Phase I Study of ONO-4007, a Synthetic Analogue of the Lipid A Moiety of Bacterial Lipopolysaccharide¹ Free

The possibility of treating cancer by the stimulation of the immune system has long been recognized. In the 1890s, William B. Coley inoculated cancer patients with bacterial extracts (Coley’s toxins) to activate general systemic immunity (1, 2). Tumor regression was reported in several patients, which could have been cytokine mediated in response to bacterial products such as LPS³(3). Subsequently, numerous studies were performed using intact bacteria or bacterial products as nonspecific adoptive immunotherapy of malignant diseases including acute lymphoblastic leukemia, malignant melanoma, and bladder cancer (4, 5, 6). There are a wide variety of biological and immunological responses associated with Gram-negative bacteria, mainly due to the lipid A moiety of the LPS contained within the cell wall (7). The administration of LPS can induce several harmful effects including pyrogenicity, septic shock, and disseminated intravascular coagulation. However, LPS also induces significant antitumor and antimetastatic activity (8). ONO-4007 is a low molecular weight synthetic analogue of the Lipid A subunit of LPS developed by ONO Pharmaceutical Co., Ltd. by chemically modifying the structure of lipid A (9).

In preclinical models, ONO-4007 demonstrated antitumor activity with considerably less toxicity than LPS. In vivo studies in mice bearing the murine tumors Meth-A (sarcoma), MH-134 (hepatoma), M-5076(reticulum cell sarcoma), and MM-46 (mammary carcinoma) demonstrated significant tumor inhibitory effects, in some cases with complete regression of tumors (9, 10). Additional studies conducted with nude mice bearing transplanted human pancreas cancer cells, MIA paca-2 and Panc-1, demonstrated significant inhibition of tumor growth and prolongation of survival compared to controls (11). These effects were associated with increased intratumoral TNF-αactivity. In vitro studies with peripheral blood monocytes or myelomonocytic cell lines cultured with ONO-4007 demonstrated induction of TNF-α production (12). These investigations have shown that ONO-4007 induces the secretion of TNF-α, IFN-γ,IL-1α, IL-6, and GM-CSF in a dose-proportional manner in splenic mononuclear cells. There appears to be no cytotoxic effect on tumor cells in the absence of macrophages, suggesting that the antitumor effect of the drug is dependent on the stimulation of macrophages to express these cytokines. These findings are in keeping with previous studies performed on murine tumor cells transfected with human TNF-α (10). In addition, lipid A analogues also inhibit tumor angiogenesis (13), partly through induction of TNF-α (14).

Preclinical toxicology studies⁴ revealed significant interspecies variation, with rats being the most sensitive species. In rats, death occurred after a single administration at 50 mg/kg or higher, with toxicological changes at doses above 5 mg/kg when administered repeatedly (every day for 4 weeks to 6 months, or intermittently for 6 months). In dogs, no marked toxicological changes were noted after a single administration of 50 mg/kg or repeated administration (every day for 2–4 weeks) at a dose of 10 mg/kg. In monkeys, no toxicological changes were observed after a single dose of 50 mg/kg. However, anemia, thrombocytopenia, prolonged coagulation time, changes in liver function tests, hemorrhagic findings in various organs, necrosis of hepatocytes, and death were reported with daily repeated administration (daily for 2 weeks or intermittently for 4 weeks) at 25 mg/kg or higher. The symptoms of toxicity seen in monkeys on repeated administration are thought to be due to the maintenance of very high blood levels of the drug for extended periods of time.

Based on this preclinical data, a three-center Phase I study was performed. The main aims of this study were as follows: (a)to determine the safety and toxicity of an initial single dose of ONO-4007; (b) to determine the MTD of ONO-4007 when administered as an intermittent treatment cycle; (c) to evaluate the PK and PD of ONO-4007; and (d) to record any antitumor activity.

Eligible patients were recruited at three collaborating centers(Beatson Oncology Center, Glasgow; St. George’s Hospital, London; and City Hospital, Nottingham). The protocol was approved by the local research ethics committee at each institution. All patients gave written informed consent. Patients with histologically or cytologically confirmed malignancy unresponsive to or untreatable by standard therapy were eligible for this study. Other eligibility criteria included:(a) the presence of measurable or evaluable disease;(b) age ≥ 18 years; (c) Karnofsky performance status ≥ 70%; (d) life expectancy of ≥4 months; (e) body weight between 40 and 90 kg; (f)adequate hematological (hemoglobin ≥ 10 g/dl; leukocytes ≥4 × 10⁹/liter; granulocytes ≥ 2 × 10⁹/liter; platelets ≥ 100 ×10⁹/liter), renal (serum creatinine < 1.5×the upper limit of normal), and hepatic function (serum bilirubin < upper limit of normal; transaminases and alkaline phosphatase < 2.5× the upper limit of normal); (g) no chemotherapy,radiotherapy, immunotherapy, endocrine therapy, or major surgery within the 4 weeks prior to commencing ONO-4007; and (h) no systemic corticosteroids within 2 weeks of starting treatment. Exclusion criteria included cerebral metastases, pregnancy or lactation, psychiatric illness, clinically significant medical conditions, major fluid collections, a history of infection with HIV or hepatitis B or C, a history of autoimmune disease or organ transplantation, and concurrent systemic Gram-negative bacterial infections.

The study was performed according to good clinical practice guidelines. Baseline laboratory evaluations included a full blood count with differential, clotting factors, urea and electrolytes, serum creatinine, calcium, phosphate, total protein, albumin, glucose,alkaline phosphatase, aspartate aminotransferase, alanine aminotransferase, bilirubin, thyroid function tests, and uric acid. Complete medical history, physical examination, radiological assessment of disease, urinalysis, chest X-ray, electrocardiogram, and tumor markers (if appropriate) were also performed. In addition, a full blood count with differential, clotting factors, biochemical parameters, and vital signs was assessed weekly during the study. Response evaluation was performed by physical examination and radiology and assessed according to the WHO criteria after each treatment cycle.

ONO-4007 was supplied by ONO Pharmaceutical Company Limited, Japan, as an injectable formulation in ampoules containing 50 or 100 mg of ONO-4007 dissolved in 1 or 2 ml of 55% ethanol, respectively. ONO-4007 was aseptically diluted to a final volume of 100 ml in 5% dextrose before administration and then stored below 5°C and used within 8 h of reconstitution. Because ONO-4007 has similarities to bacterial cell wall, patients previously exposed to this antigen could potentially develop an allergic reaction. Consequently, a prick test was performed 24 h before administration of the first infusion. The prick test was considered positive if the average diameter of the wheal associated with erythema was >3 mm after 15 min, or if the erythema was >3 mm around the test site after 24 h. Patients with a positive prick test were then withdrawn from the study. ONO-4007 was administered as a single i.v. infusion over 30 min via a peripheral venous catheter. The first infusion was followed by a 2-week observation period. The first cycle of treatment was then commenced with a repeated dose schedule in which ONO-4007 was administered weekly as a 30-min infusion for 3 consecutive weeks, followed by a 1-week rest period. This 4-week cycle was repeated up to a maximum of four cycles of treatment. The patient and the investigator had the right to stop treatment with ONO-4007 at any time for any reason, including unacceptable toxicity, intercurrent illness, treatment failure, or protocol violation. Patients who experienced drug-related NCI-CTC grade 3 toxicity could be retreated after complete resolution of the adverse event at the dose level of the preceding cohort.

Pulse, blood pressure, and body temperature were recorded immediately before and at the end of each infusion of ONO-4007, and these vital signs were also recorded 2 h after the end of the first single infusion and after the first infusion of cycle 1 of therapy. Diclofenac sodium was used for both the treatment and the prophylaxis of acute flu-like symptoms induced by ONO-4007. Body weight was also recorded at baseline and at the start of each new treatment cycle. Furthermore, the full blood count was measured 2 h after completion of the first single infusion and the first infusion of cycle 1 of therapy.

The starting dose in this study was based on preliminary information available from unpublished studies in the Japanese population.⁵ The starting dose was 75 mg/patient. This dose level was felt to be safe based on the safety profile of the drug at 50 mg/patient in the multiple administration study in Japan and accounting for the difference in body surface area between the Japanese and Occidental populations (1.5 m² compared with 1.75–1.8 m²). The subsequent planned dose levels were 100,125, and 150 mg/patient, with further dose escalation in 25-mg increments as necessary. There was no dose correction for body surface area, in keeping with the administration schedules of most immunotherapy agents and other agents thought to act as cytostatic agents rather than cytotoxic agents. The PK data from the preliminary Japanese studies demonstrated almost complete clearance of the drug within 1 week of administration. Given the relatively long half-life of the drug, a weekly administration was proposed that was predicted to avoid the accumulation of the drugs. A rest period (1 week) was incorporated into each 4-week treatment cycle to allow recovery from any possible toxicity and to avoid or reduce the rebound down-regulation of the immunological reaction observed with other immunotherapies. Cohorts of six patients were entered at each dose level, with no dose escalation in individual patients. Toxicities were recorded in all patients on a weekly basis using the standard NCI-CTC. DLT was defined as any toxicity of ≥NCI-CTC grade 3, with the exception of pyrexia treatable by nonsteroidal anti-inflammatory drugs or toxicities considered by the investigator to be unrelated to the study drug. Dose escalation was performed when all six patients in any cohort had received at least the first single infusion of ONO-4007. The MTD was defined as the dose level of ONO-4007 at which two of six patients in one cohort expressed DLT. An additional cohort of six patients was then recruited at the dose level immediately below the MTD to further characterize the toxicity profile.

ONO-4007 was discontinued if a patient experienced any NCI-CTC grade 4 toxicity, unless the investigator considered it in the patient’s best interest to continue at a lower dose. If a patient experienced drug-related grade 3 toxicity, drug treatment was withheld until the toxicity had recovered to ≤grade 1, with subsequent treatment being at the dose level of the preceding cohort or reduced by 50% if the toxicity occurred at the first dose level. If the same grade 2 toxicity recurred, the drug was again withheld until recovery to ≤grade 1 and then recommenced at the dose level of the preceding cohort (or at 50%of the dose if this toxicity occurred at the first dose level). There were no dose escalations.

PK and PD samples were analyzed at Quintiles Scotland Limited(Edinburgh, United Kingdom). The PK of ONO-4007 were studied after both the initial single infusion and the multiple dose administration for the first treatment cycle only. For the initial single infusion, blood samples (5 ml) were taken predose, at the end of the infusion, and 0.5, 1, 1.5, 2, 6, 12, 24, 72, 120, and 168 h after the end of infusion. For the first repeat administration cycle, blood samples were taken predose, at the end of the infusion, and 1 h after the end of infusion for the first two weekly administrations; for the third week, the sampling times were as described for the initial single infusion administration. All blood samples were collected into tubes containing sodium heparin as anticoagulant and centrifuged at (300 rpm)for 10 min at 4°C, and at least 2 ml of plasma were retained in a separate tube and stored at −20°C until analysis. Plasma concentrations of ONO-4007 were measured using reverse phase-HPLC. The following PK parameters were estimated using noncompartmental methods and WinNonlin Version 1.1. from the drug concentration-time data obtained: C_max, T_max, apparent terminal rate constant(k), apparent terminal half-life(t_1/2), AUC_last,AUC_inf, V_ss and clearance, observed degree of accumulation(AUC_last day 29:AUC_last day 1) and theoretical degree of fraction.

For PD analyses, 10-ml blood samples were collected pre-dose, and at 1,2, and 6 h after the end of the initial single infusion administration. For the first repeat administration treatment cycle,blood samples were collected pre-dose and 1 h after the end of the infusion for the first 2 weekly administrations and collected pre-dose and at 1,2, and 6 h after the end of the infusion for the third week. All blood samples were collected into serum separating tubes, and at least 5 ml of serum were transferred to separate tubes and stored at−20°C until analysis. TNF-α, IL-6, GM-CSF, and IFN-γ were measured using the Immunotech enzyme immunoassay at Quintiles Laboratories. Neopterin was measured using IBL enzyme immunoassay.

Patient characteristics are summarized in Table 1. Twenty-four patients, 11 females and 13 males, with a median age of 53.5 years (range, 33–77 years) were treated with ONO-4007 at three escalating dose levels. The most common tumor types included malignant melanoma (seven patients), renal cell carcinoma (five patients), and colorectal carcinoma (four patients).

All patients had a negative prick test and received at least the initial single administration of ONO-4007. A total of 23 patients went on to complete a total of 47 cycles of treatment (median number of cycles, 2). Five patients completed the planned four cycles of treatment, with one patient receiving a fifth course at the investigator’s discretion. Reasons for discontinuing therapy before completion of the planned four cycles included treatment failure (12 patients), disease progression (3 patients), toxicity (1 patient),and withdrawal of consent (1 patient).

All 24 patients who had received at least the single administration of ONO-4007 were evaluable for toxicity (Table 2). The most common toxicities were acute reactions, usually commencing soon after the start of the infusion and occurring in over 20% of patients. These included rigors, fever,nausea, vomiting, flushing, dyspnea, headache, back pain, fatigue, hot flushes, chest pain, dizziness, and hypotension. In the majority of cases, adverse events were ≤ NCI-CTC grade 2, of short duration,and usually recovered completely within 24 h.

The MTD was the 125 mg dose level, with DLT being experienced by two of the six patients treated at this dose. In both cases, the dose-limiting event was NCIC-CTC grade 3 rigors (Table 3) as defined by the presence of cyanosis. In both cases, the patients made a rapid and uneventful recovery within 30 min, although both required oxygen therapy. For one patient, no oxygen saturation was determined before the initiation of oxygen therapy, and this grade 3 rigor was associated with a drop in blood pressure from 170/90 (preinfusion) to 150/60 during the rigor. For the other patient, grade 3 rigor commenced 15–20 min after completion of the infusion and with an oxygen saturation of 88%. This returned to 100% rapidly after initiating oxygen therapy. This was associated with a temperature of 39.8°C, but no episode of hypotension. Subsequently, one patient refused further treatment, and the other patient went on to complete the four planned cycles of treatment at the lower dose of 100 mg. ONO-4007 appeared to have little overall effect on total leukocyte or neutrophil counts. However, there was a significant drop in lymphocyte counts immediately after administration of ONO-4007, although the majority of patients had low lymphocyte counts before starting treatment. Eight patients experienced a decrease in their hemoglobin during the study, with three patients experiencing grade 3 or 4 anemia during the study. In these three cases, this was thought to be due to underlying disease. ONO-4007 had no significant effects on clotting parameters or platelet counts. Changes in electrolytes were generally minor and were not considered to be drug related. Hepatotoxicity was not seen.

There was no apparent increase in the severity of adverse events across the dose levels, with the exception of rigors and fever, which were more frequent at the highest dose level. Furthermore although rigors had been observed at lower dose levels, their severity and duration were more pronounced at the 125 mg/patient dose level.

Overall, ONO-4007 was well tolerated, with little in the way of significant toxicity other than this acute reaction. The majority of adverse events were mild/moderate in severity (NCIC CTC grade 1 or 2),and there was no discernible increase in toxicity with continued treatment. Indeed the frequency of rigors and fever was higher after the initial dose than that observed with subsequent treatment cycles(Table 4), most probably due to the frequent use of prophylactic diclofenac with subsequent treatment cycles (diclofenac was not to be given before the first dose). Although three patients died during the follow-up phase of the study, all deaths were considered to be due to progression of the underlying malignancy and were not associated with the study medication.

Eighteen of 24 patients were evaluable for response. The remaining six patients did not have repeat tumor assessments, primarily due to early clinical disease progression. No patient achieved a partial or complete response to therapy. Nine patients maintained stable disease for at least one cycle, with the remainder discontinuing treatment during or on completion of the first cycle of treatment, primarily because of disease progression. Seven patients maintained stable disease for at least two cycles, with five patients maintaining stable disease for the full four-cycle duration of the study. The five patients with disease stabilization included melanoma (two patients) and renal, colorectal,and non-small cell lung carcinoma (one patient each). Four patients had tumor markers sampled during the study, of which only three had postbaseline measurements. One patient with colorectal cancer maintained levels of carcinoembryonic antigen similar to those obtained at baseline, correlating with disease stabilization as assessed by radiological tumor measurements. No patient had a significant reduction in circulating levels of tumor markers during therapy.

After the 30-min i.v. infusions of 75, 100, and 125 mg of ONO-4007,plasma concentrations declined in a biphasic manner, with a mean apparent terminal half-life of 74–95 h (Fig. 1). Peak plasma concentrations of ONO-4007 increased in a dose-proportional manner from 20 μg/ml at 75 mg to 40 μg/ml at 125 mg. Mean AUC also increased approximately proportionately with dose (Fig. 2). The PK of ONO-4007 appeared to be time invariant; the accumulation of ONO-4007 in plasma was predictable and consistent with linear kinetics.

The mean AUC and C_max varied in a linear fashion with dose (Fig. 2). ONO-4007 displayed time-invariant kinetics.

ONO-4007 has a low systemic clearance (approximately 1.3 ml/min) with a small volume of distribution (6–8 liters), which results in the relatively long apparent terminal half-life (Table 5). The volume of distribution of ONO-4007 has a value equivalent to that of the distribution volume of albumin, suggesting that ONO-4007 is largely confined to the plasma.

The pharmacodynamic analyses were based on the evaluation of circulating levels of TNF-α, IFN-γ, IL-6, GM-CSF, and neopterin in blood collected on days 1, 15 22, and 29. All of the patients were evaluable at least at one time point (either on day 1 or day 15) for evaluation of pharmacodynamic effects. Furthermore all of the patients showed a significant increase in TNF-α after administration of ONO-4007, with the majority of patients having at least a 10-fold increase in levels posttreatment (Fig. 3). TNF-α levels decreased rapidly after the completion of the infusion, although they were generally still above the predose levels 6 h later. Similarly, IL-6 circulating levels increased markedly in all patients after treatment and then decreased rapidly but remained above the predose level 6 h after infusion (Fig. 4). There were no significant changes in the circulating levels of GM-CSF, IFN-γ, or neopterin after the administration of ONO-4007.

ONO-4007 was generally well tolerated at doses of up to 100 mg,when given as a once-weekly infusion for 3 weeks out of every 4 weeks. Fever and rigors were the most common adverse events, with rigors being the DLT. The MTD of ONO-4007 is 125 mg, with dose-limiting grade 3 rigors being seen at this dose level. Other adverse events included influenza-like symptoms, fatigue, headache, and dizziness. These were generally mild or moderate and reversible on interruption of treatment or with concomitant treatment and did not increase in severity with retreatment. The use of prophylactic diclofenac attenuated these acute adverse reactions. No clinically significant hematological toxicity was observed other than a mobilization of lymphocytes, with no apparent effect on blood clotting. Furthermore, no clinically significant renal or hepatic toxicities were noted.

The findings described by this study are comparable to the results obtained by the concurrent ongoing Japanese study that reported a similar toxicity profile, DLT, and MTD. In the Japanese study,ONO-4007 was well tolerated at single doses up to 75 mg, with the MTD defined as 100 mg/patient with grade 3 pyrexia, chills, cough, dyspnea,tachycardia, Raynaud’s syndrome, and abnormal coagulation being reported. When adjusted for differences in body surface area, the MTD is similar to that observed in this study. The toxicities observed with ONO-4007 are similar to those observed with other studies with purified LPS (15, 16), namely, constitutional symptoms such as fever, rigors, fatigue, and headache, and these are also among the toxicities of systemic administration of recombinant TNF (reviewed in Ref. 17). Furthermore, when ibuprofen was used to abrogate these toxicities and enable dose escalation in one of these studies with LPS (15), additional toxicities were reported including dyspnea (without a decrease in oxygen concentration),hypotension, and hepatic toxicity, the latter of which was dose limiting. A similar pattern of toxicities has also been reported with other agents that possess macrophage activator properties including GM-CSF (18), macrophage colony-stimulating factor (19), and muramyltripeptide-phosphatidyl ethanolamine (20, 21).

The PK of ONO-4007 appears to be independent of dose within the dose range of 75–125 mg and shows linearity with respect to time. ONO-4007 is a drug with low systemic clearance (1.3 ml/min) and a small volume of distribution (6–8 liters). Because this volume of distribution is equivalent to that of albumin, ONO-4007 appears to be largely confined to the plasma. This results in a long apparent terminal half-life(74–95 h). Consequently, consideration should be given to a revised schedule of administration in future studies, e.g., twice weekly, although it has previously been reported that shorter intervals between repeated injections of LPS lead to the development of LPS tolerance (15).

Pharmacodynamic studies have shown that the administration of ONO-4007 results in a significant increase in the release of TNF-α and IL-6. These levels increase and subsequently decrease rapidly, lasting for approximately 6 h. These studies also suggested that there is a trend for higher ONO-4007 doses to result in higher circulating levels of TNF-α. There was no significant change in circulating levels of GM-CSF, IFN-γ, or neopterin after ONO-4007 administration. However,all patients had advanced malignancy, and most had received at least one prior systemic treatment modality, all of which could have contributed to the suppression of the patients’ immune system such that no increase in circulating levels of these cytokines was produced on stimulation by ONO-4007.

The rise in IL-6 levels, which remain above the baseline, may in fact be a detrimental effect of ONO-4007 because increased IL-6 levels are associated with a variety of malignancies (22), especially multiple myeloma. It may be possible to enhance the antitumor activity of ONO-4007 by decreasing IL-6 (e.g., with anti-IL-6 antibody) in future studies.

No objective tumor responses were observed in this study. However,disease stabilization was observed in five patients for the duration of the study (18 weeks), including two patients with malignant melanoma and one patient with renal cell carcinoma. In addition, two patients achieved disease stabilization in tumor types not usually responsive to immunotherapy (one colorectal carcinoma and one non-small cell lung carcinoma). The prolonged disease stabilization may suggest that this agent altered the natural history of these tumors (23). However, optimal benefit from immunotherapy is more likely to be acquired in patients with minimal residual disease.

This study has therefore defined the MTD of ONO-4007 and the recommended dose for Phase II studies to evaluate the activity of ONO-4007 in tumors amenable to immunotherapy. However, the optimal dose for immunomodulating agents may not necessarily equate to the recommended dose determined by toxicity in clinical Phase I studies. Preclinical in vivo studies with TNF-α suggest that there is a therapeutic window in which the beneficial antitumor effect can be obtained without the detrimental (cachectic) effects seen with prolonged exposure to higher levels of TNF-α (24). In these murine preclinical studies, persistent serum levels of TNF-αabove 250 pg/ml were shown to cause severe cachexia, whereas lower levels appeared to be able to inhibit tumor growth in the absence of severe weight loss. In this Phase I study, TNF-α serum levels above 250 pg/ml were observed for up to at least 6 h after completion of the infusion at all three dose levels tested. Despite the limitations of mouse models for predicting the optimal biological dose, they do represent the best preclinical guide for the use of biological agents. This Phase I study has determined the MTD on the basis of clinical toxicity. However, the aim of future studies should be to determine the biologically effective dose of ONO-4007 based on serum TNF-α levels and ideally by measurement of intratumoral TNF-α levels as a pharmacodynamic end point in patients with biopsy-accessible disease. This would be particularly relevant in correlating biological activity with objective antitumor activity as the primary clinical end point within Phase II clinical trials.

We are grateful to the Cancer Research Campaign, to all of the staff involved in the care of the patients in this study, and to Michelle Herron and Lorraine Blower (Quintiles, Edinburgh, United Kingdom) for assistance in the bioanalytical and PK assessments.