A phase 1, dose-escalation study of PF-06664178, an anti-Trop-2/Aur0101 antibody-drug conjugate in patients with advanced or metastatic solid tumors

Summary

Purpose and Methods

Trop-2 is a glycoprotein over-expressed in many solid tumors but at low levels in normal human tissue, providing a potential therapeutic target. We conducted a phase 1 dose-finding study of PF-06664178, an antibody-drug conjugate that targets Trop-2 for the selective delivery of the cytotoxic payload Aur0101. The primary objective was to determine the maximum tolerated dose and recommended phase 2 dose. Secondary objectives included further characterization of the safety profile, pharmacokinetics and antitumor activity. Eligible patients were enrolled and received multiple escalating doses of PF-06664178 in an open-label and unblinded manner based on a modified continual reassessment method.

Results

Thirty-one patients with advanced or metastatic solid tumors were treated with escalating doses of PF-06664178 given intravenously every 21 days. Doses explored ranged from 0.15 mg/kg to 4.8 mg/kg. Seven patients experienced at least one dose limiting toxicity (DLT), either neutropenia or rash. Doses of 3.60 mg/kg, 4.2 mg/kg and 4.8 mg/kg were considered intolerable due to DLTs in skin rash, mucosa and neutropenia. Best overall response was stable disease in 11 patients (37.9%). None of the patients had a partial or complete response. Systemic exposure of PF-06664178 increased in a dose-related manner. Serum concentrations of free Aur0101 were substantially lower than those of PF-06664178 and total antibody. No correlation of Trop-2 expression and objective response was observed, although Trop-2 overexpression was not required for study entry. The intermediate dose of 2.4 mg/kg appeared to be the highest tolerated dose, but this was not fully explored as the study was terminated early due to excess toxicity.

Conclusion

PF-06664178 showed toxicity at high dose levels with modest antitumor activity. Neutropenia, skin rash and mucosal inflammation were dose limiting toxicities. Findings from this study may potentially aid in future antibody drug conjugate design and trials.

Introduction

Antibody-drug conjugates (ADCs) are immunoconjugates developed to improve the therapeutic index of potent cytotoxic agents by combining them with highly selective and stable monoclonal antibodies (mAbs) [1]. This strategy minimizes systemic exposure and off-target toxicity to normal tissues by releasing highly cytotoxic small-molecules when the ADC is internalized and process in the target cancer cell. Engineering more efficient and less toxic ADCs remains an intense area of research and clinical drug development. Among the challenges involved in ADC development are: identifying an ideal tumor associated antigen (TAA) that permits selective targeting of the ADC; optimizing linker stability in the blood to limit off-target toxicity while allowing ready release inside the target cancer cell; generating homogenous ADCs with more predictable pharmacokinetics; and finding the ideal drug-antibody ratio (DAR) that defines the number of cytotoxic molecules per mAbs [2].

Trophoblast cell surface antigen 2 (Trop-2), is a cell surface glycoprotein originally identified in human placental trophoblasts that have the ability to invade uterine decidua during placental implantation [3]. Trop-2 overexpression has been found in a variety of human malignancies and linked to increased tumor growth, aggressiveness, metastasis, and a poor prognosis [4–9]. Although the exact biological role of Trop-2 in carcinogenesis remains unclear, upregulation has been shown to drive cancer-growth-stimulatory signaling through NF-kB, cyclin-D1 and the MAPK/ERK pathway [10]. Moreover, Trop-2 has lower expression in normal tissues compared to tumor cells, making it an attractive TAA for therapeutic targeting [11].

PF-06664178 (also known as RN927C) is an ADC composed of a humanized anti-Trop-2 IgG1 antibody (PF-06478924, RN926) conjugated with an AcLys-VCAur0101 (PF-06380101) linker-payload at the C-terminus of the antibody heavy chain via an enzymatic process that produces highly homogenous conjugates [12, 13]. Upon binding to the extra-cellular portion of Trop-2 on the cell surface, PF-06664178 is trafficked to the lysosomes where the cleavable linker is processed by proteases to release its auristatin-based Aur0101 payload. Aur0101 is a potent inhibitor of tubulin polymerization with inhibitory concentration IC₅₀ in the picomolar range [12, 14]. PF-06664178 contains two molecules of Aur0101 per ADC producing a low antibody-drug ratio (DAR) of 2 [12].

Preclinical pharmacokinetic (PK) studies of a single 1.5 mg/kg dose of PF-06664178 in Trop-2 positive BxPC3 patient derived tumor xenografts (PDX) with blood and tumor samples taken at 0 to 336 h demonstrate an area under the curve (AUC) AUC_0–336 percentage of ADC from the total antibody population in serum of 87% [12]. This shows that the AcLys-VC has good linker stability in vivo and is a feature of the ADC that was projected to minimize toxicity by avoiding release of the potent payload in off-target tissues [13].

Preclinical in vitro studies in multiple cancer cell lines with a wide range of Trop-2 expression showed significant antitumor activity of PF-06664178, with most cell lines readily killed at an IC₅₀ below 1 nmol/L. However, the ADC was not active against Trop-2 negative SW620 cell lines at the highest dose tested (6.0 mg/kg), suggesting that Trop-2 expression is required for the ADC’s antitumor activity. The in vivo efficacy was also demonstrated in PDX models of pancreas, ovarian, breast and lung cancer, showing sustained growth inhibition and/or regression at doses between 0.75 to 1.5 mg/kg. Similar to cell line studies, PF-06664178 did not demonstrate activity against Trop-2 negative tumors. In addition, although nonclinical studies showed that the parent antibody (PF-06478924, RN926) was able to mediate antibody-dependent cytotoxicity, it did not elicit any complement-dependent cytotoxicity and did not show any in vivo efficacy despite having the same effector function [12].

Exploratory toxicology studies in cynomolgus monkeys (up to 6 mg/kg doses), demonstrated toxicity signals in normal tissues that express the Trop-2 antigen. Reversible findings included increased necrosis in multiple types of epithelial tissue including skin, upper gastrointestinal (GI) mucosa and vagina. These toxicities were likely related to on-target activity of the Aur0101 payload in Trop-2 expressing cells, which contrasts to the hematopoetic and lymphoid toxicity seen in rat studies [12].

Based on the robust preclinical findings demonstrating specific antitumor activity and adequately characterized nonclinical toxicity profile, we conducted a phase 1 dose escalation study of of PF-06664178 in adult patients with advanced solid tumors who were unresponsive to current therapies or for whom no standard therapy was available. We report our findings on the safety and preliminary antitumor activity seen in this first-in-human clinical trial of PF-06664178 in advanced or metastatic solid tumors.

Patients and methods

Study design

This was a multi-center, open-label, non-randomized, phase 1 study of PF-06664178 in sequential cohorts of patients with advanced or metastatic solid tumors for whom no standard therapy was available. The primary objective was to determine the maximum tolerated dose (MTD) and recommended phase 2 dose (RP2D). Secondary objectives were to characterize PK, immunogenicity, overall safety profile and preliminary evidence of antitumor activity.

The study was designed to include three parts comprised of dose escalation and expansion cohorts for specific tumor types. Part 1 was to enroll a maximum of 40 patients at three sites. Once the MTD was established, the study would have then proceeded to part 2 that would have tested 10–20 patients with non-small cell lung cancer (NSCLC) at the MTD. Part 3 would have been initiated once preliminary signs of efficacy with a target overall response rate (ORR) of approximately 30% was observed. However, parts 2 and 3 were not conducted due to excess toxicity seen in the conduct of part 1.

In Part 1, eligible patients were enrolled and received multiple escalating doses of PF-06664178 in an open-label and unblinded manner based on a modified continual reassessment method (mCRM) [15, 16]. PF-06664178 was administered to cohorts of two to four patients, escalating to the next dose level after all patients in the current level completed the assessments required in cycle one. The criteria for dose escalation followed a mCRM for a target dose limiting toxicity (DLT) rate of 25%. The mCRM model determined the next dose level based on the current DLT rate of all patients in the study with the dose increase at a rate of ≤30%, ≤60% and ≤100% dose increment for zero, one or two doses skipped respectively. For any given dose, the escalation stopped if the rate of toxicity appeared to be higher than approximately 30%, and the lower doses were evaluated and/or re-evaluated. To prevent overly aggressive dose-escalation, the maximum allowed dose increase was limited to three increments at a time, (i.e., no more than 100% dose increase at a time). In addition, if two clinically significant grade 2 toxicities of the same type were observed in one cohort, dose escalation was to be limited to no greater than 60% for the next cohort. If there was one additional case of the same grade 2 toxicity or two other cases of clinically significant grade 2 toxicities of the same type, dose escalation was again limited to no greater than 60% for the next cohort; otherwise dose escalation following the mCRM algorithm was resumed.

Approval was obtained from the ethics committees at the participating institutions and regulatory authorities. Patients gave written informed consent. The study followed the Declaration of Helsinki and the International Conference on Harmonization Good Clinical Practice guidelines. The final protocol, any amendments and informed consent documentation were reviewed and approved by the institutional review boards at each of the investigational centers participating in the study. The study was supported by Pfizer Inc. and registered at ClinicalTrials.gov (ID: NCT02122146).

Patient selection

For inclusion into the study, eligible patients were required to have advanced solid tumors resistant to standard therapy, or for which no other therapy was available, and at least one measurable lesion defined by Response Evaluation Criteria in Solid Tumors (RECIST version 1.1). Patients also needed to have adequate bone marrow, renal and liver function and an estimated life expectancy of >12 weeks. Patients were excluded if they had a history of clinically significant dermatologic or mucosal disease (including mucositis induced by previous treatments), significant ocular disease (particularly of the anterior chamber), symptomatic brain metastases requiring maintenance steroids, and/or had current use or anticipated need for treatment with moderate/strong cytochrome P450 3A inhibitors or inducers.

Treatment and dose limiting toxicity

PF-06664178 was administered once every 21 days as an intravenous infusion over approximately 60 min. Dose escalation of PF-06664178 started with a dose of 0.15 mg/kg for the first cohort. The proposed doses explored were obtained from a fine grid of doses starting from 0.15 mg/kg to 6.14 mg/kg. Patients were registered and successively assigned to the next available treatment slot at a dose level determined after the previous cohort’s safety evaluation and ongoing observations of patients enrolled earlier. Day 1 safety laboratory results were reviewed by the primary investigator prior to dosing at the beginning of each cycle. Administration of the study drug continued until disease progression, patient withdrawal or unacceptable toxicity occurred.

Any of the following adverse events (AEs) occurring in the first cycle (length of 21 days) were classified as DLTs (unless due to a clear alternative explanation): hematologic: grade 4 neutropenia lasting >7 days, febrile neutropenia or infection while having grade 3 neutropenia, grade 4 thrombocytopenia or grade 3 thrombocytopenia with significant bleeding, grade 4 anemia; non-hematologic: grade ≥ 3 toxicities that were considered clinically significant, except those that had not been maximally treated or could have been easily treated and grade ≥ 3 cytokine release syndrome. Trop-2 tumor expression was not required for entry into this part of the study. In addition, clinically significant or persistent toxicities not included in the above could also be considered a DLT.

Assessments

Safety

Patients were assessed for safety at baseline, days 1, 2, 4, 8, and 15 for cycle 1, on days 1, 8 and 15 for cycle 2, on day 1 for subsequent cycles, and at the end of treatment. AEs were graded by severity using the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI-CTCAE version 4.03). Patients were followed for AEs 28 days after the last treatment administration or until all drug related toxicities had resolved.

Antitumor activity

Tumor assessments of known disease sites were performed by either computed tomography (CT) or magnetic resonance imaging (MRI) every 6 weeks from the start of study treatment until disease progression by RECIST 1.1 or death, or until permanent discontinuation of study treatment.

Pharmacokinetics and immunogenicity

Blood samples for PK assessments were collected on days 1, 4, 8, 15 for cycles 1 and 4, on day 1 for all other cycles, and at the end of treatment. Drug concentrations in serum of the PF-06664178 ADC and total antibody were measured using validated electrochemiluminescent (ECL) assays. Concentrations of unconjugated Aur0101 payload were measured using a validated ultra-high performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) assay. The limit of quantitation for ADC, total antibody, and unconjugated payload were 10.0, 100, and 0.015 ng/mL, respectively. PK parameters were not calculated for this terminated study. Samples for anti-drug antibody (ADA) and neutralizing antibody (Nab) analyses were collected on day 1 prior to administration of the study drug, day 15, and on subsequent dosing days prior to each subsequent administration of the study drug. ADA samples were analyzed using a validated ECL method with a screen, confirm, titer approach, and Nab was assessed for ADA positive samples using a validated competitive enzyme-linked immunosorbent assay.

Tumor biopsy and immunohistochemistry assay for Trop-2 expression

A Trop-2 specific immunohistochemistry (IHC) assay designed for use in routinely-collected formalin fixed paraffin embedded (FFPE) and analytically validated for sensitivity, specificity, precision, and accuracy, was used to assess Trop-2 expression in patient primary tumor samples. This assay was first optimized by identifying controls with Trop-2 expression confirmed by orthogonal methods. To this end, cell lines were characterized for Trop-2 expression by flow cytometry, western blot, quantitative real-time polymerase chain reaction (qRT-PCR), and affymetrix chips, and 4 control cell lines that represented negative, low, moderate, or high Trop-2 expression were identified. Using xenograft tissues prepared from the four control cell lines, IHC assay conditions were optimized on the Dako autostainer platform using Dako IHC reagents [12]. A scoring algorithm based on the prevalence of Trop-2 expression was developed from evaluation on human tumor tissue microarrays. Trop-2 IHC expression was scored on an increasing scale of 0 to 3+. Patients enrolled into part 1 did not require Trop-2 expression for study enrollment and IHC testing for Trop-2 expression was optional. Patients enrolled were not excluded from the study if they did not have a tumor block or slides available, although if samples were available patients must have consented to provide access to those samples for Trop-2 expression analysis. Patients in parts 2 and 3 would have been pre-selected according to moderate/high levels (as defined by ≥50% cells with 2+ or 3+ expression intensity) of target tumor expression levels, but such was not conducted as the study was terminated in part 1.

Sample size and statistical analysis

The dose escalation study employed a mCRM algorithm that utilizes Bayesian methodology to characterize the dose-toxicity relationship after each cohort’s DLT response became available. This dynamic process would continue until a maximum of 40 patients were enrolled, with each cohort enrolling 2–4 patients at each dose level. A minimum of 6 patients on MTD dose in Part 1 was required to establish such dose as RP2D. Descriptive statistics were used to summarize continuous, categorical, and time-to-event variables which included DLTs, AEs, PK analysis and overall response.

Results

Patients

A total of 31 patients were enrolled and received treatment with PF-06664178. Patient characteristics are outlined in Table 1. The four most common tumor types in patients that underwent treatment were ovarian cancer (6 patients), NSCLC (6 patients), breast cancer (4 patients) and squamous cell cancer of the head & neck (SCCHN, 3 patients). All other tumor types were diagnosed in two patients or less. Most patients were heavily pre-treated, with about 26 (84%) receiving more than three lines of prior systemic therapy. All patients had an Eastern Cooperative Group Performance Score (ECOG PS) of 0 or 1.

Table 1.

Baseline patient demographics and characteristics

Mean Age	57 years
Male : Female	13:18
Race
Caucasian	27 (87%)
Asian	1 (3%)
Other	3 (9%)
ECOG PS
0	12 (39%)
1	19 (61%)
Primary Tumor
Ovary	6 (19%)
NSCLC	6 (19%)
Breast	4 (13%)
SCHNN	3 (10%)
Colon	2 (6%)
Neuroendocrine	2 (6%)
Prostate	2 (6%)
Sarcoma	2 (6%)
Stomach	1 (3%)
Cholangiocarcinoma	1 (3%)
Esophagus	1 (3%)
Cervix	1 (3%)
Number of Prior Systemic Therapies
1	2 (6%)
2	0
3	3 (9%)
> 3	26 (84%)
Prior Surgery	29 (94%)
Prior Radiation	20 (65%)

Dose-limiting toxicities

Seven of the 24 DLT evaluable patients experienced at least one DLT during the course of the study. A summary of DLTs is presented in Table 2. During the initial dose escalation (from 0.15, 0.3, 0.6, 1.2, 2.4, 3.6 to 4.8 mg/kg), no DLTs were observed until the highest dose of 4.8 mg/kg dose was reached. The 4.8 mg/kg dose was toxic and deemed intolerable as four out of eight patients experienced a DLT attributable to the study drug. In this dose cohort, three patients developed a grade 3 rash during the first 2 cycles resulting in temporary discontinuation in one patient and permanent discontinuation in another. All cases of rash in this cohort resolved. In addition, grade 4 toxic epidermal necrolysis (TEN) occurred in one patient and grade 3 bullous dermatitis in another. The patient who presented with TEN presented with symptoms as early as the fourth day of the first cycle of treatment. This patient presented with a swelling sensation in the throat, dysuria and a rash extending from the groin to the gluteal cleft. The patient later developed grade 4 febrile neutropenia associated with mucosal inflammation and a skin biopsy of the rash was consistent with TEN. Treatment with PF-06664178 was permanently discontinued and TEN resolved with sequelae on day 15 while neutropenia resolved to less than grade 2 on day 22.

Table 2.

Dose limiting toxicities by dose level

Dose (mg/kg)	No. of DLT evaluable patients	No. (%) of patients with DLTs	DLT
0.15	2	0
0.30	2	0
0.60	4	0
1.20	4	0
2.40	4	0
3.60	6	2 (33.3%)	Gr 4 neutropenia, Gr 3 mucosal inflammationa
4.20	1	1 (100%)	Gr 3 maculopapular rasha
4.80	8	4 (50%)	Gr 4 febrile neutropenia, Gr 4 TEN, Gr 4 dehydration, Gr 3 rash (1/4 was bullous dermatitis)

Abbreviations: DLT = dose limiting toxicity; TEN = Toxic Epidermal Necrolysis

^aThese DLTs occurred during exploration of intermediate doses after the 4.8 mg/kg and was deemed intolerable

Owing to the intolerability of the 4.8 mg/kg dose level, subsequent lower dose levels of 4.2 mg/kg and 3.6 mg/kg were evaluated. The only patient evaluated at 4.2 mg/kg experienced a DLT of grade 3 rash and withdrew consent with rash ongoing at the time of study termination. Of the six patients dosed at 3.6 mg/kg, two experienced DLTs comprising of grade 4 neutropenia and grade 3 mucositis. Both events necessitated dose-reductions and temporary discontinuation in one patient, which resulted in resolution of both AEs. Four patients were evaluated at the 2.4 mg/kg dose level during dose escalation and none experienced a DLT. However, due to the potential lack of a therapeutic window, the study was terminated before this dose level was further evaluated to confirm the MTD.

Safety

Multiple doses of PF-06664178 from 0.15 mg/kg through 2.4 mg/kg were generally safe and tolerable. However, the 3.6 mg/kg, 4.2 mg/kg and 4.8 mg/kg doses were considered toxic and intolerable. The most commonly reported all-causality treatment emergent adverse events (TEAEs) for all cycles were fatigue (42%), constipation (36%), nausea (32%), chills (29%), infusion related reaction (26%), neutropenia (26%), rash (26%), weight decreased (26%), arthralgia (23%), decreased appetite (23%), diarrhea (20%), dyspnea (20%), mucosal inflammation (20%) and pruritus (20%). The majority of these adverse events were low grade except for neutropenia, rash and mucosal inflammation, as summarized in Table 3. Despite some observations with prior ADC drugs, no ophthalmologic toxicity was observed in the evaluable patients though there were some patients who did not complete the ophthalmologic examination at the end of study. All causality grade 3 or 4 AEs were reported in 20 (64.5%) patients, 14 (45.2%) of which were determined to be treatment related. The most common grade 4 treatment related AE was neutropenia (6) and the most common grade 3 AE was rash. Of note, most of the AEs for rash and neutropenia were either grade 3 or 4 (Table 3).

Table 3.

Treatment emergent adverse events by dose level

	0.15 mg/kg (n = 2)		0.3 mg/kg (n = 2)		0.6 mg/kg (n = 4)		1.2 mg/kg (n = 4)		2.4 mg/kg (n = 4)		3.6 mg/kg (n = 6)		4.2 mg/kg (n = 1)		4.8 mg/kg (n = 8)
	Gr 1/2	Gr 3/4	Gr 1/2	Gr 3/4	Gr 1/2	Gr 3/4	Gr 1/2	Gr 3/4	Gr 1/2	Gr 3/4	Gr 1/2	Gr 3/4	Gr 1/2	Gr 3/4	Gr 1/2	Gr 3/4
All Causality
Fatigue	0	0	0	0	3	0	1	0	1	0	3	1	0	0	3	1
Constipation	0	0	0	0	2	0	1	0	2	0	1	0	0	0	5	0
Nausea	0	0	0	0	2	0	0	0	2	0	3	0	0	0	3	0
Chills	0	0	1	0	0	0	0	0	1	0	3	0	0	0	4	0
Infusion Reactions	0	0	0	0	0	0	1	0	1	0	4	0	0	0	1	1
Neutropenia	0	0	0	0	0	0	0	0	0	1	0	2	0	0	1	4
Rash	0	0	0	0	0	0	0	0	2	0	1	1	0	0	1	3
Decreased Appetite	0	0	1	0	3	0	0	0	1	0	1	0	0	0	1	0
Arthralgia	0	0	0	0	2	0	1	0	0	0	2	0	0	0	2	0
Mucosal Inflammation	0	0	0	0	0	0	0	0	0	0	1	1	1	0	1	2
Diarrhea	0	0	0	0	2	0	0	0	2	0	2	0	0	0	0	0
Pruritus	0	0	0	0	0	0	0	0	2	0	1	1	0	0	2	0
Treatment Related
Fatigue	0	0	0	0	1	0	1	0	0	0	3	1	0	0	2	0
Constipation	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2	0
Nausea	0	0	0	0	2	0	0	0	1	0	3	0	0	0	3	0
Chills	0	0	1	0	0	0	0	0	1	0	1	0	0	0	2	0
Infusion Reactions	0	0	0	0	0	0	1	0	1	0	4	0	1	0	1	1
Neutropenia	0	0	0	0	0	0	0	0	0	1	0	2	0	0	1	4
Rash	0	0	0	0	0	0	0	0	2	0	1	1	0	0	1	3
Decreased Appetite	0	0	1	0	2	0	0	0	0	0	1	0	0	0	1	0
Arthralgia	0	0	0	0	0	0	0	0	0	0	2	0	0	0	2	0
Mucosal Inflammation	0	0	0	0	0	0	0	0	0	0	1	1	1	0	0	2
Diarrhea	0	0	0	0	0	0	0	0	0	0	2	0	0	0	0	0
Pruritus	0	0	0	0	0	0	0	0	1	0	1	1	0	0	1	0

Seven patients had dose reductions and another 14 patients had temporary discontinuation due to adverse events at the higher doses. Three patients discontinued permanently due to treatment related AEs: all of them in the 4.8 mg/kg treatment cohort and all of them experienced at least a grade 3 rash. One patient discontinued after a grade 3 infusion reaction, another due to persistent rash and the other due to TEN. Two patients died during the serious adverse event (SAE) reporting period, which encompasses the period from consent through and including 28 calendar days after the last administration of PF-06664178. Another two reported deaths occurred after the safety reporting period. None of these were drug related, and all were attributable to progression of disease and non-treatment related dehydration.

With regards to the extent of exposure, duration of treatment ranged from 1 to 318 days with a median of 23 days. The median number of cycles administered was between 1 and 3.5 for all groups except for the 2.4 mg/kg where the median was 6.5 cycles, suggesting that this dose level was well tolerated.

Efficacy

Of the 31 patients who were treated, two patients did not have post-dose tumor evaluations (one due to rapid clinical deterioration and the other withdrew consent after a grade 3 rash). In the 29 response-evaluable patients, the best overall response observed was limited to stable disease (SD) in 11 patients (37.9%) with no patients achieving a partial response (PR) or complete response (CR) based on RECIST 1.1 criteria. Furthermore, three patients had short lived regression that did not meet the durability criteria for best overall response of SD. The waterfall plot for best change in tumor size is shown in Fig. 1 and the spider plot for change in tumor size over time is shown in Fig. 2.

Fig. 1.

Waterfall plot of best change in tumor size

Fig. 2.

Spider plot of change in tumor size

Trop-2 expression results

Optional Trop-2 expression was performed on primary tumor samples from 19 patients.

Thirteen (68.4%) patients were identified as having medium to high expression (≥50% of tissue cells have Trop-2 expression levels of 2+ or 3+ pathology grading) with the remaining 6 patients having low expression (< 50% of 2+ or 3+). There was no apparent relationship between the strength of Trop-2 expression and depth of antitumor response.

Pharmacokinetics

Pharmacokinetic data were available for dose ranges 0.15 to 4.8 mg/kg. Serum concentration of ADC, total antibody, and unconjugated Aur0101 were determined. The mean serum concentrations for PF-06664178 ADC, total antibody and the unconjugated Aur0101 payload increased with increasing dose. Moreover, a parallel trend between the total ADC, total antibody and unconjugated Aur0101 was observed when mean plasma concentrations were plotted against time, demonstrating consistent payload dissociation kinetics as measured in serum across dose groups (Fig. 3). Serum concentrations of free Aur0101 were substantially lower than those of PF-06664178 ADC and total antibody. Twenty-six of 30 patients (86.7%) evaluated for anti-drug antibody (ADA) post-baseline had confirmed positive ADA. The effect of ADA on PF-06664178 PK was not evaluated. Twenty-four of the 30 patients (80%) tested positive for neutralizing anti-drug antibody in a competitive ligand binding assay.

Fig. 3.

Mean concentration of PF-06664178 (a), total anti-Trop2 antibody (b), and (c) unconjugated Aur0101 in serum. Error bars represent 1 standard deviation

Discussion

This was the first-in-human phase 1 trial of PF-06664178, an ADC targeting the Trop-2 TAA. Patients received escalating doses of PF-06664178, starting with 0.15 mg/kg up to 4.8 mg/kg. PK exposures generally increased in a dose-related manner (Fig. 3). Doses of 3.6 mg/kg, 4.2 mg/kg and 4.8 mg/kg were considered intolerable with DLTs primarily due to skin/mucosal and neuropenia AEs. The next lower dose of 2.4 mg/kg was well tolerated with no DLTs and the longest exposure time for patients. However, excess toxicities with minimal antitumor activity precluded further testing and resulted in the early termination of the study due to a potential lack of a therapeutic window. As such, the MTD and RP2D were not determined in this study.

The most common treatment-related AEs were fatigue, nausea, infusion reactions, neutropenia, rash and mucosal inflammation. Rash and neutropenia were the most common treatment related grade 3 and 4 AEs respectively. The extent and severity of rash varied from a maculopapular rash covering more than 30% of the body surface area, to overt TEN with involvement of mucosal surfaces of the oral cavity (also complicated by febrile neutropenia). Notably, two of the three cases of permanent discontinuation due to a DLT were due to grade 3 rash.

As such, although some degree of skin and mucosal toxicity was expected based on the known low-levels of Trop-2 expression in normal tissues, the degree of toxicity seen in humans with this study was out of proportion to the relatively mild toxicity seen in preclinical cynomolgus monkeys toxicology studies [12, 17]. The pattern and severity of skin and mucosal tissue toxicity seen with PF-06664178 in this study is notably different from that of another anti-Trop-2 ADC in clinical development, sacituzumab-govotecan (IMMU-132). Sacituzumab-govotecan is a humanized anti-Trop-2 antibody conjugated to SN-38, the active component of irinotecan. Early phase trials of IMMU-132 show rash as uncommon and mild toxicity, with cytopenias and diarrhea predominating, reflecting off-target toxicities of SN-38 [18–20]. As both drugs specifically target the same Trop-2 TAA, a possible explanation for the distinctly more severe skin and mucosal toxicities seen with PF-06664178 is its highly potent payload. Most clinical stage ADCs have highly potent chemotherapeutic payloads leading to a low DAR. PF-06664178 follows this paradigm by incorporating an auristatin-based payload with a picomolar IC₅₀ anda DAR of 2 [12]. IMMU-132, on the other hand, has a high DAR of 7.6, as it carries a moderately toxic SN-38 payload (nanomolar IC₅₀). In addition, significant toxicities in other tissues where Trop-2 is normally expressed at intermediate levels (bronchi, prostate, breast) were not observed, suggesting that skin and mucosal cells may have a particular sensitivity to the Aur0101 payload [3].

The degree of neutropenia seen in this trial was not expected since Trop-2 is not expressed in neutrophils or the bone marrow [17]. Neutropenia was also not observed in preclinical cynomolgus monkey studies, although neutropenia was observed in mice receiving PF-06664178 and is an expected toxicity of auristatin based cytotoxics [12]. This designates neutropenia as an off-target toxicity to the free Aur0101 payload and implicates ADC linker chemistry and stability. ADC linkers are generally classified into either being cleavable or non-cleavable and demonstrate varying degrees of serum stability which impact release of free drug and off-target toxicity. In general, non-cleavable linkers are more stable but release drug less readily and require degradation of the ADC complex for payload release. Cleavable linkers are more variable in their stability but release drug more readily once enzymatic conditions are met to release the payload [2]. PF-06664178 was designed to incorporate a valine-citrulline ceavalbe linker that is stable in circulation and readily cleaved intracellularly by lysosomal proteases. Preclinical non-human studies have demonstrated the linker and the ADC complex to be stable in the blood stream with little free payload [13]. Similarly, PK data obtained during this study generally mirrored the preclinical data, showing that serum concentrations of free Aur0101 were substantially lower than those of PF-06664178 ADC and total antibody, though we were not able to conduct definitive statistical analyses due to the early termination of the study (Fig. 3). It is also possible that human neutrophils are exquisitely sensitive to auristatin-based cytotoxics despite low levels of free payload [21, 22].

We were not able to observe any objective responses following treatment with PF-06664178. None of the patients had a PR or CR and the best observed response was SD in 11 patients. Although an exploratory analysis of preliminary data looking at dose versus percent change of tumor size from baseline suggested a trend of increasing response with increasing dose (Fig. 4), on-target toxicities with higher doses projected a potential lack of a therapeutic window, leading to early termination of the study. The modest antitumor activity at lower doses contrasts to the potent activity seen in preclinical studies which show efficient lysosomal trafficking and internalization led to significant tumor kill across all cell lines and PDX models [12]. Effective lysosomal trafficking and processing is another essential step for ADCs like PF-06664178 that require lysosomal cleavage for antitumor efficacy. However, an accurate in-vivo measure of lysosomal trafficking is not available and how normal cells process and traffic the internalized ADC is also unknown. It is possible that skin and mucosal cells process the ADC and release free payload more readily than tumor cells. In this study, responses were limited to SD with no patients achieving a PR or a CR. There was a trend to reaching greater antitumor activity with the increasing doses but concomitant increase in toxicities prevented achieving a meaningful therapeutic window.

Fig. 4.

Scatter plot and dose response curve based on E-max model

Future development may incorporate modifications of either the payload or the linker. Off-target toxicity may be further reduced by incorporating a non-cleavable linker that is efficiently trafficked to the lysosome for drug release while maintaining optimal serum stability [2]. Recent studies describe a strategy of employing a bispecific antibody that couples Trop-2 to an efficiently trafficked surface antigen such as amyloid precursor like protein 2 (APLP2). Devay and colleagues noted significant antitumor responses even at low Trop-2 expression with using this strategy [23]. This redirection strategy also has the potential to overcome loss of efficient ADC processing due to receptor recycling to the cell surface.

Conclusion

In conclusion, PF-06664178, a humanized anti-Trop2 ADC showed significant DLTs in skin and mucosa in the varying dose ranges tested. Objective tumor responses were limited and the study was terminated before the MTD could be fully determined. Future development of this ADC is unlikely, but the inferences and implications gathered from this trial may aid in future design of more effective and more tolerable ADCs.