|Home | About | Journals | Submit | Contact Us | Français|
This phase 1 study was conducted to determine the recommended phase 2 dose of the selective insulin-like growth factor type 1 receptor (IGF-IR) inhibitor figitumumab (F, CP-751,871) given in combination with paclitaxel and carboplatin in patients with advanced solid tumors.
Patients received paclitaxel 200 mg/m2, carboplatin (area under the curve of 6), and F (0.05–20 mg/kg) q3 weeks for up to six cycles. Patients with objective response or stable disease were eligible to receive additional cycles of single agent F until disease progression. Safety, efficacy, pharmacokinetic, and pharmacodynamic endpoints were investigated.
Forty-two patients, including 35 with stages IIIB and IV non-small cell lung cancer (NSCLC), were enrolled in eight dose escalation cohorts. A maximum tolerated dose was not identified. Severe adverse events possibly related to F included fatigue, diarrhea, hyperglycemia, gamma glutamyl transpeptidase elevation, and thrombocytopenia (one case each). F plasma exposure parameters increased with dose. Fifteen objective responses (RECIST) were reported, including two complete responses in NSCLC and ovarian carcinoma. Notably, levels of bioactive IGF-1 seemed to influence response to treatment with objective responses in patients with a high baseline-free IGF-1 to IGF binding protein-3 ratio seen only in the 10 and 20 mg/kg dosing cohorts.
F was well tolerated in combination with paclitaxel and carboplatin. Based on its favorable safety, pharmacokinetic, and pharmacodynamic properties, the maximal feasible dose of 20 mg/kg has been selected for further investigation.
The insulin-like growth factor (IGF) system comprised the IGF ligands (IGF-1 and IGF-2), the IGF binding proteins (IGFBPs 1–6) that regulate their bioactivity, and the cell surface receptors IGF-1R and IGF-2R.1 Signaling through the IGF-1R plays important roles in normal growth and development and in cancer progression.1 In NSCLC, the IGF-1R has been shown to be frequently expressed in tumor tissue and to mediate the proliferation of lung cancer cell lines.2 Also, high IGF-1 and low IGFBP-3 levels have been associated with a higher incidence and severity of the disease.3 These data suggest that targeting the IGF-1R could be a viable approach in the treatment of NSCLC.
Figitumumab (F) is a selective inhibitor of the IGF-1R, which has been well tolerated in initial studies. F increases the tumor growth inhibition activity of chemotherapy and targeted agents in preclinical models.4 This phase 1b study was conducted to define the recommended phase 2 dose (RP2D) of F when given in combination with standard doses of paclitaxel and carboplatin in patients with advanced solid tumors.
Patients with advanced tumors in whom the combination of paclitaxel and carboplatin was a reasonable treatment option were candidates for this study. Additional eligibility criteria included: age ≥18 years; Eastern Cooperative Oncology Group performance status ≤1; recovery from any prior treatment toxicity to the National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0 (CT-CAE 3.0) grade ≤1; adequate hematopoietic (ANC ≥1500/μL, platelets ≥100,000/μL), hepatic (total bilirubin ≤1.5 normal, aspartate aminotransferase and alanine aminotransferase <5 times normal), and renal (serum creatinine ≤1.5 times normal or calculated creatinine clearance ≥60 mL/min) function; serum albumin ≥2.7 g/dl; 12-lead electrocardiogram with normal tracing, or nonclinically significant changes that would not require medical intervention. Enrollment exclusions included evidence of moderate or severe mitral valve regurgitation on transthoracic Doppler echocardiogram; females of childbearing potential who refused adequate methods of contraception; active gastrointestinal abnormalities; symptomatic brain metastasis (patients with brain metastases were enrolled if stable); coexisting uncontrolled medical condition; neuropathy >grade 1; known hypersensitivity to Cremophor; and major surgical procedure or radiation therapy within 4 weeks or 1 week, respectively, before study drug administration. The protocol was conducted in accordance with Good Clinical Practice guidelines. All patients provided written informed consent.
The primary objective of the study was to determine the RP2D of F in combination with standard doses of paclitaxel and carboplatin. Toxicities were assessed using the National Cancer Institute CTCAE 3.0. Dose-limiting toxicities (DLTs) included grade ≥4 treatment-related neutropenia (ANC <500 cells/mm3) lasting >7 days, complicated by fever and/or requiring hospitalization; grade 4 thrombocytopenia (platelets <25,000 cells/ mm3); ≥grade 3 gastrointestinal toxicity despite optimal intervention and/or prophylaxis; any other F-related ≥grade 3 nonhematological toxicity; and any structural and/or functional changes in transthoracic Doppler echocardiography parameters. Patients received paclitaxel 200 mg/m2 intravenously (IV) more than 3 hours and carboplatin at an area under the curve (AUC) of 6, IV more than 15 to 60 minutes every 3 weeks. In the original protocol, F was administered IV after chemotherapy administration at doses of 0.05 to 10 mg/kg according a dose-doubling dose escalation schema.5 Based on the favorable safety profile of F in three ongoing dose escalation trials,6,7 this study protocol was amended to omit the 0.2 and 0.4 mg/kg F dose escalation cohorts, and only two patients were enrolled at the 0.1 mg/kg dosing cohort. Furthermore, based on the favorable safety and tolerability profile observed at the 10 mg/kg cohort, an additional cohort investigating the F dose of 20 mg/kg was introduced by protocol amendment. A dose of 30 mg/kg given every 3 weeks as a single agent was investigated in another study, not yet reported, and found to be not well tolerated. Consequently, doses higher than 20 mg/kg were not considered for testing in this study. Patients were eligible to receive up to six cycles of chemotherapy combined with F unless disease progression or unacceptable toxicity developed. On completion or early discontinuation of chemotherapy, patients were eligible to receive additional cycles of single agent F at a dose found to be well tolerated in two phase I dose escalation studies of single agent F conducted in parallel to this trial.6,7 To open a new dosing cohort, one patient in the previous cohort had to complete at least 3 weeks, and two additional patients at least 2 weeks, of safety follow-up without identification of DLTs. The 10 and 20 mg/kg cohorts were empirically extended to 17 and 7 patients, respectively, to obtain further safety and pharmacodynamic information.
Before enrollment, at each treatment visit and either at the completion of the study or premature discontinuation, patients underwent safety laboratory testing (blood chemistry, hematology, and urinalysis) and were queried for adverse events and concomitant medication use. Doppler echocardiograms were recorded at baseline, end of cycles 1 and 6, or at premature discontinuation. Plasma samples for detection of antidrug antibodies were collected at baseline, end of cycles 1 and 6, and 150 days from the last F infusion.
All patients had pharmacokinetic (PK) blood samples taken 30 minutes before each infusion,1 hour after the end of infusion during cycles 1 to 4, and 1, 3, and 7 days after the end of infusion at cycles 1 and 4. Additionally, PK samples were collected 30 minutes before infusion in cycles 5 and 6, at the end of study, and at 150 days after the last F dose. Plasma concentrations of F were analyzed as described previously.7
Biomarkers investigated included the enumeration of circulating tumor cells (CTCs), CTCs expressing the IGF-1R and circulating endothelial cells (CECs).8 Additional biomarkers included the measurement of plasma levels of the circulating cleaved soluble IGF-1R extracellular domain (sIGF-1R), free (bioactive) IGF-1 (fIGF-1), and IGF binding protein 3 (IGFBP-3) using enzyme linked immune assays (DSL, Webster, TX). Blood samples for the determination of CTCs and serum markers were collected before dosing at each treatment cycle. Additional samples were collected on day 8 and at the end of the study.
Tumor assessments were conducted at baseline and repeated every 6 weeks until disease progression. Disease status was assessed by investigator according to RECIST.
A total of 43 patients were enrolled between February 2005 and December 2007. One patient discontinued protocol before receiving treatment. The rest of the patients received F alone or in combination with paclitaxel and carboplatin for a median of five cycles. Demographic characteristics, performance status, disease stage, and tumor histologies are described in Table 1. Table 2 summarizes the treatments administered. No apparent dose-effect of F on the cumulative dose of chemotherapy was observed. Two study discontinuations due to combined treatment were reported and are described later. Paclitaxel was dose reduced in seven patients, six at the 10 mg/kg cohort (n = 17), and one at the 20 mg/kg cohort (n = 7). Carboplatin was dose reduced in three patients, two at the 10, and one at the 3 mg/kg (n = 3) dosing cohorts. Dose reductions were attributed to neutropenia, weight loss, and creatinine increase. Seventeen patients received single agent F on discontinuation of chemotherapy. Median and range of F single agent therapy were three and 1–62 cycles, respectively. One patient has received a total of 68 cycles, approximately 4 years, of F treatment (Table 2).
One patient died of disease progression while receiving study treatment with chemotherapy and 20 mg/kg of F. Table 3 summarizes all adverse events with a frequency higher than 5% and any ≥CTCAE grade 3 event reported by investigator as possibly related to F. Seventy-one adverse events possibly related to F were reported in 28 study subjects. Five of these events were grade 3 and included fatigue, diarrhea, hyperglycemia, gamma glutamyl transpeptidase (GGT) elevation, and thrombocytopenia (one case each). All of them were reported in patients enrolled in the 10 mg/kg dosing cohort. The grade 3 events of GGT elevation and thrombocytopenia were attributed to the combination regimen and resulted in treatment discontinuation. The event of hyperglycemia was manageable with antidiabetic medication. No DLT was reported at the 20 mg/kg cohort. Grade 2 events included arthralgia, weight loss, and anemia. Cardiac function was monitored using Doppler echocardiogram. Ninety-one Doppler echocardiograms were conducted without treatment related findings reported. No quantifiable antidrug antibodies were detected.
F plasma concentration-time data were available from 40 patients. One patient was excluded from the analysis because of incomplete data. Plasma concentrations of F decreased in a multiexponential manner after intravenous infusion (Figure 1). As shown in Table 4, the plasma concentration at the end of infusion (C1 hour) and the AUC within a cycle (AUC0–day 22) increased with dose at both cycles 1 and 4. The increase in AUC0–day 22 was approximately dose proportional at dose levels higher than 0.8 mg/kg. After repeated administration every 3 weeks, there was a moderate accumulation in plasma exposure of F at dose levels ≥3 mg/kg. Mean accumulation ratio reached approximately two-fold at 10 and 20 mg/kg q3 weeks dosing (Figure 1 and Table 4), supporting an effective half-life approximating the 3-week dosing interval.6,7
The effects of F on CTCs, IGF-1R expressing CTCs and CECs were investigated. Thirty-one patients provided blood samples (n = 674) for the enumeration of these biomarkers, including 26 patients with NSCLC. Seventeen patients had detectable CTCs at some point during the study; however, only one to five cells were detected in most cases, with all but one of the patients with NSCLC having ≤10 CTCs at study entry. Because IGF-1R has been described as an upstream regulator of VEGF expression,9 CECs were enumerated to evaluate any potential effect of F on this angiogenesis marker. CECs were detected in all patients, with a median of 52 cells per blood sample. Mean CEC counts increased with treatment cycle but no effect of F treatment on this pharmacodynamic parameter was observed (Figure 2).
Additional blood samples (n = 351) were analyzed for the determination of plasma levels of sIGF-1R, fIGF-1, and IGFBP-3. An inverse correlation (Rho = −0.426, p = 0.03) between baseline fIGF-1 and sIGF-1R was observed. No correlation with demographic parameters was identified. At low-F doses, transient decreases followed by rebound increases in circulating sIGF-1R levels were observed (Figure 3). In contrast, sIGF-1R levels were maximally suppressed for the entire dosing period at F doses levels of 3 mg/kg and above. Meanwhile, fIGF-1 and IGFBP-3 increased in patients in response to F treatment in a dose-dependent fashion, although the magnitude of increase in IGFBP-3 was generally more modest than that of fIGF-1 (Figure 3).
Fifteen objective responses (RECIST) were observed (36%), including two complete responses, one in a patient with ovarian carcinoma, and one in a patient with stage IV NSCLC. Fourteen objective responses were seen in patients with stages IIIB and IV NSCLC (40%). Median duration of response was 6.5 months (range, 2–49). In addition, 16 patients with NSCLC experienced a best response of stable disease. The median duration of disease stabilization was 2 months (range, 1–11). Six patients were stable for ≥6 months on single agent F maintenance therapy.
No apparent relationships between efficacy and PK, CTC, or CEC parameters were identified. Baseline blood samples for the analysis of plasma markers were available from 40 patients, including 13 patients with objective responses. Patients responding to treatment at F doses ≤6 mg/kg seemed to have high baseline IGFBP-3 and low fIGF-1 levels (Figure 4). As both parameters are likely interrelated, we analyzed the ratio of fIGF-1 to IGFBP3 according to F dose (≤6 mg/kg versus 10–20 mg/kg) and objective response. Of interest, patients responding to treatment at F doses ≤6 mg/kg had fIGF-1 to IGFBP-3 ratios lower than patients responding in the 10 and 20 mg/kg cohorts (p = 0.04) (Figure 4), suggesting that low bioactive IGF-1 levels may be associated with response to chemotherapy in the absence of IGF-1R blockade. Indeed, nine responses were observed in 16 patients with a fIGF-1/IGFBP3 ratio ≤0.018%. On the other hand, only five of 24 patients with higher fIGF-1/IGFBP3 ratios responded (p = 0.04), and all these responders received F at 10 or 20 mg/kg. No association of sIGF-1R levels at study entry or degree of down-regulation post-treatment with patient response was identified.
An every 3-week regimen of the anti-IGF-1R monoclonal antibody F at doses up to 20 mg/kg in combination with paclitaxel and carboplatin was well tolerated. Grade 3 events of hyperglycemia, fatigue, diarrhea, GGT elevation, and thrombocytopenia (one case each) were reported as possibly related to F. Hyperglycemia seems to be characteristic of this class of compounds,10 although contributions of comorbid conditions, such as diabetes and steroid premedication for paclitaxel cannot be excluded. Importantly, the event of hyperglycemia was generally manageable with antidiabetic medication. Mild to moderate fatigue and sporadic isolated GGT elevations have been observed in previous studies of F. In contrast, severe thrombocytopenia has been rare when F was given as single agent.6,7 This is important because hematological toxicity has been reported to be dose limiting for other anti-IGF-1R antibodies.11–13 It is possible that differences in IgG subtype may be responsible for some of these differences. F is a fully human IgG2 antibody4 and, consequently, is expected to be a poor activator of antibody mediated cytotoxicity and complement fixation.
The extended effective half-life of F of approximately 3 weeks at 10 and 20 mg/kg doses compares favorably with other monoclonal antibodies in the anti-IGF-1R class.7 Mean terminal half-lives of 4 to 11 days have been reported for other IGF-1R monoclonal antibodies.9 These differences could be due in part to their IgG backbone.14 PK exposure parameters of F given in combination with paclitaxel and carboplatin were similar to those previously observed with single agent F,9 suggesting that the combination with paclitaxel/carboplatin does not significantly alter the PK properties of F. Based on its favorable safety profile and its optimal PK and pharmacodynamic properties, the maximal feasible dose of 20 mg/kg has been selected for further investigation.
CTCs were detected in approximately 40% of the patients, but the number of circulating cells was too low for systematic analysis. CTC counts have been proposed as a novel and noninvasive approach for pharmacodynamic studies in NSCLC and in breast and prostate cancer.5,8,15 Further research would be required to improve the sensitivity of current methodologies in NSCLC. Intriguingly, the median number of CECs increased during the study in a manner independent to F treatment. This finding may reflect a chemotherapy effect or patient selection based on prognosis, with those patients with a higher number of CECs receiving a larger number of treatment cycles. In that respect, it has been previously reported that baseline CEC counts are associated with better patient response and outcome in NSCLC.16 No relationship between CEC counts and response was apparent in this study. Further research should be undertaken to determine the prognostic value of CECs in the treatment of NSCLC.
Finally, we found that IGF-1 bioactivity may be critical for NSCLC response to chemotherapy and that blockade of the IGF-1/IGF-1R interaction by F may translate to clinical benefit, particularly in settings of high IGF-1 bioactivity. These findings should be considered with caution due to the small data set investigated in this study. However, these results are consistent with previous reports associating high-IGFBP-3 levels with longer overall, disease-free, and event-free survival in NSCLC.17,18 Additional research will be needed to determine the predictive value of fIGF-1 and IGFBP-3 levels and their potential use for patient selection or stratification in studies of anti-IGF-1R therapy. Pharmacodynamic measures are currently underway in ongoing studies to validate these provocative observations.19,20
Supported by Pfizer Inc.
Disclosure: A.G.- is employed with Pfizer, Inc; L.G.P-A-was paid an honorarium to serve on the Advisory Board panel for Pfizer; F.C. – was paid for clinical trials during the last 3 years; P.E. – practice was paid for entering patients into the trial; K.L. – received a research grant from Pfizer; D.Y. and spouse are employees of Pfizer. D.K. – served as a non-paid consultant and received payment for travel expenses only; R.B.C. – served on the Advisory panel for a diabetes drug and his institution was paid a grant to cover the cost of performing the study; the other authors declare no conflicts of interest.