Medical Device Link.

Originally Published IVD Technology March 2003

In Person

IVD manufacturers develop and perfect assays, but fundamental advances in diagnostics begin in the research laboratory.

Herbert A. Fritsche, PhD, is professor and chief of clinical chemistry in the department of laboratory medicine at the University of Texas MD Anderson Cancer Center (Houston).  He can be contacted at hfritsche@mdanderson.org.

The influx of information about complex diseases, international health crises, and the mounting need for biodefense create an ever-increasing demand for new, more efficient, and less costly assays. Although IVD companies often make the final developmental contributions to assays before they are distributed to the public, they rely heavily upon academia and other outside researchers to discover and develop the fundamental technology behind the tests. 

To determine where assay development is headed and what the exchange between industry and academia can contribute to this process, IVD Technology editor Richard Park spoke with Herbert A. Fritsche, PhD, professor and chief of clinical chemistry in the department of laboratory medicine at the University of Texas MD Anderson Cancer Center (Houston). In this interview, Fritsche talks about the most effective advancements in assay development, tackling roadblocks, and working with industry.

What research projects related to assay development and diagnostics is your lab currently involved in?

We are involved in the validation of new cancer markers in preparation for their submission for the FDA clearing process. We try to prove the reliability of a particular assay method by establishing its reproducibility, long-term precision, method accuracy, and overall robustness.

Then, once we begin the clinical studies, we focus on validating the claims the manufacturer will make to FDA, whether those are claims related to early detection, utility as a diagnostic aid, or monitoring of patients. We help the manufacturer design the appropriate claims, and then we validate those claims.

We are also attempting to discover new markers by utilizing proteomics to identify proteins associated with the disease process and distinguish which proteins may have utility as markers for early detection. Similarly, we are trying to identify proteins that are reflective of a patient’s prognosis or of their response to treatment.

How have efforts to develop tests for cancer diagnosis evolved over the past few years? Have technological advancements helped in your research efforts?

The advancements made in the area of genomic and proteomic analysis have contributed greatly to identifying proteins that might be used as early detection markers. So we have many more leads to focus our work on now than we did in the past.

Certainly the DNA hybridization chip array approach gives us some idea of which genes are overexpressed. We can then identify potential products of those genes. Finally, we can make peptide sequences and raise antibodies to then search for those particular proteins in serum. From the proteomic standpoint, we can identify proteins through various mass-spectrometric-based methods and even determine the protein sequence from that mass spectral data.

Looking to the Genome

How has molecular diagnostics been beneficial in your research?

It has been helpful because it tells us what to look for. There is not as strong a correlation as we would like to see between gene expression at the tissue level and protein overexpression in the body fluids, whether blood or urine. Although there hasn’t been a direct correlation, molecular diagnostics provides good leads for where to focus our research efforts.

Do you find that genomics and proteomics complement each other, or is one a stronger tool than the other?

I think proteomics is a stronger tool because it’s a quicker route to the answer. In proteomics, we can take serum and actually identify which proteins are overexpressed, which is a more direct route than working backwards from the RNA measurements reflecting gene expression.
This second approach requires establishing that a gene is overexpressed, then determining which associated protein is likely to be present in the serum, and finally building an assay to detect that protein in the serum. Although it’s less direct, genomics is a little bit easier and more reliable, at least at the present time. So in truth, they are complementary.

Will the improvement of molecular tools improve genomic correlations and enhance your research?

No, I don’t think the lack of correlation is the fault of the genomic tools. I think the genomic production of proteins is a more complicated process than we realized.

Not every gene that is overexpressed actually leads to the identification of a specific gene product. Some proteins are not produced from genes that are overexpressed. Alternatively, the expressed proteins might have been modified so that they are not detectable.

If proteins have been altered through either alternative splicing or posttranslational modification, their detectability by immunoassays may be hindered.

Assay Evolution

What are the latest advances in assay development?

In the area of discovery, the greatest progress has been made from using the advancements in 2-D gels and high-performance liquid chromatography fractionation of proteins coupled to mass-spectrometry. Additionally, developments in mass spectrometry instrumentation and protein profiling have generated more discovery information.

From the more-quantitative aspects of development, the immunoassay continues to be the most valuable method for developing marker assays. The biggest development in immunoassays has been in the production of antibodies, either through recombinant DNA techniques or through phage display technologies. The display technologies present a wide range of antibodies, allowing scientists to identify, select, and produce those antibodies with the appropriate binding characteristics.

What roles do molecular technologies currently play in assay development?

For circulating DNA markers, DNA hybridization technologies utilized on chips or real-time PCR assays have provided a very significant contribution.

One way to measure gene expression is to measure the level of the messenger RNA (m-RNA). However, for specific analyte measurement, molecular techniques are limited when locating circulating mRNA for a particular gene, unless the measurement is in tissues or serum. Some molecular techniques form the basis of signal amplification techniques that can significantly enhance the detection limit of immunoassays.

What types of assays are most useful for IVD companies?

It depends on the analyte. When detecting a particular circulating protein, predominantly immunoassay-based methods will be used, but when detecting DNA or RNA, molecular techniques will be used. 

Immunoassay development offers the greatest opportunities for IVD companies. By the time a company acquires a new marker, the analyte will have been characterized, and the most economical analytical route is the immunoassay.

What future advances in assay development do you foresee?

Multiplexing of analytes is going to become more important, much like the current multiplexing of cytokine assays. Assays will also become more 
micro, in that they will use smaller sample volumes. Increasing the sensitivity of assays to detect low-abundance proteins that were not easily recognizable in the past will become even more of a necessity. Modern instrumentation methods are already able to detect these proteins at very low concentrations.

Which of those goals is proving to be hardest to attain? 

The multiplexing approach has not been as easy to accomplish for proteins as it has been for DNA-based analytes, simply because of the binding characteristics of antibodies and the inability to get quantitative data in a multiplexed format from mass-spectrometry methods. So I think multiplexing has probably been the most difficult issue to deal with.

Are IVD manufacturers involved in funding up-and-coming research and researchers?

Generally IVD companies do not fund external developmental projects. IVD companies typically conduct their own development after they acquire exclusive rights or intellectual property from investigators and institutions. Companies primarily conduct their own in-house method development, establish their own data reliability. Perhaps FDA requires that sort of hands-on experience.

Clinical Collaboration

When IVD manufacturers look to collaborate on clinical trials with an academic or other research laboratory, do they use the lab’s published work as their principal selection criterion?

There are probably several things that the IVD industry looks at. Certainly expertise in a particular field, as you mention, is one of them. But a more important issue can be access to the appropriate samples. A researcher’s commitment to conducting the clinical trial in the manner in which it needs to be conducted is important, because not everyone has the same level of interest in conducting clinical studies.

Many people are interested in basic work and don’t really possess the expertise or the resources to do a good clinical study. Others may have expertise in clinical studies and not necessarily in that particular marker technology. There are no clear-cut guidelines, but if a lab doesn’t have the basic resources required for a particular study, the manufacturer won’t pursue a collaboration.

Please elaborate on the role of an institution’s samples in an IVD company’s study.

A few companies would like to have full access to tissues and blood samples so they can apply some unique analytical capability that they may have and may want to develop on their intellectual property. For instance, if they’re using mass spectrometry, they may wish to do their own discovery. The best example of that is the SELDI technology from Ciphergen (Fremont, CA) currently used by many IVD companies and academic institutions for discovery.

However, during the developmental stage, IVD companies are inclined to take in the new technology and develop it for their own intents and purposes without conferring or consulting with the originator. 

For example, an IVD company might become excited about a new discovery—some academic report on preliminary clinical studies of a new tumor marker, for example. The company would then acquire that methodology and decide that the discovery technology needs to be replaced with a more robust method that is suitable for clinical studies. Typically, companies will do this sort of development themselves to get a better grasp of the technology, rather than ask the inventor to do it.

Working with Academia

What is the best way for IVD companies to develop and maintain good working relationships with clinical laboratories and with academics and other researchers?

Establishing good relationships requires keeping the candidate validation groups informed of what is in development and asking for their participation in the development of clinical trials. They should consider collaboration both in writing the initial protocols and in evaluating the data.
As a researcher, what have been your experiences in working with the IVD industry?

The work has been mutually beneficial. In the academic and clinical arena, one of our roles is to help facilitate the implementation and use of new cancer tests, either for early detection or monitoring patient care. We feel that this work must be done in an academic and clinical setting, under the direction of clinical investigators, 

Copyright ©2003 IVD Technology