A Phase I Dose-Escalation Study Evaluating the Safety Tolerability and Pharmacokinetics of CUDC-427, a Potent, Oral, Monovalent IAP Antagonist, in Patients with Refractory Solid Tumors

Apoptosis is a physiologic cell death process that, at least in part, is responsible for normal tissue homeostasis and results in the removal of nonessential cells, including those with mutations or alterations in cell-cycle control. Malignant cells, unlike their normal counterparts, are under physiologic stress and are highly dependent on aberrations in the apoptosis signaling pathways to maintain viability.

CUDC-427 is a potent, oral, monovalent, and selective antagonist of the inhibitor of apoptosis (IAP) proteins. The IAP proteins confer protection from death-inducing stimuli by binding to, and inhibiting activated caspases that are critical for the execution of apoptosis (1–3). In contrast to other apoptotic pathway–targeting therapies, CUDC-427 promotes activation of the postmitochondrial downstream apoptotic pathways, creating a potential therapeutic intervention for tumors that exhibit resistance to chemotherapy and radiotherapy from a host of upstream aberrant apoptotic pathway alterations.

CUDC-427 demonstrated single-agent antitumor activity in mouse xenograft models bearing subcutaneous MDA-MB-231 tumors. In addition, CUDC-427 demonstrated additive activity in combination with chemotherapeutic agents and TRAIL receptor–targeting antibodies in xenograft tumor models. Preclinical safety and pharmacokinetic testing established an expected therapeutic range of doses as well as elucidated possible toxicities including inflammatory cytokine and chemokine proteins that could cause an inflammatory reaction and reversible mild to moderate inflammation in the lungs and liver.

The impetus for the clinical development of CUDC-427 included the critical role of apoptosis pathway aberrations in neoplasia and the aforementioned safety and antitumor activity of CUDC-427 in preclinical models. The objectives of this phase I first-in-human and pharmacokinetic study of CUDC-427 were to determine (i) the safety and tolerability of daily oral administration of CUDC-427; (ii) the dose-limiting toxicities (DLT) on a 14-day on/7-day off schedule; (iii) the pharmacokinetic and pharmacodynamic properties; and (iv) preliminary evidence of anticancer activity in patients with advanced solid malignancies and lymphomas.

Eligible patients had pathologically confirmed solid malignancies or lymphoma refractory to standard therapy or for which no standard therapy existed: age ≥18 years; life expectancy ≥12 weeks; an Eastern Cooperative Oncology Group performance status of 0 to 1; resting oxygen saturation ≥92% on room air; previous chemotherapy ≥4 weeks (2 weeks for hormonal or kinase inhibitors and 6 weeks for prior mitomycin C or a nitrosourea); hemoglobin ≥9 g/dL; absolute neutrophil count ≥1,500/μL; platelets ≥100,000/μL; creatinine ≤1.5× upper limit of normal (ULN), bilirubin ≤1.5× ULN; aspartate serum transaminase (AST), alanine serum transaminase (ALT) ≤3× ULN; absence of pregnancy; prothrombin time/international normalized ratio ≤ 1.5× ULN; and no coexisting severe medical conditions. Patients were excluded if they had a clinically significant history of pulmonary, renal, or cardiovascular disease; chronic liver disease, active hepatic infection, evidence of hepatic cirrhosis, current alcohol abuse, active infections with hepatitis B or C virus; untreated brain metastases, uncontrolled ascites, known human immunodeficiency virus, gastrointestinal bleeding within past 6 months, history of malabsorption, or history of QT prolongation. Patients gave written informed consent according to federal and institutional guidelines before treatment.

CUDC-427 was administered orally daily on a 14-day on/7-day off dosing schedule (21-day cycles), except for cycle 1, which used 24 days to permit detailed single-dose pharmacokinetic sampling on day 1 and subsequently restarted on day 4 through 18. The starting dose, 5 mg, represented one tenth the rat severely toxic dose in 10% of animals (STD₁₀). Planned dose levels of 5, 10, 20, 40, 60, 90, 200, 300, 450, and 600 mg were examined using modified continuous reassessment method (mCRM) methodology. DLT was defined as any grade ≥3 nonhematologic toxicity and grade ≥4 hematologic toxicity except grade 3 nausea or vomiting or diarrhea responding to standard-of-care management, alopecia, and grade 3 fatigue lasting <72 hours. If one patient experienced DLT, the cohort was expanded to 6 patients. The MTD was defined during cycle 1 as the highest dose at which <33% patients experienced treatment-related DLT. Toxicity was graded according to the NCI's Common Toxicity Criteria, version 4.0.

Patients were evaluated for tumor response using CT or MRI per RECIST 1.1 criteria every 6 weeks (4). Patients received CUDC-427 until disease progression, intolerable adverse event (AE), or consent withdrawal.

Pretreatment evaluation included a complete history, physical examination, and routine laboratory studies, including a complete blood count (CBC), white blood count, chemistry (electrolytes, BUN, creatinine, uric acid, glucose, alkaline phosphatase, lactate dehydrogenase, ALT, AST, total bilirubin, calcium, total protein, albumin, cholesterol, and triglycerides), amylase, lipase, serum C-reactive protein (CRP), fibrinogen, urinalysis, pulse oximetry, diffusion capacity of the lung for CO (DL_CO), electrocardiogram, relevant radiologic studies, and tumor markers. During the study, radiologic studies for disease status were repeated after every other cycle, and tumor response was assessed by RECIST v1.1.

Repeat safety examinations, including history, relevant physical exam, CBC, and blood chemistry, were collected weekly. Amylase, lipase, CRP, fibrinogen, and coagulation studies were performed on day 1 of each cycle.

Plasma samples for pharmacokinetic parameters and MCP-1 levels were collected on day 1 (predose, 0.5, 1, 2, 3, 6, and 9 hours); days 2, 3, 4 (predose day 4), and 11 (predose); day 17 (predose and 0.5, 1, 2, 3, and 6 hours after dosing); and days 18 and 24 (predose). Urine was collected on day 1 predose, 0 to 6 hours postdose, and 6 to 24 hours postdose. Serum pharmacokinetic parameters were estimated using noncompartmental methods with WinNonlin Professional (Pharsight Corp).

Whole-blood samples were collected on day 1 (predose), 6 and 72 hours postdose, cycle1 day 18 (24 hours after day 17 dosing), and then predose on day 24. Detailed methods are described in the Supplementary Methods section.

Optional paired tumor biopsies were performed in consenting patients predose after 14 days of dosing on cycle 1 (day 18). Archived tumor specimens were obtained for molecular genetic characterization if available.

Forty-two patients, whose pertinent characteristics are displayed in Table 1, received a total of 119 cycles of CUDC-427 at doses ranging from 5 to 600 mg orally daily. The median number of cycles administered per patient was 2 (range, 1 to 15).

The mCRM specified 3 patients at dose level 1 and then a minimum of 2 patients at subsequent doses until the occurrence of a DLT or a non-DLT AE of interest. At cohort 5 (60 mg), 2 additional patients were entered to replace 2 nonevaluable patients (early progression and hypersensitivity). Six patients were entered at the 300 mg dose level to collect additional safety data after a grade 2 pneumonitis was observed in a patient in cycle 2. A DLT (grade 3 fatigue) in 1 patient and 1 patient nonevaluable due to disease progression necessitated expansion at the 450 mg dose level to 7 patients. At the 600 mg dose level, the allometrically scaled dose and target plasma concentrations exceeded the IC₉₀ in preclinical models and dose escalation ceased. The dose escalation, as well as DLT rate as a function of dose, is summarized in Table 2.

The principal toxicities related to CUDC-427 were mild to moderate in severity and included fatigue, nausea, vomiting, and rash. Severe (grade ≥ 3) toxicities were rare and included ALT and AST increase (2 patients each), anemia, fatigue, neutropenia, pruritus, pyrexia, and rash (1 patient each). The most frequent related AEs as a function of dose and grade are summarized in Table 3. Beyond the aforementioned single case of pneumonitis, none of the other 41 subjects had evidence of pneumonitis.

The specific rash related to CUDC-427 was characterized by faint, dry erythema distributed on the facial brow, bridge of the nose, malar eminence, and upper extremities, accompanied, in some cases, by pruritus.

All patients had measurable plasma CUDC-427 concentrations with the pharmacokinetic parameters on days 1 and 17 listed in Table 4 and concentration–time profiles illustrated in Fig. 1. Following single-dose oral administration, CUDC-427 was rapidly absorbed, with peak concentrations (C_max) generally occurring at 1 to 3 hours postdose. Mean half-life (t_1/2) ranged from 3.95 to 7.56 hours across all dose cohorts. The parameters that reflect exposure (C_max and AUC_0-t) increased proportionally over the doses range examined. Consistent with the short terminal t_1/2, accumulation of CUDC-427 in plasma was insignificant.

Statistically significant decreases in cellular cIAP-1 were measured in peripheral blood mononuclear cells (PBMC) samples from all patients on the study 6 hours after cycle 1 day 1 dosing of CUDC-427, regardless of dose level, although greater diminution (>90%) occurred consistently at doses that equaled or exceeded 300 mg (Table 5). Partial recovery of cIAP1 levels were observed by day 4 (after no treatment days 2 and 3) but remained below baseline levels. After 14 days of continuous dosing, cIAP-1 levels were significantly decreased from baseline, with many patients having undetectable cIAP-1 protein, and levels recovered only after the 1-week dose holiday (prior to cycle 2 day 1, data not shown). Four patients underwent optional paired tumor biopsies for cIAP-1 levels. In 2 patients, the tumor specimens were insufficient for analysis, whereas in 2 patients with sufficient paired tumor specimens, cIAP-1 levels decreased 75% of predose concentrations consistent with target activity of CUDC-427.

Blood specimens were analyzed for MCP-1, TNFα, IL6, IL8, and IFNγ. There was a trend toward higher plasma concentrations of MCP-1 at 9 hours after a single dose and 6 hours after 14 daily doses of CUDC-427, although there were no statistically significant changes in plasma MCP-1 observed in response to higher CUDC-427 dose cohorts. In addition, there were no statistically significant changes to plasma concentration of TNFα, IL6, IL8, and IFNγ in response to CUDC-427 treatment (data not shown).

Thirty-six patients with diverse malignancies were evaluable for response and 2 patients had evidence of a complete response (CR). The remainder of the 34 patients did not have an objective response.

One patient at the 300 mg dose level with a mucosa-associated lymphoid tissue (MALT) lymphoma of the stomach had a CR on PET after 3.4 months of study treatment, remained on treatment for 5 cycles completing treatment with a sustained CR, and is alive without treatment 2+ years after discontinuing CUDC-427 without evidence of disease recurrence.

Another patient with a BRCA1 mutation (germline) and platinum-refractory ovarian cancer with biopsy-proven recurrent peritoneal disease had a CR of visible disease 6 weeks after beginning therapy at 450 mg dose level. The patient only received 4 cycles, two of which required 2 dose reductions due to severe rash, mucositis, and neutropenia before discontinuing due to AEs. Despite discontinuing CUDC-427, the patient sustained a CR for 18 months before relapsing in the original peritoneal site. This ovarian cancer patient experienced toxicities of greater severity, as well as some that were unique to this patient (mucositis, photosensitivity with generalized rash, and neutropenia), compared with other patients at the same or higher dose levels.

Abrogation of apoptosis is a hallmark of malignancies, yet despite the recognized importance in our understanding of cancer, there are currently no approved therapies that directly target the apoptosis-regulatory pathways.

In this study, CUDC-427 was well tolerated and could be safely administered to the highest planned dose level of 600 mg. At the three highest dose levels, a constellation of mild or modest nonhematologic AEs occurred and included fatigue, nausea, vomiting, and rash. One DLT grade 3 fatigue was observed at the 450 mg dose during cycle 1, with additional toxicities in another patient (BRCA1 ovarian) requiring dose reduction on subsequent cycles. Cumulative but reversible toxicities, although not dose limiting, included fatigue, rash, and pruritus, were important and did contribute to the 2 patients who discontinued even though they were responding to treatment. Abnormal liver function tests occurred in 11.9% of patients; none of these met Hy law criteria for idiosyncratic, potentially fatal liver disease.

In general, the AUC and C_max values were dose proportional up to 600 mg and accumulation was insignificant. The pharmacokinetic profile for CUDC-427 behaved predictably, and concentrations could be achieved at dose levels that exceeded those that predicted antitumor activity in preclinical models. The highest planned dose of 600 mg achieved in this study exceeded the IC₉₀ of CUDC-427 in preclinical models.

Pharmacodynamic studies confirmed target engagement with the 3 highest dose levels of CUDC-427 profoundly inhibiting cIAP-1 protein levels in PBMCs, and in many patients, below the level of detection for this assay. The inhibitory effect persisted, albeit to a lesser degree, for 72 hours after discontinuation of the study drug.

The profound activity observed in 2 patients with CRs that were durable warrants further study of this drug in selected indications and patients. MALT lymphoma is characterized by several genomic events including trisomy 3, 7, 12, 18, and disease-specific translocations of t(1:14)(p22;q23), t(14;18)(q32;q21), t(8:11)(q21;q21), and t(3;14)(p13;q32). Many of these trisomies and translocations involve the activation of NF-κB, which regulates the expression of genes implicated in the survival and proliferation, including BIRC3 and MALT1, of B cells (5, 6). In the patient that responded, molecular analysis did not reveal the aforementioned translocations but rather gene amplification in the genes for BIRC3 and MALT1 genes by array CGH. In contrast, the ovarian cancer patient who experienced a CR, but also had profound normal tissue toxicity, had a germline BRCA1 mutation. BRCA1 has been linked to aberrant apoptosis execution through the regulation of SIRT1, a regulator of survivin and the apoptosis pathway (7, 8). This finding raises the question whether cIAP-1 inhibition in normal tissues bearing the BRCA1-inactivating mutation was responsible for the toxicities observed.

Further work with CUDC-427 has begun utilizing continuous dosing schedules as well as exploration in the non-Hodgkin lymphoma indication. IAP inhibitors in general may have the greatest utility in combination with apoptosis-inducing agents. Preclinical data indicate that CUDC-427 can be combined and enhances antitumor activity with cytotoxics and TRAIL-targeting agents (Curis and Genentech unpublished data).

In conclusion, CUDC-427 can be safely administered to patients with advanced malignancies at doses up to 600 mg orally daily for 14 days every 3 weeks, attains plasma concentrations that exceed the IC₉₀ in preclinical models, profoundly decreases cIAP-1 protein levels, and demonstrated promising durable CR in 2 patients.

A. Tolcher is an employee of Symphogen and is a consultant/advisory board member for Akebia, AP Pharma, Arqule, Asana, Astex, Bayer Schering Pharma, Bind, Blend, Celator, Dicerna, Eli Lilly, Endocyte, Genmab, Heron, Idea Pharma, Janssen, Johnson & Johnson, LiquidNet, MedImmuner, Mersana, Merus, Nanobiotix, Pharmacylics, Pierre Fabre, Symphogen, Proximagen, Valent, and Upsher-Smith. N. Budha and M. Mamounas have ownership interest (including patents) in Roche. No potential conflicts of interest were disclosed by the other authors.

Conception and design: A.W. Tolcher, H.A. Burris, W.J. Fairbrother, H. Wong, N. Budha, M. Mamounas, S. Royer-Joo, R. Yu, C.C. Portera, J.R. Infante

Development of methodology: A.W. Tolcher, H.A. Burris, H. Wong, W.C. Darbonne, R. Yu, C.C. Portera, J.R. Infante

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A.W. Tolcher, J.C. Bendell, K.P. Papadopoulos, H.A. Burris, A. Patnaik, W.C. Darbonne, F. Peale, J.R. Infante

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A.W. Tolcher, H.A. Burris, A. Patnaik, H. Wong, N. Budha, R. Yu, J.R. Infante

Writing, review, and/or revision of the manuscript: A.W. Tolcher, J.C. Bendell, K.P. Papadopoulos, H.A. Burris, A. Patnaik, W.J. Fairbrother, H. Wong, W.C. Darbonne, S. Royer-Joo, R. Yu, J.R. Infante

Study supervision: A.W. Tolcher, H.A. Burris, M. Mamounas, C.C. Portera, J.R. Infante

This trial was sponsored by Genentech Inc: ClinicalTrials.gov ID, NCT: 01908413.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.