There have been some exciting developments in work to expand the number of markers that can be measured in tissue. The constraints are cost and regulatory acceptance. Multiplexing of proteins in soluble samples (e.g. serum, plasma, urine) have advanced to true high-throughput, with arrays of ELISA-related technologies. However, tissue multiplexing remains difficult.
Immunofluorescence can easily do 3-4 antibodies on a tissue, and quantum dot approaches promise more, but quantum dots are unlikely to be run in the clinic for a number of reasons. Immunofluorescence in general suffers from two disadvantages — regulatory and tissue context. There are a total of 16 IHC based FDA clearances for protein expression with digital imaging using brightfield, and none using fluorescence (FISH doesn’t count). In companion diagnostics development, pharma cannot afford the risk to be first, waiting on clearance and acceptance for a novel technology. There are enough difficulties and risk along the long road of therapeutic development, adding in an unproven regulatory path on a companion diagnostics won’t help. The other difficult with immunofluorescence is the lack of the tissue context. Despite claims to the contrary (watch for these in presentations!), there are no antibodies that will perfectly bind to tumor with no other unwanted binding effect. If you happen to have one, immediately start your own biotech company, you have the holy grail of tumor targets in your hand. Thus any approach that looks to mask tumor with an antibody will require QA by a technician and then a pathologist. We do this regularly at Flagship in our digital pathology services. In brightfield with whole slide images this pathology review of image analysis is fast and cost-effective.
There are at least three emerging technologies to keep an eye on in tissue multiplexing.
20/20 Gene Systems in Maryland have a novel approach called L-IHC (layered IHC), where they sequentially pull off an antibody onto a absorbed layer, with a DNA probe that can hybridize to a fluorochrome.
GE has an approach that was presented at Pathology Informatics where they sequentially bind a fluorochrome-labeled antigen, image, remove, and bind the next one, image, remove, all on the same slide.
And finally, at Flagship we have just launched a new technology called FACTS (Feature Analysis on Consecutive Tissue Sections), where we utilize thin histology sections with image registration techniques from radiology to multiplex from 4 to 8 tissue biomarkers. The new techniques are exciting, with the end goal being an expansion of usage of tissue-based IVDMIAs.
