NEW YORK (Genomeweb) — A study comprehensively analyzing the genomes of more than 1,500 neuroblastoma cases has better determined the prognostic significance of ALK mutations and other alterations in neuroblastoma, and has biochemically investigated many of these variants to determine which ones confer sensitivity to the ALK inhibitor crizotinib, Pfizer's Xalkori, and which do not.
This knowledge will hopefully be harnessed to direct treatment to children with sensitive mutations in a planned Phase III trial of crizotinib in newly diagnosed neuroblastoma cases, intended to begin in 2016, Yael Mosse, one of the study's lead co-authors and a pediatric oncologist at The Children's Hospital of Philadelphia, told PGx Reporter this week.
Mosse, who spearheaded the earlier trials of crizotinib in neuroblastoma — a completed Phase I and an ongoing Phase II — said that the drug clearly helps some patients, and being able to better identify those who are likely to benefit will help solidify the effort to demonstrate efficacy in this subset in Phase III.
"What we found with this study was that 14 percent of kids who have aggressive form of neuroblastoma have a mutation or some abnormality in ALK that really turns it on and drives progression of the disease," Mosse explained. "This gives us the opportunity to really try to see if we can improve outcomes using crizotinib."
"This work also allows us now to understand which patients with which mutation will benefit and which won't, as well as developing a computational algorithm to predict with a new mutation we've not seen before whether that is likely to be activating or not," she added.
The study, published Monday in Cancer Cell, sequenced germline and somatic ALK exons in 1,596 newly diagnosed neuroblastomas. The analysis found that ALK mutations, which were identified several years ago as a promising potential drug target in these childhood cancers, occurred in about 8 percent, or 126 of the study cohort's samples.
Alterations occurred over three significant hotspots, which contained about 85 percent of the identified mutations, as well as 13 more minor sites, the authors reported.
The group also evaluated copy number variations, which occurred in 17 percent of a subset of cases for which copy number was measured. Gain of between three and 10 copies occurred in 195 cases. Higher amplifications and full deletion of the gene were both more rare, occurring in 24 and six samples, respectively.
Analysis of the correlation between ALK status and outcomes revealed that the presence of an ALK mutation in the cohort was associated with a 1.4-fold greater risk of disease events, and lower event-free survival and overall survival.
But the researchers did not stop at detection and cataloguing of these alterations. They also engaged in a complementary effort to biochemically and computationally study the identified mutations to better understand which are actually oncogenic drivers of the disease and which are more likely to be incidental.
Mark Lemmon, chair of biochemistry and biophysics at the University of Pennsylvania's Perelman School of Medicine and a co-lead author of the study, told PGx Reporter that this type of biochemical and functional analysis may need to become a wider player in the application of genomic findings to clinical practice.
"It has gotten to the point where you see a mutation and that's a box that's checked.
But of course, the mutations are changing the biochemistry of the system, so to fully appreciate what there is in that information you have to understand the biochemistry," he said.
This holds true, he added, not just for ALK, but also for other mutations that have become widely targeted in cancer therapy.
"For example, [targeting] the EGF receptor in non-small cell lung cancer has become almost standard of care in some settings. But if you go onto databases like COSMIC and others, there are tens of thousands of mutations now because so many patients have been sequenced," Lemmon said.
"In the EGFR kinase domain, 80 percent of sites have a mutation in at least one patient, and there are certainly hotspots where we know from the literature that they are activating and those patients should get the drugs, but there are an awful lot of people now who have mutations that we just don't understand."
Clinically, he said, this leaves two choices. "Either you say we don't understand these mutations so let's not treat, in which case you might be missing an opportunity. Or you can say, if you have any EGFR mutation we'll treat you, in which case you may be giving some people the wrong treatment."
In the neuroblastoma study with Mosse's team, Lemmon and his colleagues found that about 11 of 24 mutations, representing about nine percent of the patients in the study, appeared to be silent or even ALK inactivating based on biochemical analyses.
Of the activating mutations, in vitro analyses suggested that only six were sensitive to inhibition by crizotinib. Some others, though, might be sensitive to other ALK inhibitors, the authors suggested.
In addition to exploring the biochemistry of identified mutations, Lemmon and his group also developed a computational approach to predict whether ALK mutations were likely to be activating or not, which they showed was more successful than other currently used algorithms.
According to Lemmon, the team sees future possibilities for tools like this in clinical practice, especially as more and more novel mutations in targetable genes like ALK and EGFR are identified.
"The computation modeling is set up in principle so that it could be set up on an iPad or something like that. What we'd love is to develop this so that a physician could plug in a new mutation and get an answer based on our kinds of considerations," he said.
"There are potentially thousands of patients where the biochemistry really would make the decision about treatment."