NEW YORK (GenomeWeb) – While widespread clinical implementation likely remains some ways off, MALDI imaging technology and workflows have advanced to the point where they can feasibly be used for pathology applications, suggested a leading researcher in the field.
Until recently, "the pieces weren't all in place on the mass spec side," said Richard Drake, director of the Medical University of South Carolina Proteomics Center, where he has been working to apply imaging mass spec to glycoprotein cancer biomarkers.
Today, though, researchers have relatively complete sets of mass spec tools and workflows they can offer pathologists, Drake said. "So [pathologists] don't have to worry about developing [the technology."
Shannon Cornett, mass spectrometry applications development manager at Bruker, which has collaborated with Drake on his MALDI imaging work, noted that the company has begun working with commercial and academic pathology labs in Europe with the aim of developing MALDI-based assays that could prove clinically useful.
"We are at the stage where those initial assays have been identified," Cornett said. "Those labs are under contractual agreements to work on those assays and through that process eventually [develop] something that is filed to regulatory agencies for approval."
Capable of providing large-scale spatial information on protein expression with cells and tissues, imaging mass spectrometry could potentially bring additional molecular information to traditional pathology while also reducing the subjectivity and intra-observer variability inherent in that process. This could allow for more accurate and reproducible versions of existing pathology assays as well as the development of more informative new assays.
Traditionally, imaging mass spec work has focused heavily on MALDI-based methods, and, as a major MALDI vendor, Bruker has been a leader in the field. At the American Society for Mass Spectrometry annual meeting in June, the company highlighted the progress of its MALDI imaging workflows with Drake presenting on glycan and glycoproteomic-based pathology work he has been doing in collaboration with the company. Speaking this week, Drake cited several advances that had progressed MALDI imaging technology to where it had become a plausible technology for pathologists.
One key has been the increased speed and resolution of MALDI instrument, he said, noting, for instance, that Bruker's rapiflex MALDI TOF/TOF enabled resolution of five microns and that other research groups were pushing towards resolutions in the one-micron range.
Additionally, improvement in software has played a key role in pushing the technology forward, Drake said.
"The evolution of software is really critical," he said. For instance, 'in the clinic, there has to be something that will facilitate looking at all this data, tools that will let the pathologist look at it as well as the mass spec person."
He cited Bruker's purchase in January of software company Scils, which offers a number of packages to support MALDI imaging work. "That, I think, is really going to pay off," he said. Bruker also offers MALDI-based histology analysis tools through its flexImaging software package.
Also significant has been advances in the sprayer technology used to treat samples with enzymes like trypsin used to prepare them for mass spec analysis. These devices have improved significantly in terms of throughput and reproducibility, Drake said, though he added that ongoing quality control would be needed as the field moved to analyzing hundreds and thousands of samples.
"So these things have evolved nicely, and they've kind of evolved together," Drake said. "And that's why I think it's time to say, okay, let's take these tools and apply them to larger clinical questions. I think the table is set."
What those clinical questions are remains to be seen, though Bruker is currently working with pathologists on several specific assays.
Drake's MALDI imaging work focuses primarily on glycoproteomics in prostate cancer. He and his colleagues are currently working on a study of some 500 prostate cancer samples, looking at how MALDI imaging profiles of different glycans are associated with things like tumor aggressiveness and the path of disease progression.
The majority of cancer protein biomarkers are glycoproteins, and "it has been known for decades that glycosylation changes on the cell surface promotes cancer progression and the way the immune systems sees a tumor or doesn't see a tumor," Drake said. However, much of the work exploring these relationships has used ground-up tumor tissue where any spatial information had been lost.
"That's the advantage of MALDI imaging," he said. "You're looking at the whole tissue micro-environment, and particularly for cancer it turns out to be important. Once we started seeing something, we were like, whoa. It was actually much more [informative] than we had expected. Though, like everything, it's brought up more questions than answers, probably."
"We thought we could make panels of glycans, pick out tumor glycans and glycans that weren't in the tumor, or maybe glycans that were immune-associated," Drake said. "It worked out that we could do that. Then, of course, [the data] starts getting parsed into different subsets of tumor glycans and different regions of tumor, and it's mushroomed into something that is much more complex than we realized. But now we have a good way of visualizing that, and we can start addressing that complexity."
Drake noted several possibilities for how MALDI imaging data could have clinical impact. In terms of cancer he suggested it might provide a link between genomic information and pathology, providing, for instance, data that recapitulates and refines at the level of pathology observed genomic subtypes.
"You've got all this genomic data, and samples get grouped out on that, but at the tissue level, they may not look any different," he said. "Even though there could be 10 different genes changed, these things just aren't obvious with a microscope. I think some of these features [observed by imaging mass spec]… will be molecular correlates that may bridge pathology with all the genomics out there."
He acknowledged that precise clinical applications remained somewhat vague, but added that "it's a good time to get these questions addressed, because the tools didn't exist until very recently. I think we're … starting to ask: How is this going work? How can this help?"
Getting pathologists involved at this stage is essential, Drake added, as their questions and expertise will be vital to assessing where the technology might be best deployed and how studies should be designed.
While he declined to specify a timeline for when he thought MALDI imaging would enter clinical pathology, he cited Bruker's MALDI Biotyper as an example of an optimistic scenario. That instrument, which has seen significant uptake within the clinical microbiology field, went from a research tool to a US Food and Drug Administration-cleared device in around 10 years.
"I don't know if that is achievable [for MALDI imaging], but there is the precedent if you get a specialized assay that has some promise," he said.
Cornett agreed that while the technology had made key advances, clinical implementation was still a ways off. He said that Bruker and its pathologist collaborators are currently focused on collecting samples and building databases for the assays they are investigating.
"We're in the early stages of database development for particular assays," he said. "Those have to be internally validated before we would go forward into a clinical trial, and then of course you've got the clinical trials, and all of that, and then regulatory submissions and probably a second-phase trial."
"I would estimate you could be thinking five to 10 years [to the clinic], easy," he said.