Here Is a Human Being (21 page)

I supposed that could be true. But I wasn’t sure that the answer was for people to willfully ignore their genomes. Hadn’t we scientists spent years telling people how fantastic this stuff was all gonna be? Didn’t we justify our ongoing suck at the federal teat by promising diagnostics and cures? Didn’t we want to encourage people to take responsibility for their health and to embrace preventive measures? Hypochondria was a risk, no question. But the price of genomic information was falling every day: indeed, in the summer of 2009 Navigenics lowered its price 60 percent to $999.
³⁶
I still believed that it was incumbent upon us—the genetics community—to view this as an opportunity, not for more CT scans, but to talk about whether people really needed CT scans, what sorts of healthy behaviors they could manage on their own, and how their genomic data could play a part in that.

For decades, medical genetics has been criticized as a field akin to bird-watching, whose credo is “diagnose and adios.” In 2010, our ability to treat strongly genetic diseases was fairly pathetic. But even in those many instances where we could do little or nothing to mitigate our genetic lot in life, a few scientists and physicians were starting to ask whether we might want to know what our risks were anyway.

I
n the late 1990s, Robert Green, a neurologist at Boston University, asked his geneticist friends a simple question: Should we start offering genetic testing that would let people know their genotypes for the APOE gene, that is, tell them if they were more likely to get Alzheimer’s? The geneticists were mostly horrified. Not only was there nothing one could do to delay the onset of Alzheimer’s, they told him, but the average person would almost certainly misunderstand his/her risk, come to serious psychological harm, and maybe even commit suicide. This was crazy talk.
¹

Green is a good-looking, bespectacled man—a dead ringer for Pittsburgh Penguins coach Dan Bylsma; as it happens, he has the coach’s same understated manner. His voice is a pleasing tenor: while he was on the faculty at Emory University, his performances as part of a chorus that sang with the Atlanta Symphony won five Grammys. Unlike the disheveled basic scientists he often hangs around with, Green favors a jacket and tie. He is a self-described “clinical trialist” who for many years studied prospective—and mostly disappointing—Alzheimer’s treatments.

But perhaps, he thought, he could do some good by providing Alzheimer families with
information.
Not only did he reject the standard “this is toxic knowledge” stance of the geneticists,
²
but the idea of disclosing risk and studying its effects actually appealed to him as both a clinician and a scientist. So he did what he always did: played by the rules and set up a randomized clinical trial. One group of people with a family history of Alzheimer’s would receive its genotype information; the other would not. The group that did could get 1) good news (no copies of the APOE4 allele and therefore a 9 percent lifetime risk of Alzheimer’s); 2) moderately bad news (one copy of APOE4 and a 29 percent lifetime risk); or 3) really bad news (two copies of APOE4 and a greater than 50 percent lifetime risk). What Green’s team found was that those getting bad news would indeed get bummed out (just as I did when I learned about my phantom multiple sclerosis) … but not for long. And virtually no one regretted learning his or her risk.
³

It turns out this is the norm, whether the devastating disease is Alzheimer’s, breast cancer, ovarian cancer, or Huntington’s disease. “There are dozens of articles and they all seem to show the same thing,” Green said. “In the short term people may become upset by positive results [showing they are at high risk], but they cope quite well. And when people learn they have negative results, they are relieved and that relief persists.”
⁴

Until recently, most of the genetics and bioethics community was having none of it. As we sat in the living room of his stately house in Wellesley, Massachusetts, Green wrapped himself in a blanket on the couch and told me that over the years his studies had been called irresponsible, unethical, foolish, and unrealistic. “They said our careers would be over.”
⁵
Now his work was held up as a template for how to disclose genetic and genomic information. Green eventually joined the PGP team as a counterweight to George’s rebellious tendencies and flights of fancy: an even-keeled, NIH-savvy physician who understood the ins and outs of delicate human subjects research on genetics. “Bob was brought in specifically to keep me on a leash,” said George with an impish smile, implicitly acknowledging that that would be no mean feat.
⁶

At least two personal genomics companies included APOE as part of their SNP-based scans (Green consulted with them pro bono).
⁷
As genes go, APOE is arguably the most powerful one yet discovered that contributes to complex traits mediated by multiple genes and the environment. This is probably why both Jim Watson and PGP-10 participant Steven Pinker declined to learn their APOE genotypes (though, as we saw in chapter 7, successfully
not
knowing was not always so easy to accomplish).

Time was short. The National Human Genome Research Institute (NHGRI) folks were about to pay a visit to the Church lab in anticipation of George submitting a renewal for his big grant, one of ten such genome grants in the country. The site visit would not be a make-or-break; indeed, it would not be evaluative at all, only advisory. The NIH wanted to know what progress the Molecular and Genomic Imaging Center had made over the last five years and what George’s team was planning for the next five. The idea was to maximize the grantee’s chances for success going forward. George thought the stakes were fairly low and refused to get worked up about it the way Bob Green and Jason Bobe had done. “The only tricky thing,” George conceded, “will be getting the PGP right.”
⁸

“Tricky” was an understatement. From the moment the NIH notified George that it would fund every aspect of the imaging center
except
the PGP, rhetoric and accusations flew back and forth between the Church camp and NHGRI. From what I could glean from these exchanges, the NIH saw George as a brilliant scientist, a world-class biologist, and a sharp and incisive mind that had always rewarded the public’s investment in him. Perhaps even more important, he was a card-carrying inventor and precocious technology developer who had a knack for seeing the future. It would be stupid not to make him a centerpiece of the institute’s efforts to realize the thousand-dollar genome. “We ignore George Church at our peril,” a powerful genome scientist once said. On the other hand, there were parts of the Churchian future that the NIH simply could not stomach. His views about genetic privacy, for example, which he began to spell out in the original 2003 Molecular and Genomic Imaging Center proposal:

For George, the answer to each of these questions was a resounding no. DNA was the ultimate digital identifier, after all: a Social Security number was nine digits, while a genome was 3 billion. George believed the NIH was balking at paying for the project (despite having approved every other aspect of his $10 million genome technology grant) because he refused to do it under the ethical paradigm set forth by the agency—that is, one in which subjects give informed consent and in return are promised, more or less, privacy and confidentiality.

And what about the long list of privacy breaches involving health and genetic information?
¹⁰
To NIH, these might have been noteworthy anecdotes that raised potentially provocative questions about the privacy of genomic data. Such questions were certainly worthy of study. But their answers were in no way self-evident. It was one thing for the Church team to come up with faster ways to sequence human DNA or better ways to interpret what those sequences meant. It was quite another to start sequencing healthy humans, sharing every detail of their genomes with them, and not even
trying
to keep their identities secret.

In early 2007, Jeff Schloss, the National Human Genome Research Institute’s point person on sequencing technology grants, reluctantly granted me a brief telephone interview. His main point was that George had not made the case, either in his 2003 grant application or in three years’ worth of subsequent correspondence, as to why the PGP should be carried out under different rules of informed consent. “We need hypotheses,” Schloss said. “We need to know what scientific ideas he’s going to be testing.”
¹¹

“We’re doing hypothesis
generation,
not hypothesis
testing,”
countered George. “What hypothesis were we testing when we sequenced the first human genome? When he sold the Human Genome Project to Congress, Jim Watson didn’t say we were going to test hypotheses, he said we were going to cure cancer.”
¹²

A few weeks later I tagged along to a meeting of the various ethics centers funded by NHGRI. There an NHGRI administrator gave a presentation outlining the institute’s plan to sequence five to ten humans in the next one to three years. It was never clear to me what NIH hoped to gain from this, as there was no discussion of scale-up or data analysis. Multiple companies had, like George, already begun to sequence humans to demonstrate the power of their technologies. But NIH itself was not in the technology development business—it gave money to academic scientists and start-ups to do that. And clearly NHGRI was not excited about publicly identified genomes. During the Q&A someone asked about the PGP-10 and how it was different from what NIH was proposing. An administrator said that George insisted on sequencing highly educated geneticists and seemed to imply that this was elitist, celebrity genomics. The interlocutor, an ethicist familiar with the PGP, said she thought that the education requirement was something that had been demanded by Harvard. The administrator said no and insisted it was George’s idea. I had interviewed Rabbi Terry Bard, a member of the Harvard Institutional Review Board that oversaw the human subjects aspects of the PGP. He told me that having the PGP-10 be highly credentialed was indeed the IRB’s stipulation, not George’s.
¹³
Yet in the coming months I would hear the same charge leveled at George from several people affiliated with NHGRI.

I paid a visit to the fortress itself. On the NIH website there was a warning about “new security procedures.” I showed up early, which was a good thing. I drove into the car inspection line and queued up, and a succession of orange-vested people approached my window. The first brandished a metal detector wand, which she waved over my steering wheel. The next collected my driver’s license. The third placed a sheet of paper on my dashboard saying that my car had been inspected and would I please open the trunk. The fourth asked me to step out of the vehicle and walk through a metal detector.

I should say that, through the entire process, there was not an ounce of DMV- or airport-security-style surliness. Judging by my experience at the entrance gate, NIH was populated by Shiny Happy Civil Servants. They preempted one’s inclination to get mad at them. They had to do this all day in the sweltering heat—so if they weren’t irritated, how could you be? I returned to my car and a smiling orange vest came back with my license and a badge with my name on it. Then I sat and waited. And waited. And waited some more.

Eventually I was waved through. I found Building 49 and waited for Les Biesecker to come downstairs and guide me over the final security hurdle. Biesecker ran the Genetic Disease Research Branch at NHGRI. He was tall with short gray hair and glasses on the end of his nose. Biesecker was in charge of ClinSeq, which its website described as an attempt to “pilot large-scale medical sequencing in a clinical research setting.”
¹⁴
He told me that he and his colleagues wanted to bring genomics to bear on medical problems. The idea was to find a population with a medical condition and sequence a couple of hundred genes in each patient where there’s likely to be a “hit”: a mutation or mutations that are important in that disease. The National Heart, Lung and Blood Institute bit on this idea: cardiovascular disease would therefore be ClinSeq’s initial focus.
¹⁵

Like the PGP, ClinSeq would return results to participants. But unlike the PGP, it would return them only if they were clinically relevant. This meant 1) known genetic variants that led to known clinical manifestations; 2) variants that appeared to be severe because they stopped a complete protein from being made, they inserted/deleted sequence, or they drastically rearranged the gene; and 3) more ambiguous changes that altered amino acids within a protein but whose effects were not known. “Those will be tough,” said Biesecker of the last category. “The good thing is that for cardiovascular disease, there are interventions, so if there is bad news then there are potentially things that can be done [in those cases].”
¹⁶
For each gene analyzed then, ClinSeq staffers would have to investigate the variants they found and decide if a case could be made for returning results from that gene to subjects.
¹⁷