Climbing the Informatics Mountain—It’s Not a Bubble: Highlights From the 2011 Tri-Conference

With each year, Cambridge Healthtech Institute’s (CHI) Molecular Medicine Tri-Conference continues to expand. The 18th edition, held February 23–25 at the North Hall of San Francisco’s Moscone Convention Center, increased the parallel tracks to 13. Attendance was up by 10%. Tracks included Circulating Tumor Cells (CTCs), Informatics, Medicinal Chemistry, Translational Science, and Cancer Biologics. Of course, it would take a team of journalists to even begin to completely cover this multifaceted meeting, so you are getting only the highlights.

Circulating tumor cells

Over 140 years ago, CTCs were implicated in tumor metastasis. The problem has been how to isolate and measure them. This requires analytics, but products are only now starting to appear. Several alternatives were on the floor at the Tri-Conference, including 13 of the 120 posters. Most involved antibodies but two use dielectrophoresis.

Dielectrophoresis for CTCs

Silicon Biosystems (Bologna, Italy) uses dielectrophoresis (DEP) to select individual cells in a microchip. Recall that DEP involves placing a suspension of particles (in this case a population of cells) in an asymmetric electric field. Even if the particles are uncharged, the field will induce a dipole. The amount and direction of the dipole are a function of the properties of the particle and the surrounding liquid. If the field is not uniform, the particle experiences a net force directed toward locations with increasing (positive dielectrophoresis [pDEP]) or decreasing (negative dielectrophoresis [nDEP]) field intensities.

Silicon Biosystems applies the electric field to a 2-D array of microchambers in a fluidic chip. Each chamber has an electrode, sensors, and logic. Each electrode creates a DEP cage in which individual particles are trapped and levitated. Individual particles of a heterogeneous population can be localized into individual chambers for counting and further characterization with dyes, etc. The device concentrates by sorting the abnormal members of the population and releasing the normal members to waste.

DEP + field flow fractionation

ApoCell (Houston, TX) introduced an instrument that quickly separates CTCs and other rare cells from whole blood. The instrument uses dielectrophoresis coupled with field flow fractionation (DEP-FFF) for the cell separation.1 As in DEP, an alternating current electric field is applied, thereby generating DEP forces onto cells which act perpendicular to the flow axis of the FFF channel. The transchannel motion of the cells depends on the differential permittivity between the cells and suspending liquid. Particles experiencing strong attractive DEP forces are pulled toward the floor of the flow chamber, where the liquid flow velocity is low, and are collected by a skimming process. Particles that experience repulsive forces migrate to the center of the flow chamber and are eluted quickly.

The separation seems quite general and robust as it is based on inherent differences between cancer cells and normal cells. ApoCell has examined over 20 cancer cell lines indicating the DEP-FFF CTC isolation technology is applicable to a broad range of cancer types. The normal components of blood, including monocytes, granulocytes, B and T lymphocytes, and erythrocytes, respond differently from cancer cells under proper DEP conditions enabling isolation of cancer cells. As no tagging or labeling is required, isolated cells are intact, viable, and cultureable. Thus the cells are available to measure drug sensitivity, which could guide treatment. Potentially, this one could harvest and expand cancer cells from an individual and then challenge aliquots with different drugs to measure efficacy as is done today with bacteria.

Poster awards

Traditionally, CHI selects two posters as most outstanding at the meeting. Each lead author is presented with a $500 check. This year, one went to Dr. Cheryl H. Cui of the Harvard-MIT Division of Health Science and Technology (Cambridge, MA) for the poster, “Capture and Detection of Circulating Tumor Cells Using Three Dimensional, Multivalent DNA Aptamer Network.” Dr. Cui explained that aptamers are superior ligands for affinity enrichment since they can contain several binding domains for the cell’s surface markers, which provide exceptionally high binding avidity. Plus, the DNA strands are long enough to extend into solution to aid in the capture of the CTCs even at a high flow rate. The aptamers are bound to a glass surface. Successful capture of different CTCs was demonstrated. The cells can be released by treating the slide with dilute restriction enzymes, which cleaves the DNA into small sections that float away.

Informatics

The other award-winning poster, “NCBO Resource Index: Ontology-Based Search and Mining of Biomedical Resources,” by Patricia L. Whetzel of Stanford University (Stanford, CA) described a resource index that can be used to facilitate location of biomedical information irrespective of source; database format; and names, including native language and jargon. The key is to create and maintain field-specific ontologies that can provide a uniform structure to inherently disparate databases. The National Center for Biomedical Ontology (NCBO) (Stanford, CA) curates and manages resources that record and cross-reference terminology used by different investigators involved in biomedical research. Currently, the NCBO maintains more than 200 ontologies. Each is context specific. With these, scientists can find and extract data from relevant datasets without imposing inflexible standards or reconstruction of the database to a common structure.

During the Q&A following the lectures, one person suggested creating a master ontology that would not depend on context. Prof. Mark Musen, also of Stanford and a coauthor of the poster, responded that ontologies are inherently context specific. Understanding the context is essential to valid use. “Sensitivity” came to mind. In analytical chemistry, sensitivity is used to describe detection limits, as in detection sensitivity or detection limit, but in clinical chemistry, sensitivity is a measure of the true positive calls. Even I could see the attraction of mapping ontologies across existing boundaries to improve understanding. But jargon exists and efforts to eradicate it are probably doomed to failure.

Later, Dr. Venkat Koka of Janssen Pharmaceutical, a J&J company (Titusville, NJ), described the alternate approach. About eight years ago, Dr. Koka was part of a team directed to replace paper with electronic information technology. The registration process had to be bulletproof, since it involved the discovery engineering of new chemical entities (NCEs) from antibodies. Since litigation was predictable, they needed to provide traceable and verifiable data to support decision-making and patent filing and, most importantly, patent defense. The team found that the existing paper process and jargon were investigator specific, which was not acceptable. The notebook-based system was changed to a rigid, projectwide work flow and supporting IT system. Most of the department acquiesced and adopted the new work flow, since Dr. Koka was a recognized laboratory expert in antibody engineering. The specific work flow accomplished the job, but it would not be suitable for other purposes.