Treatment with onartuzumab did not provide any additional clinical benefit when added to bevacizumab (Avastin) in the treatment of patients with recurrent glioblastoma multiforme (GBM), according to a study recently published in the <em>Journal of Clinical Oncology</em>.
Timothy Cloughesy, MD
Treatment with onartuzumab did not provide any additional clinical benefit when added to bevacizumab (Avastin) in the treatment of patients with recurrent glioblastoma multiforme (GBM), according to a study recently published in theJournal of Clinical Oncology.1
Although clinical improvement was not seen with the addition of onartuzumab in unselected patients, Timothy Cloughesy, MD, and colleagues noted in the study that patients with high hepatocyte growth factor (HGF) expression or unmethylatedMGMTmay benefit from treatment with the monovalent MET inhibitor. Further study was recommended to validate these prognostic markers.
The phase II GO27819 trial investigated the safety and efficacy of onartuzumab on top of bevacizumab (Avastin), which was approved by the FDA in May 2009 as a second-line treatment for patients with glioblastoma as a single agent. Exploratory biomarker analyses were also conducted to find an association between GBM subtype,HGFexpression, orMGMTmethylation and improved outcomes from the combination.
In the study, 129 patients were randomized 1:1 to either 15 mg/kg of onartuzumab once every 3 weeks plus 15 mg/kg of bevacizumab once every 3 weeks (n = 64) or bevacizumab with placebo (n = 65). Patients had recurrent disease after prior chemoradiation and had never received bevacizumab before.
An additional arm was originally included in the trial studying onartuzumab plus placebo, however, the cohort was placed on hold during a safety assessment.
Baseline patient characteristics were well balanced between the 2 arms of the study. The median age was 57 in the onartuzumab/bevacizumab arm and was 55 in the bevacizumab/placebo arm. More than half of the patients (56.6%) had a Karnofsky performance score (PS) of 70% or 80%, and the rest had a PS of 90% or 100%.
After a median follow-up of 9.8 months for the onartuzumab arm and 9.9 months for the bevacizumab plus placebo arm, 92% and 85% of patients had discontinued treatment, respectively. The median progression-free survival (PFS) in the intention-to-treat population was 3.9 months with onartuzumab/bevacizumab and 2.9 months with bevacizumab/placebo (HR, 1.06; 95% CI, 0.72-1.56;P= .7444). The 6-month PFS rate was 33.9% in the onartuzumab and bevacizumab arm versus 29% for the bevacizumab and placebo arm (P= .5555).
Median overall survival (OS) was 8.8 months versus 12.6 months for onartuzumab/bevacizumab and bevacizumab/placebo, respectively (HR, 1.45; 95% CI, 0.88-2.37;P= .1389). The objective response rate (ORR) was 22.2% in the onartuzumab arm, which included 1 complete response (CR), and the ORR in the bevacizumab and placebo arm was 23.7%, including 3 CRs.
General disorders of any grade occurred in 70.8% of patients in the onartuzumab/bevacizumab arm and 57.8% in the bevacizumab/placebo arm, including peripheral edema (44.6% vs 14.1%, respectively), asthenia (24.6% vs 21.9%), and fatigue (15.4% vs 20.3%). Nervous system disorders were also common, experienced by 49.2% of patients in the onartuzumab group and by 75% in the bevacizumab group.
Grade ≥3 adverse events (AEs) were noted in 25 patients (38.5%) in the onartuzumab/bevacizumab arm and by 23 patients (35.9%) in the bevacizumab/placebo arm. A greater proportion of peripheral edema and hypoalbuminemia were both noted in the group receiving onartuzumab.
Two deaths occurred in the onartuzumab/bevacizumab arm, both due to intestinal perforation, and 1 death occurred in the bevacizumab/placebo arm due to intracranial hemorrhage.
More patients had mesenchymal GBM (n = 56) than the proneural (n = 28) or proliferative (n = 20) subtype. No difference in PFS was noted with the mesenchymal subtype over the other subtypes, as had been seen in the AVAglio trial.2Retrospective analyses of the AVAglio trial, which explored bevacizumab plus radiotherapy and temozolomide in GBM, had demonstrated a greater benefit for patients with mesenchymal or proneural GBM when receiving the bevacizumab combination, although only the proneural subtype showed a benefit in OS.
The AVAglio study also showed an association between the mesenchymal phenotype and expression of MET ligand HGF, which suggested “that this may be a common occurrence in glioblastoma biology,” the study authors noted.
In the GO27819 trial, high expression ofHGF(upper 25% by polymerase chain reaction) seemed to be predictive of efficacy with the onartuzumab/bevacizumab combination in patients with the mesenchymal subtype. Patients with higherHGFlevels in the onartuzumab/bevacizumab arm had longer median PFS than patients in the bevacizumab/placebo arm of 6.1 months versus 2.8 months, respectively (HR, 0.37; 95% CI, 0.16-0.86;P= .0201). The opposite was seen in patients with lowerHGFexpression.
MGMTmethylation, a well-known prognostic marker in glioblastoma, was also explored among 47 patients with methylatedMGMTand 57 with unmethylatedMGMT.Patients with unmethylatedMGMThad better outcomes in the onartuzumab/bevacizumab arm, and multivariate analyses confirmed the potentially predictive effect of treatment with the combination for patients with unmethylatedMGMT.
MET status was intended to be a part of the analyses, yet only 5 patients were MET-positive according to a ≥50% staining cutoff by immunohistochemistry, which did not allow for a large enough population to study.
“Although the biomarker data are compelling and should inform future trials targeting the MET pathway, their exploratory nature and relatively small sample size make it difficult to infer firm conclusions on the clinical utility of HGF expression or MGMT methylation as predictive markers in glioblastoma,” the authors noted in the study. “Future prospective studies are necessary to validate these results.”