Researchers from Italy's University of Genoa and other institutions have produced unpublished early data suggesting that digital PCR assays run on Life Technologies' recently launched QuantStudio 3D platform may be useful in measuring HER2 copy number in borderline-amplified breast cancer research samples.
In particular, the digital PCR assays showed good concordance with both immuno-based techniques, such as fluorescence in situ hybridization, and molecular methods like real-time quantitative PCR, quantitative reverse-transcriptase PCR, and array CGH, and may be an ideal method to quantify and resolve HER2 expression in difficult-to-interpret cases, Gabriele Zoppoli, lead author on the study, told PCR Insider last week.
Encouraged by these early results, the group plans to compare the ability of digital PCR with the other techniques in a much larger cohort in the hopes that the digital PCR assays will prove to be as accurate while remaining more reproducible, cheaper, and faster than its counterparts, Zoppoli said.
"So far [we cannot] say that digital PCR works better — it works differently," Zoppoli said. "What we observed is a very accurate way of measuring difficult samples that correlates with both traditional and non-traditional methods, which are supposed to be highly unbiased."
Zoppoli presented his group's work with the QuantStudio 3D — the first set of data from the platform to be presented by an early access user — this past weekend at a user meeting sponsored by Life Technologies as part of this year's American Association for Cancer Research meeting in Washington, DC. PCR Insider did not attend the user meeting but spoke with Zoppoli beforehand.
Life Tech in November launched the QuantStudio 3D — a dedicated, chip-based digital PCR platform that costs about $30,000 — with the hope of driving more widespread adoption of digital PCR for life science research (PCR Insider, 11/8/2012).
The platform, which was set to begin shipping commercially in the first half of this year, will be able to generate up to 20,000 data points from a given sample in a single run on its first generation of chips, according to the company.
Zoppoli said that he decided to test out the QuantStudio 3D because it offered the promise of providing "absolute" and reproducible quantification of HER2 copy number variation — an area where established methods can fall short.
"HER2 amplification status is really important for targeted therapy," Zoppoli said. "Those patients who have amplified HER2 will undergo targeted treatment, whereas those that don't, won't, so it is very important to have precise and reproducible results."
The current consensus in the clinical community is that FISH serve as the second-tier test after immunohistochemistry. And even though FISH can be extremely accurate "it still has gray areas in which you either have heterogeneity in your tumor sample, or the material is a little bit degraded, so you don't see it very well; or you have a very small sample, like fine-needle aspirate biopsy for neoadjuvant treatment," Zoppoli said.
It is in these cases — so-called "equivocal statuses" — in which a clinician might want to turn to molecular methods such as qPCR. However, another technical issue with qPCR is its reliance on normalized reference genes, "so you want to go for an absolute quantification," Zoppoli said. "This is the case for digital PCR, for example, and is the case for FISH, as well, but l don't think that counting 40 or 100 nuclei is representative of a population … you can just get unlucky and select the wrong place ... whereas array CGH or digital PCR actually go across the whole sample."
In order to assess whether and how well digital PCR could sub-classify patient samples that fell into gray areas using other methods, the group chose "a very tricky set of breast cancer samples ... which were or were not HER2 amplified ... and we ran array CGH, FISH, and digital PCR," Zoppoli said. "We also did gene expression [using qPCR] for HER2, which is a nice proxy for copy number, because HER2 status really drives its own transcript, whereas some other genes won't."
Zoppoli said that data from the experiment were not convincing enough to be able to conclude that digital PCR alone could "rescue" samples that were considered HER2 negative and recategorize them as HER2-positive.
However, the group observed that, "in general, we have a positive correlation with single-color FISH, which is to be expected." Digital PCR also correlated with array CGH, "which was nice because the unbiased methods [of] digital PCR and array CGH were more correlated to each other than single-color FISH."
Zoppoli stressed that the group's sample size was small, and their study did not approach that of a comparison to a gold standard. However, the results were promising enough to further explore digital PCR as a tool to determine HER2 amplification status.
"As a physician I'm reluctant to abandon traditional methods, because they have worked so far," Zoppoli said. "What I want is to integrate what can be done with FISH, which is pretty much the gold standard, with methodologies that are less operator-dependent, more reproducible, and cheaper … and this looks like ... a promising technique to do that," he added.
The next step for Zoppoli and colleagues is to compare digital PCR with the gold standard of FISH, as well as other molecular methods, in a much larger study. The researchers have already laid the groundwork for such an undertaking: At AACR, they presented a poster demonstrating the general agreement of immunohistochemistry, FISH, real-time qPCR on DNA, and qRT-PCR on RNA in assessing HER2/Neu status in a single-center, retrospective study of 130 breast cancer patients.
Zoppoli said that he believes this study is largest comparison to date of various methods to assess HER2 amplification status, and the group hopes to soon publish its data.
In the meantime, "it would be really interesting to do the next step, which is to compare digital PCR to the other methods," he said. "My wish really is to expand this data set to include a broader spectrum of HER2 cancer samples, and by then, if we adopt a centralized in situ hybridization analysis and proper techniques, we could really go for a comparison."