A French team is exploring the feasibility of routinely using next-generation sequencing, together with array comparative genomic hybridization, to molecularly screen tumor biopsy samples from individuals with advanced cancers. The group's goal is to place those patients into phase I trials of treatments targeting the alterations identified in their tumors.
Institut Gustave Roussy researcher Antoine Hollebecque provided an overview of the triage trial — known as molecular screening for cancer treatment optimization, or MOSCATO 01 — at the American Society of Clinical Oncology's annual meeting in Chicago this past weekend. During the presentation, Hollebecque also touched on interim findings from the trial, which kicked off in 2011.
The ongoing MOSCATO effort hinges on the use of "on purpose" biopsies to nab bits of tissue from primary or metastatic tumors in individuals with advanced solid tumors, Hollebecque said, particularly those who have experienced cancer progression after receiving multiple standard cancer treatments.
Depending on the proportion of tumor cells present in a given sample, members of the team are subjecting DNA in the tumor biopsy samples to one or more forms of molecular profiling.
In situations where 30 percent or more of the sample is comprised of tumor cells, the group does targeted sequencing using Life Technologies' Ion AmpliSeq hotspot cancer panel.
Analysis on a given sample is expanded to include array CGH analysis — a technique for tracking down chromosomal alterations — when tumor cells make up at least half of those in the sample.
At weekly tumor board discussions, Hollebecque explained, a group of clinicians, biologists, and bioinformaticists review the genomic alterations identified in each individual's tumor to determine whether they include specific mutations that might be susceptible to existing or experimental treatments.
From there, MOSCATO investigators track progression-free survival patterns in those who do or do not receive a targeted treatment — for instance, through a phase I trial —selected based on the genomic profiles of their tumors.
For each individual receiving targeted therapy, progression-free survival can be compared with time to progression on the most recent, non-targeted treatment that the patient received.
The motivation for the trial was a desire to provide a "molecular triage" for phase I patients, MOSCATO 01 Principal Investigator Jean-Charles Soria, with Institut Gustave Roussy, told Clinical Sequencing News in an email message.
The effort is intended to test the notion that genomics-based targeted treatment will provide clinical benefit to at least one-quarter of those enrolled — an improvement compared to the dismal outcomes currently seen in the advanced cancer patient group being considered.
In the 129 cancer patients enrolled in MOSCATO to date, researchers have gotten access to biopsy material for 112 of the individuals. All of the individuals participating in the study had experienced cancer progression following treatment with one or more existing first-line therapies.
The team has tested samples from 98 biopsied individuals by next-generation sequencing using Life Technologies' Ion AmpliSeq cancer hotspot mutation panel, Hollebecque said.
For 94 cases, the researchers have also looked for chromosome-level alterations by aCGH using an Agilent array. Molecular analyses for the study are all done in house at Institut Gustave Roussy.
Actionable mutations have been found in 53 individuals so far — representing just under half of those advanced cancer patients whose tumors were biopsied.
Almost three-quarters of the apparently actionable alterations turned up in samples tested by aCGH, Hollebecque noted. On the other hand, researchers found around 28 percent of the most intriguing genetic glitches through targeted gene sequencing alone.
The median time between biopsy and the identification of molecular targets in the study is currently 21 days.
Among the most commonly found mutations being uncovered are alterations affecting genes that code for the fibroblast growth factor and its receptor. Mutations in EGFR, HER2, and members of the cyclin and PTEN/PI3K/AKT/mTOR pathways are relatively common, too.
In 33 of the 129 cases, researchers have been able to match patients with targeted treatments based on mutation patterns in biopsied tumor samples. For a few of the cases, that has involved off-label use of existing treatments indicated for other cancer types, Soria noted, though most patients with targetable mutations are being enrolled in one of the 57 phase I trials currently underway at the Institut Gustave Roussy.
In most cases, it has been fairly straightforward to match patients with clinically relevant mutations to appropriate phase I trials of targeted therapies, Soria said, though some of the structural DNA changes that have been found are not yet actionable.
It can also be tricky to select the most appropriate treatment for individuals whose tumors contain many apparently actionable mutations. In the MOSCATO study, for instance, the researchers have already encountered patients with as many as five relevant targets.
The team's preliminary follow-up on the first 25 individuals receiving genomics-guided treatment revealed that five had had at least a partial response to treatment. Another 14 had stable disease and three of the 25 patients had progressive disease.
From their findings so far, Hollebecque, Soria, and their colleagues concluded that it's possible to incorporate high-throughput molecular analysis into clinical care, particularly to enrich for individuals with appropriate genomic alterations in phase I trials of treatments meant to target such changes.
Whereas phase I trials with all-comers typically have partial response rates of between 7 and 10 percent, the team noted in an abstract accompanying the MOSCATO presentation at ASCO, the partial response rate in phase I trials stemming from the more molecularly targeted trial enrollment is currently closer to 20 percent.
The group plans to assess at least 900 individuals with advanced cancer for the ongoing effort, which is scheduled to wrap up in the fall of 2015. A unique pediatric cohort will also be tested, Soria noted.
Going forward, members of the MOSCATO team are interested in finding ways to generate deeper sequence data in order to assess copy number patterns in patient tumors. They may also take a crack at doing whole-exome sequencing of patient tumor samples, Soria said, at least in the research setting.
He noted that the second generation of personalized trials in France will involve broader studies done at the national level, similar to efforts by those involved in a French study known as SAFIR.
At this week's ASCO meeting in Chicago, investigators from the Institut Gustave Roussy and several other centers in France, presented information on the first phase of the SAFIR study, known as SAFIR01.
That prospective trial is designed to address the potential of using molecular profiling to funnel individuals with metastatic breast cancer into appropriate phase I and phase II trials of targeted treatments. There, too, the study design hinges on finding genetic glitches in advanced tumor samples tested by biopsy.
Researchers have already attempted to test samples from more than 400 individuals with metastatic breast cancer for the SAFIR01 effort, which couples aCGH analysis of metastatic tumor biopsy samples with Sanger sequencing of stretches of the PIK3CA and AKT genes.
Findings presented by the SAFIR01 researchers at ASCO suggested that molecular testing on metastatic tissue is a feasible means of improving targeted treatment rates, though low tumor cell representation precluded genomic testing on roughly one-quarter of metastatic biopsy samples they hoped to test.
Data collection for the first stage of SAFIR is set to wrap up next spring. Members of the group plan to embark on a related study — a randomized trial known as SAFIR02 — sometime this summer.
Members of the MOSCATO and SAFIR studies are not the only ones looking at ways of using next-generation sequencing and other genomic profiling methods to more routinely guide treatment for cancer patients.
At another ASCO session, for example, representatives from several research teams outlined schemes for using molecular profiling methods — including sequencing, in some cases — to place individuals with glioma into appropriate targeted treatment trials.
And in the past few years, several centers in the US have made forays into the clinical sequencing realm for patients with rare diseases, advanced cancer, and/or undiagnosed congenital disorders (see, for example, CSN 5/29/2013, CSN 3/6/2013, CSN 1/2/2013, CSN 10/3/2012, CSN 3/7/2012, and CSN 2/29/2012).