For years preceding the completion of the Human Genome Project, genome researchers were trumpeting the revolutionary breakthroughs, discoveries, and treatments that would result from this work.  When this failed to immediately materialize, due to a gross underestimation of the true complexity of the human body and disease -  investors, media, and the general public quickly lost interest.  Fast forward 10 years, and it appears that the genomics industry is gun shy choosing to talk about their tools and not how an Ashkenazi Jewish rabbi has to tell two in-love members of his synagogue that they can’t marry because genetic testing says they are both carriers for canavan disease.

From my position here at GenomeQuest, I have been fortunate to work with leading edge customers in the clinical space who are using next generation sequencing for disease diagnostics today, not tomorrow and not after “we figure out how to deal with the data”.   It’s great to tell stories about terabytes and petabytes of data, but until we can translate that into treatments and prescriptions it remains a distant pipe dream in the public’s eyes. After all, isn’t that why we build these tools? Not to talk about how much data they can churn out but how they can be used  to affect people’s everyday lives.

Talk about night and day!

While it’s reasonable to question if and when core reseachers will understand the full extent of NGS data, it’s equally reasonable to get excited about applying what we know already.  Simply stated, I would argue that diagnostics — specifically consolidated molecular testing (CMDx) — is a clear killer app for NGS.

No doubt, labs are aching for and advancing to CMDx today.  And, no doubt, their patients are realizing undeniable and measureable rewards of faster/better/cheaper diagnostics.  Moreover, a standard platform (NGS) creates a wonderfully virtuous circle for them to provide increasingly better diagnostics for more and more diseases.

When it was introduced, the Mac was an amazing technology looking for a critical mass application — then along came desktop publishing and computing history was changed forever. I think the same opportunity exists for NGS, CMDx, and healthcare.

As evidence, I offer you real world “faster/better/cheaper” excerpts from two simple questions in our grant applications:

– What are the health benefits to your patients?
– What are the business benefits to your labs and partners?

See if you tend to agree.

Below are my top “takeaways” from the conference and workshop:

1. 80% say sequencing value will be in “interpretation”

In the case study on “sequencing technology” run by Richard Hamermesh of HBS, the audience was asked:

In 2021, which of the following sequencing segments will be strongest: hardware, consumables, service, or interpretation?

The answer: eighty percent said interpretation.  The viewpoint that interpretation of the data will be challenge #1 in PM was repeated by several presentors, including: Lee Hood of ISB, Stephen Spielberg of the FDA, Dave King of Labcorp, John Nierderhuber of Inova, and Hakan Kakul of Pfizer.

Immediate good news here: I think that we’re getting over the initial fear of the whole-genome sequencing (WGS) “ocean of data“.  Yes, we’re understanding that WGS is a technology and that we don’t have to digest everything it produces — we can focus on the data of most interest and use.  Just like we don’t have to watch all 800+ TV channels to justify and enjoy cable service, there’s amazing value in “targeted interpretation”.  And, over time, we’ll learn more and expand our targets.

2. PM industry integration is happening.

In order to take hold, PM requires a tighter integration of the healthcare pieces — otherwise, benefits won’t accrue to the investors and the new economic wheel won’t turn.  Here, I am encouraged as I see a combination of top-down and bottom-up change agents at work.   Industry leaders, such as PMC/FDA/VA,  are proactively creating an environment for motivating and guiding this integration.  And, industry players — including VCs, Dx companies, labs, sequencing vendors, and payors — are working amongst themselves to understand and incrementally effect this integration.

(BTW, progress and thoughts on the regulation front: we heard from many Dx leaders that the FDA is open-minded and willing to be led in PM –- so meet early and often in the development/approval process.)

3. Our aim should be squarely on whole-genome sequencing.

While the intemediate technologies of gene-panels and whole-exome sequencing will offer substantial rewards and lessons learned, the economies-of-scale and medical revelations granted by WGS justify that our major investments and best minds focus on this final destination.  After all, according to Partners Scientific Director Scott Weiss, “WGS will render targeted sequencing obsolete in 4 years“.

As breathtaking as the falling cost-curve of sequencing has been in the past decade, I think we’ll be equally awed by sequencing cost and quality improvements in the next few years.   Specifically, I believe that the market force of an expanding set of 3rd generation vendors and technologies (over 10 listed in HBS case study) will enable whole new PM applications and create whole new markets.

4. VCs and Wall St. are becoming increasingly supportive of PM/MDx.

I appreciated the inflective thoughts of Brook Byers of Kleiner (“we are entering the early days of PM“) and Amanda Murphy of William Blair (“we’re at an economic peak of uncertainty in PM“).  Byers was most enthused about the Dx side and influenced by long-term, wellness managers and payors.  Also indicating enthusiasm are the funding of the above 3rd generation vendors and warming of MDx acquisition activity, including the $.5B acquisition of Clarient by GE Heathcare.

One of the more interesting dynamics as we arc from genomics for drug/disease research to clinical application is the associated pivot from a largely foundation-funded market to a commercial market — aiming to improve and provide economic value to a $5T global healthcare market.   In my opinion, this adjustment to a commercial market is where much of the management challenge and opportunities lies.

5. The industry needs to agree on an global “atlas” of genotype/phenotype database.

This is front and center of many thoughful PM talks over the past year.  At PMC, I began to see significant movement in this direction: a. organizations planning the integration of genomics with the EHR which will speed genotype/phenotype associations, b. teams structuring phenotype-rich, genotype-enabled clinical trials, and c. a proposal for a standard platform for research labs to establish and share clinical evidence.

6. Much effort/$ should be put into enabling PM at community hospitals

Gregory Feero of NHGRI thought it worth noting — from a conference operating deep inside HMS — that 95% of heathcare happens outside academic medical centers.  Well played.  One example of progress is the KEW Group, which is building a national network of PM-based community cancer care centers.

7. The Dx industry is preparing its economic case for MDx.

On the one hand, Dx informs 70% of medical decisions and MDx holds immense promise for fundamentally improving healthcare.  On the other, Dx amounts to just $40B of the $5T healthcare industry and the assigned skeptic at the conference (artfully played by Dr. Ezekiel Emanuel) waged a blistering attack that PM makes little/no economic sense.

What gives?

Clearly, MDx is undervalued in the healthcare industry — our payment system favors therapy/technology over diagnostics/interpretation.  One giant step forward would be to argue the economic case for MDx.  And I foresee the PM Coalition and industry preparing two pieces: a holistic, multi-factor economic argument for MDx to spur reimbursement change and associated decision support to inform day-to-day care.

8. We continue to be guided and inspired by Dr. Lee Hood

In accepting his leadership award, Lee Hood offered a wonderful combination of science vision and medical practicality to guide the path forward.  In particular, he was most excited about: a. family-based studies, proven to reduce errors by 70% and shrink solution space by 100x, b. analysis of single cells, and c. system-level considerations and tools, required  to addresss the “grand challenge” of PM (complexity).

Overall, it provided a wonderful and thorough update on the state of genetic research, practice, and plans – bravo to the show organizers and all presenters!

Below are the top seven thoughts I heard in my sessions and around the conference:

1. “We’re all a brotherhood of mutations”

… a wonderful expression of community by Michael Hayden of Canada’s Center for Molecular Medicine that was met with loud applause in the Medical Genetics panel debate.  It was a welcome counter to Dr. James Watson’s disturbing “some people are genetic losers” from the starting panel the day prior.

2. “I was assigned this case by a computer and it was a stroke of great providence that my clerk had a PhD in molecular biology”

…  testimony from U.S. District Court Judge Sweet who, of course, ruled to invalidate the Myriad BRCA patents.  The audience in this “Who Owns the Gene” panel was wowed with his presence and clearly leaned anti-patent — best expressed by a member who proclaimed “Darwinian selection will eliminate gene patents”.

3. “Targeted NGS for the clinic is ‘ready-to-go’”

… a confident claim made by Ming Qi of the Zhejiang Medical School and buttressed by Joris Veltman who is today using NGS for clinical work in blindness at the University of Nijmegen.  Hayden raised serious warnings of over-promises and incidental consequences but the Medical Genetics debate settled on “acting on what we know” and coined the handy term “targeted interpretation”.  Representing the general MDx market and in related conversations, Dr. Roberta Pagon of GeneTests suggested that consolidated gene panels would be a common entry point to NGS for many labs.

4. “This would have saved $7M and prevented 400 adverse events”

… the result of a retrospective PGx simulation by Vanderbilt on 53K patients over a six month period.  Dr. Schildcrout also reported that ADR is the 4th largest cause of death in the US, killing 106K patients per year, and that PGx has grown from a start by the FDA in 2007 to over 100 drugs now carrying PGx labels.  Dr. S. Qin reported that China now runs a nation-wide network to collect and share adverse events.

5. “Physicians don’t want to learn it ALL, they want to learn ENOUGH”

… telling observation made by Dr. Farndon of the UK Genetics Education Center, which has been training physicians in genetics for over six years.  ACHG president Lynn Jorde stressed the need for this continuing education in U.S., noting that only 26% of physicians have a genetics education and only 10% make use of PGx.  Good news on when that happens — MEDCO reported that, in a Labcorp-supported study, 50% of physicians took advantage of a piloted PGx system when deciding treatment and 40% of the time changed treatment when directed.

6. “We all seek a global phenotype/genotype database”

… perhaps the most resounding agreement by the Medical Genetics panel and declared by Han Brunner of the University of Nijmegen (and one of the better moderators I’ve come across).  In separate discussions with lab leaders Heidi Rehm (Partners) and Madhuri Hegde (Emory), it’s clear to me that, while such a “mother of all databases” will be some time coming, indeed academic/commercial collaborations will make transformational and practical progress for health care in this direction in 2012.

7. “Give patients the choice”

… not exactly a quote but my paraphrasing of the ethics panel consensus on the issue of informing patents of incidental findings from whole-genome sequencing.  More generally, I’m a believer that, as more personalized medicine services emerge and news of their life-changing benefits escapes, patients will demand more information, become the fastest learning group in health care, and serve as a positive and major force in this disruption.

Oh, and…

8. “Maybe we spend a couple weeks in Burlington every winter”

… from a quite smitten skier and wife, after our stay in northern VT on the drive from Boston.

]]> http://blog.genomequest.com/2011/10/top-7-thoughts-heard-at-ichg/feed/ 0 http://blog.genomequest.com/2011/09/personalized-medicine-for-cancer-on-cbs-evening-news/ http://blog.genomequest.com/2011/09/personalized-medicine-for-cancer-on-cbs-evening-news/#comments Fri, 23 Sep 2011 18:51:35 +0000 Tony Flynn http://blog.genomequest.com/?p=456 here [...]]]> On the 40th anniversary of the National Cancer Act, the American Association of Cancer Research (AACR) this week issued its first “Cancer Progress Report”.

It’s an outstanding, 84-page narrative that outlines the progress in our understanding/treatment of cancer and the criticality of genomics in transforming patient care going forward.

You can download the full report here and watch the CBS Evening News story here.

Congratulations to the AACR on this clear and powerful piece.

They have much to be proud of – the use of VistA has enabled the VA to reach a pharmacy prescription accuracy rate of almost 100%, and the VA outperforms most public sector hospitals on a variety of criteria, enabled by the implementation of VistA. VA hospitals using VistA are one of the few hospital systems that have achieved the qualifications for HIMSS stage 7, the highest level of EHR integration. The goal is to create a usable, next generation EHR that can anticipate and be flexible enough to accommodate where clinical informatics will be headed over the next decade, and the big target is genomic data from the VA’s own Million Veterans Program (MVP).

The ambitious MVP is an ongoing initiative, according to Veterans Affairs Eric K. Shinseki, “It is my honor to join with so many fellow Veterans in keeping VA at the leading edge of genomics research.  This innovative research program will support VA’s mission to provide Veterans and their families with the care they have earned.” The Million Veteran Program is a trailblazing VA effort to consolidate genetic, military exposure, health, and lifestyle information together in one single database.  The database will be used only by authorized researchers with VA, other federal health agencies, and academic institutions within the U.S.—in a secure manner—to conduct health and wellness studies to determine which genetic variations are associated with particular health issues.  By identifying gene-health connections, the program could consequentially advance disease screening, diagnosis, and prognosis and point the way toward more effective, personalized therapies.

In a recent workshop held at the prestigious Institute of Medicine, called “Integrating Large-Scale Genomic Information into Clinical Practice”, several key scientists and clinicians from Genome Centers described the evolution of clinical genetic diagnostics – from gene test panels to whole exome sequencing to whole genome sequencing. With the precipitous drop in the cost of human genome sequencing, the strong recommendation from participants at the workshop was, although we have interim genome-based diagnostic tests, within 2-3 years everyone will be performing whole genome sequencing and analysis because it reveals so much more information about the correlation between genetic variation and disease phenotype^1,2. This mantra is now resounding within the VA’s clinical MVP program, which is evolving with the technology:

¹ Ley at al (2008) DNA sequencing of a cytogenetically normal acute myeloid leukaemia genome.  Nature 456, 66-72 (6 November 2008) | doi:10.1038/nature07485.

² Welch et al (2011) Use of Whole-Genome Sequencing to Diagnose a Cryptic Fusion Oncogene.   JAMA, April 20, 2011. 305 (15): 1577-1584

GQ-Dx is our response to an overflowing demand by the testing market to move from Sanger sequencing-based diagnostics to next-generation sequencing. Even as recently as one year ago, the medical community was outright reticent that this new technology was years away from clinical adoption. Today, we are aware of nearly 50 academic and commercial labs that employ next-generation sequencing directly. The first movers, like our customer the University of Iowa, recognized that gene panels could be screened rapidly using high-throughput sequencing, and then a more directed Sanger sequencing approach could dramatically lower costs and turn-around times.

The gene panel market in mid 2011 is small but growing quickly. And labs who sequence the same one or two genes for thousands of patients are also beginning to consider the use of massive multiplexing to pool many patient samples for high-throughput screening followed by confirmatory Sanger.

Of course to reach $3.6B the market needs to move from panels to exomes to genomes. Already we’re in discussions with the heads of labs who see gene panels as an unnecessary step, and are moving to validate whole-exome diagnostic panels with software probes for specific variations that drive clinical action. And the dreaded incidentalomas that paralyzed the medical community even six months ago seem more manageable by some: they’ll be classified by protein and pathway impact, and assessed by qualified pathologists and clinical geneticists. Reporting of the potential impact of such incidental findings will be done based on prior informed consent between doctor, patient, and family, and will depend on a variety of factors from age to severity of disease.

Perhaps the government will slow down the pace, but until then (and hopefully thereafter), individual medical institutions are innovating, and patients are benefiting, and the clinical sequencing market is growing.

Included in the July 8, 2011 report:

• In as soon as three years, NGS testing could reach $3.6B in the U.S. (others have noted that MDx is the fasting growing segment of the $40B diagnostics market)
• Full-genome sequencing will eventually replace single mutation testing/analysis
• GenomeQuest was singled-out as a “technology leader” in informatics for analysis of tests — a critical area, it said, for commercial labs
• Full “democratizing” of sequencing will require the cost of whole-genome sequencing to fall to $100-200 per test.

Below are excerpts from the full Macquarie report, which is available by emailing dane.leone@macquarie.com.

Click here for the announcement of GQ-Dx — the molecular diagnostic decision support system from GenomeQuest highlighted in the report.

————————————————————————————————-

“The detail is in the report but the quick answers are: 1) NGS diagnostic services will likely be supported by a coordinated care model that involves a primary clinician, genetic counsellor, technology team and a medical review board; 2) We estimate the refractory cancer patient testing market could be ~US$150m within the next three years but under a preventative care scenario, NGS testing could reach ~US$3.6bn in the United States; 3) Adoption of NGS diagnostics could likely erode the market for „single-marker‟ gene tests, create a significant opportunity for IT companies and will drive the development of new sequencing technologies.”

“Given the size and complexity of data that is generated by NGS, even for targeted or exome only studies, the analytical portion of the experiment has continued to be bottle-neck for commercial organizations (independent laboratories) to feasibly offer NGS based testing to clinicians. In our view, informatics companies are just beginning to tackle this complex issue and the technology leader in the space currently is a private company called GenomeQuest.”

“In our view, a conceptual future in which every human is sequenced prior to birth, after birth and on a regular basis throughout their lifetime, would require a new generation of “smart‟ sequencers. The criteria for a truly democratizing sequencing technology, in our view, would require 1) Sample preparation on par with the current generation clinical molecular diagnostic platforms; 2) Complete time to result of several hours; and the 3) Cost of a whole-genome analysis ~US$100 – 200.”

Based on previous recommendations of the “IT Assessment Report of the National Institute of Health (March 28, 2011)”, and the recent IOM report called “Generating Evidence for Genomic Diagnostic Test Development”, there has been a universal plea from the Federal Healthcare IT community that we begin the process of molecular diagnostic test aggregation based on individual patient genomes.

Some of the barriers to adoption include:

• Education of clinicians to better understand molecular diagnostic tests based on whole genome data.
• Enhanced storage and management of genomic data in our Clinical IT systems.
• Implementation of Clinical Decision Support (CDS) systems, to enable the busy physician to make rapid, accurate judgments.

So what does this all mean for the hospital CIO and administrators? I would suggest that we be prepared for data management and storage adequate for whole genome analysis of patients. And the application of decision support tools to make more informed diagnostic decisions. Even if  many physicians don’t yet recognize the value of these data, you know that patients will be demanding this capability, to create a more efficient and lower cost healthcare environment.

This is a question that is often asked when I demo the GenomeQuest platform to potential customers. I always answer that question in three phases.

The first phase goes like this: “It’s really fast, it’s certainly not any slower than BWA/Bowtie or anything else out there.”. Next question is always: “Well, do you have any benchmarks?”.

Which nicely transitions me into the second phase of the answer. This phase is much more rigorous and usually starts with: “Well, it depends. Let me try to explain.”:

• Any good computer scientist can write a mapping algorithm that is really fast. However, that doesn’t mean anything unless it produces the kind of results that are needed. The NGS application you’re working with matters here. For example, finding genetic variation in human disease requires very accurate mapping with mismatches and indels. In contrast, with digital gene expression in maize you can cut a lot of corners. You just need to have a rough idea of the number of alignments on a transcript.
• Then there is the matter of the data and the technology that produced it. Do you have long reads (more than 120 bp)? Will your mapper handle them? Will it also increase the number of mismatches / indels it can use to align a read? Will it significantly slow down execution, or eat up all your memory when reads get longer? Does this mapper also support local alignments when you need them? Will it align in colorspace? In paired end mode? I could go on.
• Next there is the matter of connecting results to the following step in your pipeline. How long does it take to de-duplicate a  1TB alignment file in SAM/BAM format? Or to find those alignments who’s position overlaps with your exome capture experiment, or all dbSNP entries? At GenomeQuest we have a very efficient way of storing and handling alignments (including sequences/annotation). This saves real time, especially when compared to the alignment step itself (it saves a lot of disk space as well by the way).

By the time we get to the third phase of the answer I’m usually much more confident: “Well, how fast do you need our mapping algorithm to be?”.

• Does it really matter how fast the read mapper is, as long as it’s comparable in performance to other algorithms for most common use cases? Does it matter if you have the alignments in 2.5 hours instead of 3? Maybe if you analyze thousands of samples per week it matters, but then other things like reliability and professional software support should matter as well.
• Do these other algorithms scale with the hardware you throw at them? How easy is it to run a read mapping on 64 compute nodes, with 2 CPUs, 8 cores per CPU per node? What about if you double the amount of hardware? Will you go twice as fast? With the GQ- Engine  you will. Want to run on 1024 nodes? That’s possible.
• Are you asking about speed for a single run, or the throughput for a bunch of runs? Last weekend I ran 2000 NGS read databases though our read mapping workflow (low coverage genome sequencing, about 80M reads per database). I started them on Friday afternoon, went for drinks with my friends that evening, had a nice family dinner on Saturday afternoon, and watched a movie with my kid afterwards. The runs were finished before I woke up on Sunday. No hiccups, no failed runs, no logs to monitor, and – best of all – no “one million-file” directories to organize. There were a lot of other customers on the system that weekend, doing their NGS analysis as well.

If we ever meet for a demo, please ask me this question. I love to talk about it.

Henk Heus, Ph.D.
VP Product Management & Services
GenomeQuest Inc.

At GenomeQuest we use an extended version of the GASSST read mapping algorithm (among others). Read about it here in Bioinformatics here: http://bioinformatics.oxfordjournals.org/content/26/20/2534.abstract