What are you going to do if a student commits suicide after finding out they have the genetic marker for early onset Parkinson’s disease?
This is not a typical discussion between professors and university presidents. However, a new class at the Tri-Institutional M.D.-Ph.D. program at Weill Cornell Medical College, Sloan-Kettering and Rockefeller University engendered such concerns. In fall 2011, Dr. Chris Mason, Cornell professor and head of the integrative functional genomics laboratory, envisioned a class for M.D.-Ph.D. students in clinical genomics. Mason’s work focuses on developing the promises made over a decade ago during the Human Genome Project. From the entire 3.1 million base pair genetic code, the field of clinical genomics seeks to extrapolate not only expected reactions to drugs, but susceptibility to disease and predicted long-term health outcomes. Mason proposed an unorthodox textbook for the class: the students would perform their class research on themselves by sequencing their full genome. So how did this goal led to a flurry of inter-institutional emails, meetings, invectives such as the one above and — to an extent — fear?
On a superficial level, performing genomic analyses of one’s own genome appealed to our scientific curiosity, and perhaps, to our narcissism. What could be more self-indulgent than gazing at your own reflection for 10 weeks, albeit genetically? Dr. Olaf Andersen, the head of the Tri-Institutional M.D.-Ph.D. program, concisely described the perils of genetic sequencing, saying, “The chance of finding something as deleterious as early-onset Parkinson’s or Huntington’s disease is extremely small. Still, the chance is finite and non-zero and we have to be prepared for that possibility.”
Fortunately — or perhaps unfortunately — few genomic conditions are as strikingly binary as the example of early-onset Parkinson’s disease. Far more likely is that an analysis will reveal a 30-percent increased chance of heart disease or a 15-percent decreased chance of diabetes. Mason contextualized, “Clinical genomics is all about probabilities, and nothing is set in stone.” Andersen countered by describing the concept of the “incidentalome;” essentially, everyone is walking around with 100 or so mutations that according to current scientific literature, should have immediate phenotypic expression (e.g. congenital deafness) but in reality, do not.
Probabilistic correlations and the incidentalome highlight the work needed in clinical genomics. From a clinical perspective, such in-depth genomic sequencing is akin to a whole body CT screen. Some medical centers market this as a proactive way to stay healthy, by imaging the body in detail. However, in reality, such screens either reveal nothing, giving a patient a possibly false sense of health, or reveal abnormalities that will eventually prove unremarkable but will induce multiple follow-up tests and associated costs and morbidities. Andersen concluded, “We simply do not have the infrastructure or knowledge in place to responsibly care for outcomes students may discover. Clinical genomics is a powerful tool, but we must proceed cautiously.”
Mason discovered that the legal implications lagged even behind logistical considerations. Administrators conferred with legal counsel at Cornell’s affiliate, NewYork-Presbyterian [??1] Hospital, and discovered that the Health Insurance Portability and Accountability Act has no provisions for this type of information. Approval from the WCMC Institutional Review Board was also unnecessary as full genome sequencing was not technically an experiment. The Genetic Information Nondiscrimination Act of 2008 also did not apply.
Ironically, such ambiguity — both scientific and legal — was a driving force behind sequencing students’ genomes. As Mason explained, “the plan was to have students examine their own sequences and say, hey, I’ve got an allele for taller than average height, but I can barely qualify to ride a rollercoaster. Some of the gene interactions are beyond our current knowledge, there are environmental considerations as well, and just as in Gattaca, your genetic code is not the final determinant of your life.”
To gain more insight into the genetic counseling component, I spoke with Dr. Jessica Davis, a professor and clinical geneticist at NYPH and the Hospital for Special Surgery. Dr. Davis cautioned, “in teaching counseling it is important to not get swept away by the tech hype.” When asked to explain the expected role of the general physician, Dr. Davis prognosticated that, “genetics will be part of everything,” but intense analysis will fall to the genetic counselor. How many are in the United States and Canada? About two thousand. For something that promises to become as ubiquitous as blood tests, two thousand trained professionals for a population of over 300 million highlights the discordance between expectations and logistics.
Other universities are pushing for levels of genetic sequencing of students — with varying success. Stanford University recently conducted a similar genomics class in which students were given single nucleotide polymorphism panels. These are not as revealing as a full genome sequence but can identify known common gene variants. The University of California at Berkeley, presumably not wanting to be outdone by their perennial cross-bay rivals, attempted to test the entire incoming freshman class for SNPs in nutritional pathways. The state of California green-lighted the project but said that individual data could not be returned to students.
It may appear that researchers at WCMC are eager to develop a promising new technology while administrators seek protection from disruptive outcomes, legal and otherwise. The M.D.-Ph.D. students are symbolic for the underlying argument. Are the students in the clinical genomics class acting as researchers, clinicians or patients? This unique opportunity first enticed the students but also proved to be a frustrating conflict of interest.
Mason stipulated that in order to conduct the full genome sequencing for students next year, “the concerns of administrators and institutions will have to be mollified. But unfortunately there is always a chance that a student will discover a BRCA mutation [indicator of strong predisposition to breast and ovarian cancer].” The bottom line is, “more information is always a good thing. It may be daunting, but that is exactly why we need to train scientists to properly interpret the data.”
Earlier this year two companies, Life Technologies and Oxford Nanopore, both claimed to be able to sequence the entire human genome in less than a day for about $1000. With technology improving and costs decreasing, the question is if we as clinicians will utilize genomic sequencing, but when. And will we possess the skill set to do so as competent physicians?
Hemingway timelessly replies, “The shortest answer is to do the thing.” For $1,000 and with appropriate training and computing power, anyone can dig through their genome. I may have found my graduation present.
Daniel Rosen is a first-year M.D.-Ph.D. candidate at Weill Cornell Medical College. He may be reached at email@example.com. What’s Up, Doc? appears alternate Fridays this semester.