FDA Panel: All Thumbs Down for Antibody to Treat Metastatic Neuroblastoma

An FDA advisory committee voted unanimously against recommending approval of the radiolabeled monoclonal antibody ¹³¹I-omburtamab for progressive neuroblastoma with central nervous system (CNS)/leptomeningeal metastases.

Citing perceived shortcomings in the pivotal trial’s design, the external control group, and evidence of clinical activity, the Oncologic Drugs Advisory Committee (ODAC) voted 16-0 with no abstentions against recommending approval of ¹³¹I-omburtamab.

“The panel does not feel that the applicant has met the criteria needed to prove an overall survival benefit,” said Christopher Lieu, MD, of the University of Colorado Cancer Center in Aurora, following the vote count. “The panel would like to see increased response, obviously, if possible, and an improved comparator, and potentially even more robust preclinical data.”

¹³¹I-omburtamab recognizes and binds selectively to the transmembrane protein B7-H3, which is highly expressed on the surface of neuroblastoma cells but only minimally expressed in normal tissues. The iodine-131 component of the construct damages B7-H3-expressing cells, leading to cell death. Specific targeting of B7-H3 facilitates irradiation of tumor cells with limited toxicity to adjacent normal tissues, according to a report submitted to ODAC by Y-mAbs Therapeutics.

Principal support for the approval application came from the single-arm 03-133 trial with an external control group constructed from patients in a German registry. A propensity-matched analysis showed a 3-year overall survival (OS) rate of 54% with ¹³¹I-omburtamab versus 31% in the control group, representing a 42% reduction in the survival hazard (95% CI 0.31-1.09, P=0.0544).

An FDA staff report to ODAC cited multiple perceived deficiencies with 03-133 and a 50-patient companion trial designed to assess response to ¹³¹I-omburtamab. The 03-133 trial started out as a 30-patient dose-finding study, but additional patients were enrolled in a dose-expansion phase and added to the original protocol, which had a primary endpoint of toxicity, with 3-year OS as a secondary endpoint.

The 101 companion study is ongoing and has a primary endpoint of 3-year OS.

The FDA report stated that the 03-133 patient population and the external comparator group were not comparable. The 109 patients in 03-133 started treatment in 2004, whereas patients selected for the control arm began treatment as early as 1990. Patients in 03-133 had progressed after multimodality treatment, including craniospinal irradiation (CSI). Only a minority of the control group had received multimodality treatment, and CSI is not used in Germany to treat neuroblastoma.

“The external control population is not fit-for-purpose as a comparator due to substantive differences between the study and control populations that limit the ability to attribute survival differences to the effect of ¹³¹I-omburtamab,” FDA staff members stated in the report.

In the interest of “regulatory flexibility,” the FDA performed additional analyses that reinforced “that differences in survival cannot be reliably attributed to ¹³¹I-omburtamab,” they added.

The FDA found fault with several aspects of the 101 trial and the resulting data supporting the monoclonal antibody’s activity. The concerns related to at least one patient with no measureable target lesion, limited washout periods between treatments, variable timing of therapies, and disagreements among radiology reviewers.

“Overall, no patient in study 101 demonstrated an unequivocal treatment response that could be definitively attributed to ¹³¹I-omburtamab,” the staff report continued.

Despite high hopes at the outset, the results were a letdown for some ODAC panelists.

“I very much wanted to believe the survival differences that were shown,” said Jorge Nieva, MD, of the USC Norris Comprehensive Cancer Center in Los Angeles. “But in looking at the adjustment for confounders, this really looks to me like a lot of selection bias. Unfortunately, we don’t have any data that isolates the treatment in the absence of a lot of other treatments. We don’t have good response data. I’m very much bothered by the fact that the best picture that we had showing a response was a picture that was confounded by intervening chemotherapy and all these [other treatments].”

Multiple panelists also credited Y-mAbs’ effort and expressed hope that the company would continue to investigate promising therapies for a difficult disease with limited treatment options.

“This is very hard to do, and I think this work highlights a lot of the difficulties that we face,” said E. Anders Kolb, MD, of Nemours Children’s Health in Wilmington, Delaware.

The FDA is not bound by advisory committee decisions but usually follows the panels’ recommendations.