%) based on ipilimumab data [9]. However, 69.7% of patients in this study failed ipilimumab therapy making the relevance of this cut-off equivocal. The median PFS was 8.3 weeks in the 2 mg/kg group and 8.4 weeks in the 4 mg/kg group (Fig. 2). There was no significant difference between the 2 doses (p=0.8460).

Fig. 2.

Kaplan-Meier curves for progression-free survival for all patients by dose cohort

Progression-free survival at 16 and 52-weeks was also determined for the primary efficacy population. The overall PFS value at 16 weeks was 27.0% (95% CI: 17.3%–37.6%) and at 52 weeks was 4.3% (95% CI: 1.1%–10.8%), with no difference between dose cohorts.

A total of 60 of the 75 patients (80%) in the primary efficacy population died (71.8% in the 2 mg/kg group and 88.9% in the 4 mg/kg group). The median overall survival was 31.0 weeks (95% CI: 28.3–44.0 weeks); 40.9 weeks (95% CI: 29.0–53.3 weeks) in the 2 mg/kg group and 29.3 weeks (95% CI: 22.9–35.4 weeks) in the 4 mg/kg group (Fig. 3). No statistically significant difference in overall survival between the 2 dose groups was observed (p=0.367).

Fig. 3.

Kaplan-Meier curves for overall survival for all patients by dose cohort

Among patients who had not received ipilimumab previously (n=22), 14 (63.6%) patients died with a median overall survival of 45.1 weeks (95% CI: 24 weeks, upper CI not calculable); compared with those who had received ipilimumab (n=53), 46 (86.8%) died with a median overall survival of 29.5 weeks (95% CI: 25.9–40.9 weeks, p=0.0768). Estimates of PFS by prior ipilimumab treatment did not show this difference; the median PFS was 8.1 weeks (95% CI: 7.4, 8.7) among patients who had not received ipilimumab previously and 10.4 weeks (95% CI: 8.1, 15.4, p=0.2151) among patients who had received ipilimumab.

Because of the imbalance in patients with high LDH values in the 4 mg/kg treatment group, clinical responses by baseline LDH values were evaluated. Patients with baseline LDH values less than the median had both a significantly longer median overall survival and significantly longer median PFS. Patients with baseline LDH values less than the LDH median value had a median overall survival of 10.9 months (95% CI: 7.3–17.4 months) compared with a median overall survival of 5.9 months (95% CI:3.8–7.1 months, p=0.0002) for patients with baseline LDH values ≥ the LDH median; with a Hazards Ratio (HR) of 0.38 (95% CI: 0.22–0.64). Similarly, patients with baseline LDH values less than the LDH median had a median PFS of 2.1 months (95% CI: 1.9–3.7) compared with a median PFS of 1.9 months (95% CI: 1.7–3.1 months, p=0.0212) for patients with baseline LDH values ≥ the LDH median; with HR of 0.57 (95% CI: 0.35–0.93).

No patients achieved a complete response and 2 patients achieved a partial response during the study. One patient in the 2 mg/kg group achieved stable disease at Cycle 2, Day 22 (Study Day 54), then improved to a partial response at Cycle 4 Day 22 (Study Day 113). The partial response was sustained through Day 701, and he has continued to receive ontuxizumab treatment for over 57 months with no unacceptable toxicity. However, this patient was excluded from the efficacy analysis because he had an active malignancy (Merkel cell carcinoma) within the previous 5 years prior to randomization into this study. This patient is an 83-year-old male initially diagnosed with melanoma approximately 10 years prior to entry into the current study. Metastatic cancer was diagnosed approximately 3 1/2 years prior to study start. The patient received prior treatment with ipilimumab for 10.4 weeks, with treatment completed 70 days prior to randomization in the current study. His stage at entry into this study was M1a. Paired tumor tissue endosialin levels were not available for this patient.

A second patient in the 4 mg/kg dose group had a partial response at Cycle 8 Day 22 (Study Day 224) which was sustained through Study Day 448. The patient has continued to receive treatment for over 54 months without unacceptable toxicity. This patient is a 66-year-old male initially diagnosed with Stage III melanoma approximately 4 1/2 years prior to entry into the current study. Metastatic cancer was diagnosed 4 1/4 years prior to study start. The patient was previously treated with high dose interferon and tasisulam and progressed more than 1 year prior to randomization in the current study. His stage at entry into this study was M1c. Paired tumor tissue endosialin levels were not available for this patient.

These 2 patients remain alive and have received treatment for more than 64 months as of August 2017. Overall response rate was 3.1% among the 32 patients in the 4 mg/kg group with a baseline and at least 1 on-study radiologic evaluation (overall response rate population). The time to response was 32.1 weeks for the 1 patient with a partial response. The duration of response has been maintained for at least 56 months.

Twenty-seven out of 66 response evaluable patients (40.9%, 14 in the 2 mg/kg group and 13 in the 4 mg/kg group) achieved a best overall response of stable disease. The overall median duration of stable disease for these 27 patients was 16.9 weeks (95% CI: 15.3, 19.3). A clinical benefit (complete response + partial response + stable disease) was achieved by 28 patients (42.4%). Neither dose, age, race, ECOG performance status at baseline, or M stage was a significant factor in the best overall response using a multivariate logistic regression analysis.

Pharmacokinetics:

The mean terminal phase half-life was similar for both dose groups (11.4 days for patients in the 2 mg/kg group and 11.8 days for patients in the 4 mg/kg group).

Biomarker Studies

The relationship between clinical parameters (PFS, overall survival, and best overall response) and baseline tumor endosialin and PDGFR-β, as well as baseline serum endosialin levels were evaluated. Previous experimental data have shown a biological link between these pathways [23]. No correlations between mean baseline biomarker levels and clinical outcome were evident.

Further, immunohistochemical analysis for endosialin expression on paired tissue samples obtained from 24 patients at baseline and during on-study Cycle 1 did not demonstrate significant changes in expression levels. BRAF and NRAS mutation status was only available for approximately 49% of patients and therefore precluded detailed analyses.

Safety

All treated patients experienced at least 1 adverse event irrespective of attribution during the study. The most common adverse events overall were headache (55.3%), fatigue (48.7%), chills (42.1%), and nausea (36.8%). The majority of adverse events were Grade 1 or 2 in severity. Grade 3 or higher adverse events were similar between dose groups; 47.5% of patients in the 2 mg/kg group and 55.6% in the 4 mg/kg group. The most frequently reported Grade 3 or higher adverse event was anemia (5.0% in the 2 mg/kg group and 19.4% in the 4 mg/kg group). The frequencies of all other Grade 3 or higher adverse events were generally comparable in the 2 dose groups. Grade 3 or higher adverse events of fatigue occurred in 4 (5.3%) patients; abdominal pain, cellulitis, and pulmonary embolism each occurred in 3 (3.9%) patients; and nausea, headache, infusion-related reaction, hypokalemia, confusional state, hypertension, and lymphedema each occurred in 2 (2.6%) patients. At least 1 treatment-related, per investigator, adverse event occurred in 69 (90.8%) patients. Treatment-related adverse events occurring in greater than 15% of patients are shown in Table 3 Table. In general, the types and frequencies of treatment-related adverse events were comparable in the 2 dose groups. Chills and pyrexia occurred more frequently in the 4 mg/kg dose group than in the 2 mg/kg dose group.

Table 3:

Treatment-Related Adverse Events in 15% or More Patients

	N (%)
	Ontuxizumab
	2.0 mg/kg (N=40)		4.0 mg/kg (N=36)		Total (N=76)
Preferred term	All grades	Grade 3,4	All grades	Grade 3,4	All grades	Grade 3,4
Headache	21 (52.5)	0	21 (58.3)	2 (5.6)	42 (55.3)	2 (2.6)
Fatigue	15 (37.5)	1 (2.5)	13 (36.1)	1 (2.8)	28 (36.8)	2 (2.6)
Chills	11 (27.5)	0	16 (44.4)	0	27 (35.5)	0
Pyrexia	8 (20.0)	0	13 (36.1)	0	21 (27.6)	0
Nausea	9 (22.5)	0	11 (30.6)	1 (2.8)	20 (26.3)	1 (1.3)
Decreased appetite	7 (17.5)	0	6 (16.7)	0	13 (17.1)	0

Seven patients had fatal adverse events during the study, (6 were due to progressive disease and 1 in the 4 mg/kg group was due to pneumonia). None of the fatal events were considered by the investigator to be related to study drug treatment.

Treatment-related serious adverse events occurred in 9 patients (2 [5.0%] patients in the 2 mg/kg group and 7 [19.4%] in the 4 mg/kg group). Three treatment-related serious adverse events occurred in 2 patients each: infusion-related reaction, pyrexia, and pulmonary embolism. All other treatment-related serious adverse events occurred in 1 patient each and included anemia, cellulitis, wound infection, hypersensitivity, chills, toxic encephalopathy, and hyponatremia. All treatment-related serious adverse events resolved.

Four patients had adverse events that resulted in discontinuation of study drug; 2 patients in the 2 mg/kg group had pulmonary emboli (1 considered possibly related and 1 considered not related to study drug, 1 patient in the 4 mg/kg group had a small intestinal obstruction (considered not related to study drug), and 1 patient in the 4 mg/kg group experienced a confusional state (considered probably related to study drug). All 4 events resolved. The Grade 3 dose-limiting toxicity of vomiting observed in the Phase 1 study [26] was not observed in this study.

Adverse events of interest included chills, pyrexia, infusion-related reactions, pruritus, hypersensitivity, dyspnea, and wheezing. The most frequently occurring adverse event of interest was chills in 38.2% of patients overall.

Five patients (6.8%) had a positive ADA level during the study. Drug hypersensitivity adverse events were defined as any adverse event of interest associated with a treatment-induced or treatment-boosted ADA response. Three patients had drug hypersensitivity events, all in the 4 mg/kg dose group. Two patients each had chills and pyrexia, and 1 patient had an infusion-related reaction and pruritus. All drug hypersensitivity events, except the pruritus, resolved; and none of the events resulted in dose interruption or discontinuation.

DISCUSSION:

In preclinical studies, endosialin has been found to play a key role in tumor growth and neovascularization in numerous cancer types. This was the first study to assess the efficacy of an endosialin-targeted antibody in patients with metastatic melanoma.

In this study, a PFS value of interest among patients with refractory metastatic melanoma was originally identified as ≥35% at 24 weeks based on results from an ipilimumab study in patients who had not previously been treated with ipilimumab [9]. The 24-week PFS value in all ontuxizumab monotherapy patients in the current study was 11.4% and therefore did not achieve the pre-specified 35% PFS goal. A high target PFS value was established to determine if a strong efficacy signal was apparent in this early clinical study; this target PFS value may have been set too high. Most (69.7%) patients enrolled in this study failed prior ipilimumab therapy, therefore making it difficult to compare with response outcomes reported in ipilimumab studies. As shown in a subgroup evaluation in this study, patients who had not received prior ipilimumab treatment did better than those who had received prior ipilimumab treatment. Patients who had not received prior ipiliumab treatment had a longer (although not significant) overall survival compared to those who had received prior ipilimumab treatment. Despite the modest PFS value and the complexity in interpreting these results, disease stabilization was observed, with 27 of 66 response evaluable (40.9%) patients treated with ontuxizumab achieving a best overall response rate of stable disease.

At the time of this study report, 2 patients remain alive and continue to receive progression-free, long-term treatment with no unacceptable toxicity. The patient receiving 2 mg/kg ontuxizumab has received treatment for over 67 months and the patient receiving 4 mg/kg ontuxizumab has received treatment for over 64 months. Initial evaluation of these 2 patients did not identify any salient features to explain the lengthy response.

There were a number of limitations to this study. Overall, there was a small sample size and patients enrolled in this study consisted of a heterogeneous population of previously treated melanoma patients, with the number of prior melanoma treatments ranging from 1–12. In addition, the relative results of the 24-week PFS in this heterogeneous population are difficult to interpret without a reference control group. A better comparison is a 6-month PFS of 16% (95% CI: 10–22%) observed in a clinical study among 179 patients treated with chemotherapy following progressive disease after treatment with ipilimumab or BRAF or MEK inhibitors [31].

The doses selected to test clinically were chosen based on the best available data at the time of study initiation [26, 32]. Two doses of ontuxizumab (2 mg/kg and 4 mg/kg) were evaluated in this study, with the expectation that a dose response would be evident. Although no significant differences were apparent between the 2 dose groups for any efficacy parameter, the higher dose did not trend towards better clinical responses. This may be explained by several imbalances in baseline characteristics between the two dose groups that favored the 2 mg/kg treatment group, especially the higher percentage of patients in the 4 mg/kg group with elevated LDH at baseline, a statistically significant marker for progression in melanoma [33]. Based on preliminary pharmacokinetic data indicating potential accumulation of ontuxizumab at higher doses, no higher than 4 mg/kg was tested in this study. Since the initiation of the current study, two studies have been initiated in patients with colorectal cancer or soft tissue carcinoma using a dose of 8 mg/kg without tolerability concerns.

Baseline endosialin and PDGFR-β were assessed relative to efficacy in order to determine if expression could be predictive of a population of patients who would respond to ontuxizumab therapy. These two biomarkers were evaluated because of experimental data showing a link between these pathways and cellular proliferation [23]. Although evidence has linked the expression of endosialin with tumor growth and progression in preclinical studies, the biomarkers measured in this study showed no clear or strong predictive association with clinical response in ontuxizumab-treated patients. No notable pharmacodynamic effects were associated with baseline tumor endosialin or PDGFR-β markers, or baseline serum endosialin. These results may be explained by insufficient molecular data available for analysis. Another possible explanation for the lack of correlation of biomarkers with response is that the dose of ontuxizumab was not high enough to produce a maximum response, however a preliminary evaluation of ontuxizumab serum concentration versus response did not suggest this explanation.

No significant change in tumor endosialin expression was noted in paired tissue samples from before and after one cycle of treatment with ontuxizumab. Although this was a small sample size with limited serial collections, endosialin levels appear not to have been significantly affected. In a recent study using a human endosialin knock-in mouse tumor model, endosialin was highly expressed in the pericytes of tumor blood vessels soon after tumor implantation [34]. Treatment with ontuxizumab showed internalization and degradation of endosialin in pericytes, using immunofluorescent staining of tumor microvasculature. Additional preclinical investigations are ongoing to better understand the mechanism of action of ontuxizumab and the most suitable application for targeting endosialin with ontuxizumab, as well as potentially identifying a molecular profile for patient enrichment. Future clinical trials with ontuxizumab in melanoma will take into account the most recent developments in the field of melanoma research, including potential combinations with other active drugs.

CONCLUSIONS:

Ontuxizumab was generally well tolerated in this patient population. The primary endpoint, 24-week PFS value, was 11.4% among all ontuxizumab-treated patients. The overall response rate was 3.1% at the 4 mg/kg dose, with clinical benefit achieved in 42.4% of all response evaluable patients. Efficacy of single-agent ontuxizumab at the 2 doses used in this study in melanoma was low.