Key PointsQuestion
In patients with resistant neuroblastoma, will increasing the dose of hu3F8, a humanized anti-GD2 antibody, with granulocyte-macrophage colony-stimulating factor affect tumor response without excessive toxic effects?
Findings
In this phase 1 clinical trial of 57 patients with high-risk neuroblastoma, high doses of hu3F8 were well tolerated, without a need for inpatient monitoring and with a low incidence of neutralizing antibodies; major responses and prolonged progression-free survival occurred in the heavily previously treated study population.
Meaning
Achieving substantial antineoplastic results using a safe humanized antibody against a single carbohydrate antigen supports the further development of hu3F8, which is proceeding in a multi-institutional trial.
Importance
Chimeric and murine anti-GD2 antibodies are active against neuroblastoma, but the development of neutralizing antibodies can compromise efficacy. To decrease immunogenicity, hu3F8, a humanized anti-GD2 antibody, was constructed.
Objective
To find the maximum-tolerated dose of hu3F8 with granulocyte-macrophage colony-stimulating factor.
Design, Setting, and Participants
This phase 1 clinical trial used a 3 + 3 dose-escalation design in a single referral center (Memorial Sloan Kettering Cancer Center, New York, New York). Participants were enrolled from December 24, 2012, through May 3, 2016, with follow-up and analyses through February 28, 2018. Eligibility criteria included older than 1 year and resistant or recurrent neuroblastoma regardless of the number or kinds of prior treatments. All 57 participants met the eligibility criteria, received treatment according to the protocol, and were included in all analyses.
Interventions
Treatment cycles were monthly, if human antihuman antibody remained negative. Each cycle comprised hu3F8 infused intravenously for 30 minutes on Monday, Wednesday, and Friday as well as granulocyte-macrophage colony-stimulating factor administered subcutaneously daily from 5 days before infusion through the last day of infusion. After cycle 2, hu3F8 was increased to the highest dose level that had been confirmed as safe.
Main Outcomes and Measures
Toxicity, pharmacokinetics, immunogenicity, and disease response.
Results
Of the 57 participants, 34 (60%) were male and 23 (40%) were female (male-to-female ratio of 1.5), with a median (range) age of 6.8 (2.4-31.3) years at enrollment and a median (range) time of 3.1 (0.6-9.0) years since initial chemotherapy. Participants received a median (range) of 4 (1-15) cycles. Treatment was outpatient with reversible neuropathic pain and without unexpected toxic effects. No maximum-tolerated dose was identified. Dose escalation was associated with increased serum levels and proceeded through dosage of 9.6 mg/kg/cycle (approximately 288 mg/m2), which is more than 2.5 times higher than the standard dosage of 75 mg/m2/cycle or 100 mg/m2/cycle of dinutuximab and m3F8. Human antihuman antibody positivity developed in 5 of 57 patients (9%) after cycle 1, including in 1 of 10 patients (10%) not previously treated with anti-GD2 antibody and in 4 of 47 patients (9%) previously exposed to 1 or 2 anti-GD2 antibodies. Antineuroblastoma activity included major responses associated with higher dosing and prolonged progression-free survival despite a history of relapses.
Conclusions and Relevance
This phase 1 clinical trial found hu3F8 to be associated with modest toxic effects, low immunogenicity, and substantial antineuroblastoma activity; phase 2 trials are in progress.
Trial Registration
ClinicalTrials.gov identifier: NCT01757626
In patients with high-risk neuroblastoma, immunotherapy using the anti-GD2 chimeric monoclonal antibody (mAb) dinutuximab alone1 or with interleukin 2 and granulocyte-macrophage colony-stimulating factor (GM-CSF) is associated with improved outcome.2 The murine IgG3 anti-GD2 mAb 3F8 (m3F8) with GM-CSF is an effective consolidative therapy in patients with high-risk neuroblastoma in first or second (or later) remission3,4 and is active against primary refractory osteomedullary disease.5 With m3F8, however, early formation of human antimouse antibody could compromise efficacy by accelerating blood clearance and preventing further treatment. The removal of mouse epitopes should decrease human antimouse antibody response, but even chimeric mAbs can induce human antichimeric antibody.6 To circumvent this problem of sensitization, hu3F8, an IgG1 subclass humanized form of m3F8, was constructed.7
Preclinical studies7 revealed differences between hu3F8 and other anti-GD2 mAbs. Compared with dinutuximab, hu3F8 has 10 times higher affinity for GD2 ganglioside, which is a desirable property of therapeutic mAbs against carbohydrates.8 Compared with m3F8, hu3F8 exhibits superior antibody-dependent cellular cytotoxicity (ADCC) mediated by mononuclear cells and neutrophils, reduced but active complement-dependent cytotoxicity (CDC), and augmented efficacy in neuroblastoma xenograft models. This cytotoxicity profile made hu3F8 attractive for clinical use because enhanced ADCC is associated with better response, and complement activation is deemed responsible for the pain adverse effects that can limit dosing of anti-GD2 mAbs. Hence, we hypothesized that high dosing of hu3F8 would be tolerable in patients and would demonstrate major antineuroblastoma activity given the dose-response association between anti-GD2 mAbs and ADCC in vitro.9
Several observations supported using hu3F8 with GM-CSF. This cytokine is well tolerated clinically and enhances granulocyte-mediated ADCC of neuroblastoma, in which killing correlates with effector to target ratios and is antibody concentration–dependent.9 In high-risk neuroblastoma clinical trials, GM-CSF use has varied. With dinutuximab, GM-CSF has been administered at a dosage of 250 μg/m2/d, with the option of intravenous or subcutaneous administration.2,6,10 In contrast, with m3F8, GM-CSF was administered intravenously in initial trials and subcutaneously in later studies and always with a step-up from priming dosages of 250 μg/m2/d to 500 μg/m2/d during the period of m3F8 infusions.3-5 This increase in GM-CSF dose was based on the GM-CSF dose-response association with ADCC in vitro.9 Clinically, the subcutaneous route and the step-up in dose were associated with greater activation of granulocytes, which was an independent prognostic factor for superior progression-free survival.11 This activation of granulocytes provided a likely mechanism underlying the better outcome from subcutaneously administered GM-CSF compared with either intravenously administered or no GM-CSF.3-5,11 This article reports on the phase 1 clinical trial (NCT01757626) of hu3F8 and GM-CSF.
The primary objective of this phase 1 single-arm clinical trial was to find the maximum-tolerated dose of hu3F8 (prepared as described7) when used with GM-CSF (Leukine; Immunex Corp). Secondary objectives included performing pharmacokinetic studies and assessing antineuroblastoma activity. Participants were enrolled from December 24, 2012, through May 3, 2016. The trial protocol was approved by the Memorial Sloan Kettering Cancer Center Institutional Review Board and is available in Supplement 1. Informed written consent for this trial was obtained from patients according to the rules of the Memorial Sloan Kettering Cancer Center Institutional Review Board.
Eligibility criteria included the following: age older than 1 year; high-risk neuroblastoma, defined as MYCN [OMIM *164840]–amplified stages 2 to 4S at any age and MYCN-nonamplified stage 4 diagnosis after age 18 months; disease recurrent or refractory to standard high-risk neuroblastoma therapy with no limit to number or types of prior treatments; major organ dysfunction grade of 2 or lower, according to Common Terminology Criteria for Adverse Events, version 4.0; leukocyte count of 1000/μL or higher (to convert to ×109/L, multiply by 0.001), absolute neutrophil count of 500/μL or higher, absolute lymphocyte count of 500/μL or higher, and platelet count of 25 000/μL or higher (including with transfusion; to convert to ×109/L, multiply by 1.0); and no systemic therapy for 3 weeks or more before enrollment.
Clinical status at enrollment had to be any of the following: primary refractory disease, defined as incomplete response to high-risk neuroblastoma therapy but no prior relapse or progressive disease; secondary refractory disease, defined as incomplete response to salvage therapy for prior relapse or progressive disease; current progressive disease; or second or later complete remission, ie, no assessable neuroblastoma after salvage therapy for prior relapse or progressive disease. Previous treatment with anti-GD2 mAb was allowed if the patient’s human antihuman antibody (HAHA) titer was 1300 enzyme-linked immunosorbent assay (ELISA) U/mL or lower.
All 57 participants met eligibility criteria, received treatment according to the trial protocol at Memorial Sloan Kettering Cancer Center (New York, New York), and were included in all analyses; none were lost to follow-up (Figure 1). Participants were enrolled from December 24, 2012, through May 3, 2016, and all had chemotherapy-resistant stage 4 high-risk neuroblastoma, as evidenced by results of a radiologic study, a bone marrow histologic study, and/or at least 1 prior relapse.
Each cycle comprised (1) subcutaneously administered priming dosage of GM-CSF at 250 μg/m2/d on days −4 to 0 (Wednesday through Sunday), followed by a step-up to 500 μg/m2/d on days 1 to 5 (Monday through Friday) and (2) hu3F8 infused intravenously for 30 minutes on days 1, 3, and 5 (Monday, Wednesday, and Friday, ie, only 3 doses/cycle). The cytokine GM-CSF was not administered if the absolute neutrophil count was greater than 20 000/μL. Premedication included opiates and antihistamines.
Treatment was repeated 4 times monthly but was deferred if HAHA developed. Patients who completed 4 cycles without the development of progressive disease had the option to continue treatment at intervals of up to 8 weeks between cycles. After cycle 2, patients had the option with each cycle to increase the hu3F8 dose to the highest level at which patients had completed assessment for dose-limiting toxicity (DLT).
The 3 + 3 dose-escalation design for phase 1 trials was used.12 Thus, 3 to 6 patients were treated at each dose level, and the maximum-tolerated dose was the highest level at which 0 of 3 or 1 of 6 patients experienced DLT. Dose level 1 of hu3F8 was 0.9 mg/kg/cycle, which is approximately 27 mg/m2/cycle using the standard equivalency calculation (× 30) for converting dose by weight to body surface area.
Dose-limiting toxicity was defined as grade 3 toxicity or higher (Common Terminology Criteria for Adverse Events, version 4.0) occurring in cycle 1, days 0 to 21. All patients who received 1 dose of hu3F8 were evaluable for toxic effects. Dose-limiting toxicity included hypertension requiring medication for more than 24 hours; grade 3 or higher capillary-leak syndrome; and pain necessitating 7 doses or more of opioids within 2 hours, with 1 dose defined as 0.1 mg/kg of morphine sulfate. This DLT designation for pain was based on historical experience: wherein an unselected cohort of 50 patients who, in earlier studies, had received 435 administrations of m3F8 and GM-CSF, the proportion of patients treated with 7 doses or more of opioids in 24 hours was 0.06.13 Grade 3 toxic effects were not considered DLT if they resolved in fewer than 24 hours, which included urticaria (improved to grade ≤2), vasovagal reaction, sinus bradycardia or tachycardia, hypotension, fever, and electrolyte disturbances. Grade 3 toxic effects were not considered DLT if they resolved in fewer than 48 hours and included nausea/emesis, diarrhea, and paresthesia. Other grade 3 or 4 toxic effects were not considered DLT if they returned to baseline by day 21 of cycle 1 or were clearly associated with disease activity or cointerventions.
HAHA and Complement Measurements
Human antihuman antibody was quantified by an ELISA assay validated for its sensitivity, specificity, and interference in accordance with Investigational New Drug 12594. Serum samples were first bound to hu3F8-IgG1-F(ab’2) coated on microplates (ThermoFisher Scientific) and then to peroxidase-conjugated goat antihuman Fc-specific antibody. On the basis of a human high-titer serum reference standard, 42 normal controls were measured to establish an upper limit of normal as mean (±3 SDs) to maximize the positive or negative predictive value of the assay, which served as a cut point for HAHA positivity (1300 U/mL; 3.9 ng/mL).
Because hu3F8 mediates CDC in vitro,7 C3 complement and total hemolytic complement activity (CH50) were measured before and 3 hours after the first infusion of hu3F8 in cycle 1.
Serum samples were collected before and then after the first hu3F8 infusion on day 1 at 5 minutes, 3 hours, 6 hours, 24 hours, 48 hours (before and 5 minutes after the second hu3F8 dose on day 3), 72 hours, 96 hours (before and 5 minutes after the third hu3F8 dose on day 5), 120 hours, 168 hours, 216 hours, and 264 hours.14 Serum hu3F8 concentration was quantified using a validated ELISA in accordance with Investigational New Drug 12594. In brief, anti-m3F8 idiotype antibody A1G415 was used to capture serum hu3F8 (hu-IgG1 antibody) on microplates followed by peroxidase-conjugated mouse antihuman IgG1 Fc-specific antibody. Color reaction using hydrogen peroxide and o-phenylenediamine was measured with an ELISA plate reader at 490 nm. Serum hu3F8 was quantified by referencing to a hu3F8-IgG1 standard curve.
Pharmacokinetic analysis was performed by noncompartmental analysis of the serum concentration-time data using the Phoenix WinNonLin software, version 7 (Certara). On the basis of the dosing interval of 48 hours between doses of hu3F8, the key measurements included area under the serum concentration-time curves, measured using the log-linear trapezoidal calculation method for noncompartmental analysis suggested by the software; peak serum concentration (mean of 5 minutes after infusion on days 1, 3, and 5); trough serum concentration (mean of 5 minutes before infusion on days 3 and 5); terminal half-life; clearance and volume of distribution in steady state; and mean residence time, the theoretical mean time the drug remains in the body and calculated as mean residence time = volume of distribution/clearance.
Treatment Response Assessment
Disease status was assessed after cycles 2 and 4 (a standard for high-risk neuroblastoma) and then at least every 3 months by using computed tomography of the primary site, a 123I-metaiodobenzylguanidine scan, and a bone marrow histologic study (aspirates and biopsy specimen from bilateral posterior iliac crests and aspirates with or without biopsy specimens from bilateral anterior iliac crests). Response was defined according to the International Neuroblastoma Response Criteria,16 and 123I-metaiodobenzylguanidine scan findings were quantified by Curie scores17: complete remission, no evidence of neuroblastoma; partial response, 50% or greater decrease in neuroblastoma by radiologic study and bone marrow histologic study results showing complete remission or minimal disease (>0 to ≤5% tumor infiltration); minor response, 50% or greater decrease of any lesion with less than 50% decrease in any other lesions; stable disease, less than 50% decrease but 20% or less increase in any existing lesion; or progressive disease, new site of disease or greater than 20% increase in an existing lesion.
Comparison of responses was performed using the χ2 test for a significance level of 2-sided P = .05. Statistical analysis was performed with SPSS software (IBM).
Of the 57 participants, 34 (60%) were male and 23 (40%) were female (male-to-female ratio of 1.5), with a median (range) age of 6.8 (2.4-31.3) years at enrollment and a median (range) time of 3.1 (0.6-9.0) years since initial chemotherapy. Table 1 presents patient characteristics, and eTable 1 in Supplement 2 includes clinical features of patients treated at each dose level. All but 3 patients were previously treated with 1 or more anti-GD2 mAbs, received autologous stem cell transplant, and/or underwent 131I-metaiodobenzylguanidine therapy.
All 57 patients were evaluable for toxic effects. No maximum-tolerated dose was identified (eTables 1 and 2 in Supplement 2). Acute toxic effects were manageable, allowing outpatient treatment. No delayed or long-term toxic effects occurred. Patients received a median (range) of 4 (1-15) cycles of hu3F8 and GM-CSF. Eight patients (14%) received 10 cycles or more. Dosage escalation stopped at 9.6 mg/kg/cycle (dose level 15; approximately 288 mg/m2/cycle), based on the plateau of the area under the serum concentration-time curves in pharmacokinetic studies. No association between toxic effects and development of HAHA was found.
eTable 2 in Supplement 2 lists the toxic effects observed in 5% of patients or more in cycles 1 or 2. The common adverse effects, as expected with anti-GD2 mAbs, occurred on days of hu3F8 administration and included grades 1 to 2 pain, urticaria, and cough. Grade 3 toxic effects in major organs were limited to transient elevations in liver enzymes in cycle 1 in 10% of patients and in single patients in cycles 2, 6, 7, and 9. Five patients (9%) developed grades 1 to 2 Adie pupil (mydriasis, abnormal accommodation), including 3 patients after cycle 1, 1 patient after cycle 4, and 1 patient after cycle 10. No other neurologic deficits were noted.
Four patients (7%) experienced DLT (eTable 1 in Supplement 2). Grade 4 anaphylactic reactions (acute unresponsiveness and hypotension) accompanied the first infusion of hu3F8 in 1 patient who received 0.8 mg/kg (2.4 mg/kg/cycle, dose level 3; this patient had previously experienced an identical reaction to m3F8) and in 1 patient who received 2.8 mg/kg (8.4 mg/kg/cycle, dose level 13). One patient had acute, transient renal failure after unreported emesis during the weekend after the completion of cycle 1 (7.8 mg/kg/cycle, dose level 12). The fourth patient who experienced DLT had transient grade 3 or 4 hypertension, which developed after the completion of cycle 1 (9 mg/kg/cycle, dose level 14).
Beyond the period of monitoring for DLT (ie, cycle 1), 2 patients (4%) experienced grade 4 toxic effects that necessitated their exit from the study: 1 patient, who had uneventfully received 6 cycles, developed an anaphylactic reaction from the first infusion of hu3F8 (1 mg/kg) in cycle 7 (3 mg/kg/cycle, dose level 4); another patient developed grade 4 angioedema (tongue swelling, stridor) after completing cycle 2 (8.4 mg/kg/cycle, dose level 13).
Fifty-four of 57 patients (95%) had serum samples for pharmacokinetic analyses (2 patients experienced DLT after the initial infusion, and 1 patient had a missing peak serum concentration of the third dose). Dose escalation was associated with increased serum concentration of hu3F8. Analysis of pharmacokinetics in cycle 1 at each of the 15 dose levels (range, 0.9 mg/kg/cycle to 9.6 mg/kg/cycle) revealed a strong linear correlation between dose and peak serum concentration of hu3F8 (ie, at 5 minutes after infusion on days 1, 3, and 5) (R2 = 0.9629; Figure 2A). A linear correlation was also seen between dose and trough serum concentration (before infusion on days 3 and 5) (R2 = 0.8658; Figure 2B). Similarly, a statistically significant association was found between hu3F8 dose and area under the serum concentration-time curves (R2 = 0.9501; Figure 2C).
Dose escalation did not change the terminal half-life, with a mean (SD) of 4.08 (1.12) days (eFigure, A in Supplement 2), or the mean residence time (calculated as mean residence time = volume of distribution/clearance; eFigure, B in Supplement 2). This lack of association between dose and elimination of hu3F8 from the body can be attributed to the correlation between volume of distribution and clearance (eFigure, C in Supplement 2).
Human antihuman antibody was measured after every cycle. Human antihuman antibody positivity developed in 5 of 57 patients (9%) after cycle 1, including in 1 of 10 (10%) not previously treated with anti-GD2 mAb and in 4 of 47 (8.5%) previously exposed to 1 or 2 anti-GD2 mAbs. Among the 49 patients who received cycle 2, HAHA-positivity developed in 8 (16%), including in 2 of 10 patients (20%) not previously treated with anti-GD2 mAb and 6 of 39 (15%) previously treated with 1 or 2 anti-GD2 mAbs. Late HAHA positivity emerged in 2 patients after cycles 9 and 12. No substantial change in C3 complement or CH50 serum levels was observed after the infusion of hu3F8 on day 1 in cycle 1.
Antineuroblastoma activity (response, progression-free survival) at each dose level is included in eTable 1 in Supplement 2, and Table 2 summarizes results stratified by disease status and prior mAb treatment for the 31 patients with evaluable disease (Figure 1). Of the 31 patients, 14 (45%) had complete remission or partial response, 5 (16%) had stable disease (>5 months), 11 (35%) had early progressive disease, and 1 (3%) experienced DLT (with first infusion of hu3F8). Eight patients (26%) remained progression free at a median (range) of >37 (>22 to >54) months from the start of treatment, including 6 who received a vaccine18 for further consolidation of remission after completing this trial.
Among 24 patients who enrolled with refractory disease (excluding the patient with DLT during the first infusion of hu3F8), only 3 of the 10 patients (30%) treated with 4.8 mg/kg/cycle or less (approximately 144 mg/m2/cycle, dose level ≤7) had major responses (1 complete remission and 2 partial response), whereas 11 of 14 patients (79%) had major responses (5 complete remission and 6 partial response) with 5.4 mg/kg/cycle or higher (approximately 162 mg/m2/cycle, dose level ≥8), with an odds ratio of 2.6 (P = .02).
Among the 26 patients who enrolled in the second or later complete remission and no neuroblastoma evaluable for response (Table 1), 5 (19%) remained progression free at a median (range) of >50 (>26 to >63) months from the start of treatment, including 2 who were treated after protocol with the vaccine18 and 1 who received a cycle of low-dose maintenance chemotherapy and vaccine18 after exiting the study because of early HAHA positivity.
Notable findings regarding hu3F8 in this phase 1 trial included acceptable toxic effects without the need for inpatient monitoring and allowing timely dose escalation; low immunogenicity despite both previous treatment with anti-GD2 mAbs and repeated hu3F8 challenge; and major responses and prolonged progression-free survival in heavily previously treated patients. Pharmacokinetic findings supported using 3 doses of hu3F8 per cycle administered every other day compared with 4 or 5 doses with other anti-GD2 mAbs.1-3,19
Treatment was outpatient because the adverse effects were readily manageable, even with a dose of hu3F8 that was 2.5 times higher than the standard dosage of 75 mg/m2/cycle or 100 mg/m2/cycle of dinutuximab or m3F8. The explanation for the feasibility of such high dosing might be associated with the decreased complement activation observed with hu3F8 in vitro given that CDC is deemed partly responsible for the predominant toxic effects of anti-GD2 mAbs (ie, pain on days of treatment). The pain problem suggests that excessive CDC is not desirable. This concept prompted a study of hu14.18K322A, a humanized version of dinutuximab specifically constructed to reduce CDC and pain.19,20 Given the possible trade-off between manageable toxic effects and the well-established antineuroblastoma activity of CDC, it is reassuring that hu3F8 has less but still substantial CDC compared with m3F8.7 Other neuropathic adverse effects were limited to transient ophthalmic findings (Adie pupil), which have been noted with other anti-GD2 mAbs, including dinutuximab6,21 and hu14.18K322A.19 No transverse myelitis,22 motor neuropathy,2 capillary leak syndrome,2 or death occurred.
As previously reported, HAHA can substantially reduce serum levels of mAb.14 Fortunately, early HAHA positivity developed in only 9% of study patients after cycle 1 and 16% after cycle 2. Otherwise HAHA negativity continued despite repeated hu3F8 challenge, except in 2 patients who formed HAHA after cycle 9 and after cycle 12. The low rate of HAHA positivity with hu3F8 is especially remarkable given that prior exposure to a therapeutic mAb likely predisposes to sensitization and 47 of 57 study patients were previously treated with 1 or more anti-GD2 mAbs. This low immunogenicity compares favorably with both the 40% HAHA positivity reported for hu14.18K322A19 (humanized anti-GD2 mAb) and the 70% to 80% human antimouse antibody positivity with multiple cycles of m3F8.3-5 The HAHA positivity findings in this phase 1 trial approximate the sensitization of 8% to 28% of patients (not previously treated with anti-GD2 mAb) after 1 to 6 cycles of dinutuximab and cytokines.6,10,23-25 Clearly, the use of humanized or chimeric mAbs lessens but does not completely abrogate a humoral response that could hinder immunotherapy.14
As expected with phase 1 trials, the study patients all had a dismal prognosis given the extent of their chemoresistant disease and/or 1 or more prior episodes of progressive disease. These conditions with high-risk neuroblastoma are widely considered to be lethal.26 Yet patients with refractory disease achieved durable major responses and patients in second or later complete remission achieved prolonged progression-free survival (eTable 1 in Supplement 2). These welcome findings were noted in patients with or without prior exposure to 1 or more anti-GD2 mAbs. The substantially better response rate noted with higher hu3F8 dosages (≥162 mg/m2/cycle) is particularly relevant given that dinutuximab and m3F8 were tolerated only at a dosage of 100 mg/m2/cycle. The results in this phase 1 clinical trial and in the ongoing successor phase 2 clinical trial appear to have contributed to the US Food and Drug Administration granting, in August 2018, Breakthrough Therapy designation to hu3F8 in combination with GM-CSF for patients with persistent refractory neuroblastoma limited to bone marrow with or without evidence of concurrent bone involvement.
One limitation of this trial is the single referral center setting. The other limitation is the small number of patients at each dose level, which is a well-recognized drawback of a phase I trial.
To our knowledge, this phase 1 clinical trial with outpatient immunotherapy has exceeded expectations about dose and antineuroblastoma activity and met expectations about low immunogenicity. Further development of hu3F8 is proceeding apace, including an international pivotal phase 2 trial of hu3F8 and GM-CSF at multiple institutions in the United States and Europe, a phase 2 trial of hu3F8 and GM-CSF for first complete remission, and pilot studies of hu3F8 and natural killer cells as well as chemotherapy with hu3F8 and GM-CSF.
Accepted for Publication: July 5, 2018.
Corresponding Author: Brian H. Kushner, MD, Department of Pediatrics, Memorial Sloan Kettering Cancer Center, 1275 York Ave, New York, NY 10065 (kushnerb@mskcc.org).
Published Online: September 20, 2018. doi:10.1001/jamaoncol.2018.4005
Author Contributions: Dr Kushner had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Kushner, Modak, Basu, Roberts, N.-K. Cheung.
Acquisition, analysis, or interpretation of data: Kushner, I. Y. Cheung, Modak, Roberts, N.-K. Cheung.
Drafting of the manuscript: Kushner, I. Y. Cheung, N.-K. Cheung.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Kushner, I. Y. Cheung, Modak, N.-K. Cheung.
Obtained funding: Kushner, Modak, N.-K. Cheung.
Administrative, technical, or material support: Kushner, I. Y. Cheung, Basu, N.-K. Cheung.
Supervision: Kushner, I. Cheung, Basu, N.-K. Cheung.
Conflict of Interest Disclosures: hu3F8 Has been licensed by Memorial Sloan Kettering Cancer Center to Y-mAbs Therapeutics Inc, a company in which Memorial Sloan Kettering Cancer Center and Dr N.-K. Cheung have a financial interest. Dr Modak reported being a consultant to Y-mAbs Therapeutics Inc. No other disclosures were reported.
Funding/Support: This study was supported in part by the Band of Parents Foundation, the Robert Steel Foundation, Katie’s Find A Cure Fund, the Rhyan Loos Family Fund, the Arnold J. Jacobs Pediatric Cancer Fund, and Cancer Center Support grant P30 CA008748 from the National Cancer Institute.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
3.Cheung
NK, Cheung
IY, Kushner
BH,
et al. Murine anti-GD2 monoclonal antibody 3F8 combined with granulocyte-macrophage colony-stimulating factor and 13-cis-retinoic acid in high-risk patients with stage 4 neuroblastoma in first remission.
J Clin Oncol. 2012;30(26):3264-3270. doi:
10.1200/JCO.2011.41.3807PubMedGoogle ScholarCrossref 4.Kushner
BH, Ostrovnaya
I, Cheung
IY,
et al. Prolonged progression-free survival after consolidating second or later remissions of neuroblastoma with anti-G
D2 immunotherapy and isotretinoin: a prospective Phase II study.
Oncoimmunology. 2015;4(7):e1016704. doi:
10.1080/2162402X.2015.1016704PubMedGoogle ScholarCrossref 5.Cheung
NK, Cheung
IY, Kramer
K,
et al. Key role for myeloid cells: phase II results of anti-G(D2) antibody 3F8 plus granulocyte-macrophage colony-stimulating factor for chemoresistant osteomedullary neuroblastoma.
Int J Cancer. 2014;135(9):2199-2205. doi:
10.1002/ijc.28851PubMedGoogle ScholarCrossref 6.Ozkaynak
MF, Sondel
PM, Krailo
MD,
et al. Phase I study of chimeric human/murine anti-ganglioside G(D2) monoclonal antibody (ch14.18) with granulocyte-macrophage colony-stimulating factor in children with neuroblastoma immediately after hematopoietic stem-cell transplantation: a Children’s Cancer Group Study.
J Clin Oncol. 2000;18(24):4077-4085. doi:
10.1200/JCO.2000.18.24.4077PubMedGoogle ScholarCrossref 7.Cheung
NK, Guo
H, Hu
J, Tassev
DV, Cheung
IY. Humanizing murine IgG3 anti-GD2 antibody m3F8 substantially improves antibody-dependent cell-mediated cytotoxicity while retaining targeting in vivo.
Oncoimmunology. 2012;1(4):477-486. doi:
10.4161/onci.19864PubMedGoogle ScholarCrossref 9.Kushner
BH, Cheung
NK. GM-CSF enhances 3F8 monoclonal antibody-dependent cellular cytotoxicity against human melanoma and neuroblastoma.
Blood. 1989;73(7):1936-1941.
PubMedGoogle Scholar 11.Cheung
IY, Hsu
K, Cheung
NK. Activation of peripheral-blood granulocytes is strongly correlated with patient outcome after immunotherapy with anti-GD2 monoclonal antibody and granulocyte-macrophage colony-stimulating factor.
J Clin Oncol. 2012;30(4):426-432. doi:
10.1200/JCO.2011.37.6236PubMedGoogle ScholarCrossref 14.Cheung
IY, Kushner
BH, Modak
S, Basu
EM, Roberts
SS, Cheung
NV. Phase I trial of anti-GD2 monoclonal antibody hu3F8 plus GM-CSF: impact of body weight, immunogenicity and anti-GD2 response on pharmacokinetics and survival.
Oncoimmunology. 2017;6(11):e1358331. doi:
10.1080/2162402X.2017.1358331PubMedGoogle ScholarCrossref 16.Park
JR, Bagatell
R, Cohn
SL,
et al. Revisions to the International Neuroblastoma Response Criteria: a consensus statement from the National Cancer Institute Clinical Trials Planning Meeting.
J Clin Oncol. 2017;35(22):2580-2587. doi:
10.1200/JCO.2016.72.0177PubMedGoogle ScholarCrossref 17.Matthay
KK, Shulkin
B, Ladenstein
R,
et al. Criteria for evaluation of disease extent by (123)I-metaiodobenzylguanidine scans in neuroblastoma: a report for the International Neuroblastoma Risk Group (INRG) Task Force.
Br J Cancer. 2010;102(9):1319-1326. doi:
10.1038/sj.bjc.6605621PubMedGoogle ScholarCrossref 18.Kushner
BH, Cheung
IY, Modak
S, Kramer
K, Ragupathi
G, Cheung
NK. Phase I trial of a bivalent gangliosides vaccine in combination with β-glucan for high-risk neuroblastoma in second or later remission.
Clin Cancer Res. 2014;20(5):1375-1382. doi:
10.1158/1078-0432.CCR-13-1012PubMedGoogle ScholarCrossref 19.Navid
F, Sondel
PM, Barfield
R,
et al. Phase I trial of a novel anti-GD2 monoclonal antibody, Hu14.18K322A, designed to decrease toxicity in children with refractory or recurrent neuroblastoma.
J Clin Oncol. 2014;32(14):1445-1452. doi:
10.1200/JCO.2013.50.4423PubMedGoogle ScholarCrossref 22.Ding
YY, Panzer
J, Maris
JM, Castañeda
A, Gomez-Chiari
M, Mora
J. Transverse myelitis as an unexpected complication following treatment with dinutuximab in pediatric patients with high-risk neuroblastoma: a case series.
Pediatr Blood Cancer. 2018;65(1):e26732. doi:
10.1002/pbc.26732PubMedGoogle ScholarCrossref 23.Gilman
AL, Ozkaynak
MF, Matthay
KK,
et al. Phase I study of ch14.18 with granulocyte-macrophage colony-stimulating factor and interleukin-2 in children with neuroblastoma after autologous bone marrow transplantation or stem-cell rescue: a report from the Children’s Oncology Group.
J Clin Oncol. 2009;27(1):85-91. doi:
10.1200/JCO.2006.10.3564PubMedGoogle ScholarCrossref 24.Marachelian
A, Desai
A, Balis
F,
et al. Comparative pharmacokinetics, safety, and tolerability of two sources of ch14.18 in pediatric patients with high-risk neuroblastoma following myeloablative therapy.
Cancer Chemother Pharmacol. 2016;77(2):405-412. doi:
10.1007/s00280-015-2955-9PubMedGoogle ScholarCrossref