This is part 2 of the blog covering the 7th Personalized Medicine World Conference held at Computer History Museum in Mt. View, CA. This part covers the panel discussions from track 1 and few talks from track 2 (established companies) and track 3 (startups pitching for fund raising). The 23 companies in track 3 also competed for the “most promising company” award at PMWC. 3 companies were selected to give a final pitch in the main hall to win the award (more on this below).
Kevin Davies, Editor Bio-IT World and author of $1000 Genome, started day 2 with panel discussion titled “Killer Apps, Genome interpretation and Future of the NGS”. The panel consisting of sequencing technology developers/providers included Steve Quake from Stanford University, Cliff Reid from Complete Genomics, Paul Billings from Life Technologies (filling for Hunkapiller of Pacbio) and Steven Roever from Genia. All panelists agreed that the dreamed cost of $1000 genome is still away and the sequencing costs will continue to keep going down. Discussing genome interpretation, Cliff classified the current sequencing data into three groups; known (well understood) part of the genome (~1%), the grey part of the genome (5-10%, needs more data studies or analysis) and the dark matter or unknown part of the genome. He commented that in the coming years, as we gain more data, the grey part will stay but our knowledge of the known part will increase. Applications like the microbiome will likely play an important role in achieving that. Cliff also raised the issue of data sharing and mentioned that though Complete Genomics has data on more than 15k genomes, the IRB’s do not allow putting them together and analyzing them.
Another panel discussion on day 2 was titled “drug discovery and development” and consisted of Maureen Cronin (Celgene), Steven Stein (Novartis) and Kelly Constable (Amgen). Both Steve and Kelly shared interesting new approaches that they are currently testing to implement personalized medicine. At Novartis, they have developed the signature trial with the idea of rapidly matching patients to therapies that target their molecular abnormality in correlation with patient’s sequencing data and response. He mentioned that the well-defined contract negotiations that define the issues like subject injury, identification, drug efficacy and toxicity etc. are critical and have helped in testing the treatments with pace. At Amgen, they are building a system to improvise personalized medicine that came out based on survey and discussion with physicians and something that would fit in the current routine of care. This system based on predictive modeling gives physician’s choices of treatment and provides speculative data on the effect of treatment on the patient if or not changing a course (e.g. giving a new drug or administer blood transfusion).
I attended few of the talks from track 2 and topic “Biomarker and Companion Diagnostics”. Andre Marziali, CSO of Boreal Genomics showcased the performance of the test which is focused on detecting tumor mutations in cell free DNA in blood. Using the impressive technology based on electrophoretic separation to enrich mutated DNA, he represented the high sensitivity of mutation detection and greater than 88% concordance between the cell free DNA and tumor biopsy on grade 2 to grade 4 tumors of 17 patients. They plan to launch a 120plex based test in April this year. With a similar focus, Trovagene presented their assay to detect cfDNA in urine using ultrasensitive digital PCR and NGS. With sequencing advancements and rapid success of NIPT, I presume more of these assays will be available in the market soon.
InterLeukin Genetics presented their product PerioPredict, a test that takes into account three factors to group people in low risk and high risk for periodontics. This is based on Interleukin genes, risk factors like smoking and tooth loss. Using this classification, they can define also number of visits required to the dentist and predict pace of progression of disease. As a selling pitch, he claimed that the companies can have return of investment in 2 years and save lot more money by moving towards early and preventative medicine.
Alan Wu from UCSF gave a talk on preparing medicine students for the era of Genetics medicine. Moving forward to enable these patients to handle the genetics, they had students sequence the DNA for CYPC29 gene in UCSF lab. Importantly, for data interpretation and counselling they formed groups of students from different medical backgrounds e.g. a group consisting of pharmacy student, experienced physician, medical student and a genetic counselor. Each student in the group will be act as the patient and counseled by others.
Many companies also gave talks under the session “Big Data: Storage and Management” on solving the issues of data integration, bioinformatics analysis and better structure for –omics information. A panel discussion on “how will bioinformatics scale “ included Jonathan Hirsch from Syapse, Somali Dutta a Syapse user and head of IT for many labs at Stanford and Sasha Wait Zaranek. They envisioned that in the future with implementation of proper systems, (e.g. Syapse) we will be better able to structure and connect different maps (-omes). Currently, if there are many labs working together on project, it is very difficult to share data and meta data in a proper manner. Using Syapse, Somali and the Stanford core lab had a huge success in handling data with different collaborators on Integrated Omics profiling project (IPOP). In the end to be successful in making a system for GM, physicians want a system that fits in their workflow, provides an interactive base report and helps them making quicker, better informed decisions about patient’s treatment.
Bina Technologies also showcased their product and the new upcoming Binalite (for researchers). They also shared the experience of using their technology in area of neonatal care where speed diagnosis and rapid interpretation is very important.
22 companies that presented in track 3 out of which three finalists Epic Sciences, Neurotrack and T2 Biosystems were selected for round 2 and to give a talk in the main hall. Each of them got 8 minutes to woo the panel/crowd and win the prize of the “most promising company”. The judges in the panel included Dr. Atul Butte, Dr. Warren Hogarth and Dr. Diego Miralles and held 60% of votes and crowd 40%. Each competitor was asked one question by one of the judges.
Epic Sciences located in La Jolla, have developed an assay to identify and characterize circulating tumor cells (using some markers). These cells could be separated as well and analyzed further with technologies like NGS. Being non-invasive in nature, multiple biopsies are possible as well as the possibility to gain insight in to different clonal subtypes. They were pitching for $20M series C funding.
Neurotrack, a 2 year old company from Palo Alto, has developed a computer based test that can predict the onset of Alzheimer’s disease 3-6 years earlier than the symptoms appear. The product is based on the research done at Emory University and the duration of the test is now reduced to only 4 minutes to help in the disease prediction. Answering to the question by the panel, the absence of the treatment seemed to be the issue in some cultures especially America, overall I see it is very unique and great product. Neurotrack also won the award for the 2013SXSW best health technology startup.
T2 Biosystems from Redwood City have been working for 8 years to build a diagnostic test to detect sepsis much sooner than currently possible. They have built a test that uses magnetic resonance and can detect bacteria or fungi at concentrations as low as 1 CFU/ml. This test can reduce the sepsis detection time from 2-5 days to 5 hours. T2 Biosystems was chosen the most promising company both by the judges and the crowd.