Hacking the Code of Life

by Nessa Carey

  All the preliminary work in the lab, using human cells and animal models, looked very promising and CRISPR Therapeutics (with their partner company, Vertex Pharmaceuticals) submitted applications to the regulatory authorities in December 2017 to allow them to start testing their approach in adult humans with the disorder. If you look on the European Clinical Trials database, you can see that an application to run such a trial has been submitted,7 and in fact has now been approved.

  A similar application was submitted in the USA and seemed to be moving forwards. However, at the end of May 2018 the companies announced that the Food and Drug Administration (FDA) had put the application on hold and was requesting more information.8 Unfortunately, there has been no public announcement of what their concerns are, so we don’t know what is worrying the FDA. It’s not really surprising, however, as this will be the first large-scale trial of the technology in humans, so there are an awful lot of unknowns that need to be addressed.

  Gene editing moves into humans

  Earlier, more cumbersome versions of gene editing, based on much more expensive and less easily employed technologies, have been used on a very small scale in the recent past. There’s a condition called Hunter’s syndrome in which patients lack a key protein. Because of this, their cells are unable to break down certain carbohydrates. These build up in the cells and cause a range of symptoms that include hearing loss, breathing problems, bowel dysfunction, increased risk of infection and cognitive decline. It’s possible to give sufferers an infusion of the protein that they lack, but this is really expensive, costing between $100,000 and $400,000 a year for each individual. In November 2017, a team from UCSF injected a 44-year-old Hunter’s syndrome patient with the early-generation gene editing technology, carried in a virus vector. The aim was for the virus to reach the liver and release the gene editing machinery into the liver cells. This machinery was designed to insert the gene for the missing protein. If all went well, the liver cells would start producing the protein which would be released into the bloodstream. There was no expectation that this approach would reverse existing damage but the hope was that it would prevent any further progression.

  This human experiment triggered major excitement, and wholly inappropriate headlines such as ‘Scientists see positive results from first-ever gene-editing therapy’.9 But what actually happened was that the researchers didn’t see any major negative effects. The patient, Brian Madeux, didn’t suffer any serious adverse reaction to the gene editing and this has given the investigators the confidence to go ahead and administer it to a second person with the same condition.

  People with Hunter’s syndrome usually die between ten and twenty years of age, so Brian Madeux is quite a clinical outlier with a rather mild form of the disease. His participation in the trial will allow the investigators to answer certain important questions. The researchers involved will be able to assess critical issues around whether they have used a high enough dose; what percentage of liver cells need to be edited to cause detectable increases in the missing protein; how long the edited liver cells will survive and produce the protein; whether they will pass on the functional edit to their daughter liver cells. This will help inform the next trials of this therapy, but whether Brian Madeux’s participation in the trial will benefit him clinically is very much open to question. We often lionise those clinicians and scientists who develop new therapeutic approaches. We should never forget that without patients, many of whom agree to participate in trials more in the hope of helping others than with any real expectation of getting better themselves, there will be no progress.

  We might wonder why the trial went ahead, given that it uses an old form of gene editing that is significantly inferior to the newer technology. The most likely reason is because the company involved, Sangamo Therapeutics, had already invested heavily and for a number of years in this approach. There comes a point when you have already gambled so much money – and make no mistake, drug development is a very high-end form of gambling – that you have no choice but to keep moving forwards.

  You’ve gotta deliver

  One of the biggest challenges in getting gene editing to work in the clinic actually has very little to do with the basic technology of gene editing itself. It’s the same issue that has stymied progress with older genetic therapies, and it’s basically a delivery problem.

  We have all become very comfortable with taking tablets or liquid medicines. Pop a pill and on you go. The problem is that this only works for traditional small molecule drugs like aspirin, or antibiotics, or antihistamines. Large, complex preparations, such as gene editing reagents, can’t be administered via this route. They wouldn’t survive the journey through the highly acidic stomach.

  If you want to distribute something large and complex around the human body you generally have to inject it into the bloodstream. The blood is the transport network of our bodies, moving nutrients, gases and toxins to all the right destinations. Early on in its journey, anything that was injected will reach the liver, the giant decontamination organ. One of the liver’s main jobs is detoxification, breaking down weird foreign stuff before it can do any damage to the tissues.

  The problem is that gene editing reagents can look exactly like weird foreign stuff to the liver. So it gets on and does its job, breaking down these invaders, and the dose that finally reaches the ultimate target tissues is too low to have any effect.

  It’s not surprising that the Hunter’s syndrome trial that uses the older version of gene editing targets a condition where you don’t need the reagents to get further than the liver. In fact, the propensity of liver cells to take foreign materials inside themselves is a definite advantage in this scenario. If scientists can create the right delivery package, the normal function of the liver cells actually helps to unwrap the genetic payload, which then has a decent chance of jumping into the nucleus of the cell and editing the DNA it finds there. If this happens successfully, the liver itself will produce the missing protein and release it into the bloodstream, where it can travel to the target tissues and do its job.

  The trials under development for sickle cell disease and thalassemia also work around the delivery problem, albeit by a different route. The gene editing reagents are delivered directly to the patient’s own cells, but in a laboratory rather than in the body. Once the editing has taken place, the cells can be returned to the body, just as we might perform a typical blood transfusion.

  Although we think of all the tissues in the body being connected and integrated, especially via the blood system, there are exceptions. These are known as privileged sites, which act as independent regions within the great federal entity. This is one of those phenomena that many people are familiar with, without actually realising they know about it. We are all aware that for organ transplants such as kidney, liver, heart and lungs, and most others, it’s important to ‘match’ donor and recipient as closely as possible. Essentially, by ‘matching’ we mean we are trying to find a donor and recipient pair whose immune system identity tags are as similar to each other as possible. This decreases the chances that the transplant will be rejected by the ever-vigilant immune defenders, which have evolved to protect us from foreign pathogens. Even with reasonably well-matched pairings, recipients often have to spend the rest of their lives taking drugs which dampen down the immune protection.

  But it’s a whole different story with corneal transplants. The cornea is the transparent part at the front of the eye. For corneal transplants there is no need to match donor and recipient, and the patients don’t need to take immunosuppressive drugs either. That’s because our eyes are effectively hidden from our immune system. They are privileged. This has almost certainly evolved as a way of preventing dangerous inflammatory reactions in the eye, which risk making us blind.

  Because we don’t have to worry about the immune system attacking foreign agents that we introduce into the eye, it makes this organ a great candidate for gene editing. We can inject the reagents straigh
t into the relevant part of the eye, secure in the knowledge that they won’t be wiped out by an overly-zealous immune response. We also know that the gene editing reagents won’t get out of the eye, so we don’t have to worry that they will enter the wrong tissues and perhaps carry out editing elsewhere.

  Experiments in human cells and in animal models have already shown that gene editing works in the cells of the eye. In theory it should be possible to use this technology to stabilise and even reverse various forms of blindness. These can include the types caused by genetic mutations such as retinitis pigmentosa, or even the age-related conditions that affect the general population, including macular degeneration.

  The gene editing company Editas Medicine was moving towards clinical trials of this approach to cure a genetic condition called Leber’s congenital amaurosis. This is a common form of childhood blindness in which significant vision loss occurs before the affected child is even a year old, eventually progressing to blindness. Editas Medicine planned to treat this by using gene editing to remove the mutation that causes this disorder, via direct injection into the eye. In a setback, however, the company had to delay its plans to submit their trial for regulatory approval.10 There doesn’t seem to have been anything wrong with the gene editing approach per se, but the company was having problems manufacturing reagents of high enough quality and at sufficient scale for clinical trials. It’s perhaps no surprise that they have now teamed up with a more established company called Allergan who are a lot more experienced at the outwardly mundane but actually critical functions needed before you can really claim to have a therapeutic that’s good enough for use in humans.

  Look east

  While regulatory authorities in the USA and Europe take an understandably cautious approach to moving gene editing into humans, things are moving rather faster in China. There are claims that around 100 patients have been treated in Chinese hospitals using the most advanced forms of gene editing.11 The problem is that this statement is based on claims made by doctors to western journalists. No scientific papers or clinical reports have been published, so it’s hard to know which conditions were targeted or whether any improvements or stabilisation in disease were achieved using the new technology.

  Why is China further ahead than the rest of the world in trialling this technology? It’s hard to be sure, given the lack of detail, but some of the reason is almost certainly a medical culture that is less risk-averse, and with a lighter regulatory touch. Where you stand on this question is probably dependent on your own position. If you have a life-ending/life-limiting condition that is otherwise untreatable you might want to have access to new approaches sooner rather than later. On the other hand, reduced regulatory oversight isn’t such a great thing if the proposed approach is flaky.

  It’s not clear why the Chinese scientists and clinicians aren’t publishing their methods and clinical results in the medical literature. However, some of this may be a consequence of the apparently lower levels of regulation in China. If the trials don’t meet the ethical standards of the western authorities, many journals outside China will be reluctant to publish the papers describing the interventions.

  There may also be a pragmatic reason for the lack of publications. The technologies that underpin gene editing are incredibly valuable and the organisations that created them want to protect the intellectual property very aggressively. They will expect to be paid large amounts of money if anyone uses their breakthroughs commercially, and in China a great deal of medicine takes place in the context of a private healthcare system. If you don’t publish the details of your gene editing treatments, it’s very difficult for anyone to sue you for infringing their intellectual property.

  Whatever the reason, it’s such a shame that so little information is coming out of China. Sharing this information openly would almost certainly speed up progress globally for the benefit of patients. It would give everyone a better heads-up on what is effective and what isn’t and what the safety risks are, if any.

  Notes

  1. https://www.buzzfeednews.com/article/stephaniemlee/this-biohacker-wants-to-edit-his-own-dna

  2. https://www.insidescience.org/news/Alzheimer%27s-Drug-Trials-Keep-Failing

  3. http://www.who.int/bulletin/volumes/86/6/06-036673/en/

  4. For a historical overview from the person who led this research, see: https://iubmb.onlinelibrary.wiley.com/doi/full/10.1002/bmb.2002.494030050108

  5. https://www.cdc.gov/ncbddd/sicklecell/data.html

  6. http://www.ema.europa.eu/ema/index.jsp?curl=pages/medicines/human/orphans/2011/03/human_orphan_000889.jsp&mid=WC0b01ac058001d12b

  7. EudraCT Number: 2017-003351-38.

  8. http://ir.crisprtx.com/news-releases/news-release-details/crispr-therapeutics-and-vertex-provide-update-fda-review

  9. https://nypost.com/2018/02/06/scientists-see-positive-results-from-1st-ever-gene-editing-therapy/

  10. http://ir.editasmedicine.com/phoenix.zhtml?c=254265&p=irol-newsArticle&ID=2273032

  11. https://www.wsj.com/articles/china-unhampered-by-rules-races-ahead-in-gene-editing-trials-1516562360

  * CRISPR is the technical term for the most advanced type of gene editing.

  6

  SAFETY FIRST

  Clearly, medicines regulators have concerns about the safety of gene editing. Or at least, western ones do. But we shouldn’t take this as a sign that the technology is inherently dangerous. The very first hurdle that all new drugs have to jump over is the hurdle of safety. If a drug isn’t safe, it’s unlikely to be allowed onto the market.

  Of course, safe is a relative term. Safety has to be balanced against benefit. If the drug is something you will buy from a pharmacist to treat hay fever, the regulator isn’t likely to look favourably on it if the side effects commonly include nausea, vomiting, extreme fatigue and hair loss. On the other hand, if the drug is the only option to save the life of a person with an otherwise incurable cancer, the regulator may decide that these deeply unpleasant, but not fatal, side effects are an acceptable compromise.

  Pharmaceutical companies are actually reasonably adept at identifying and stopping the development of drugs that would likely cause large-scale safety problems. There is no point running highly expensive clinical trials if your laboratory results have already indicated that your new drug will be rejected for having an inadequate safety profile.

  The problem is that the more innovative a new therapeutic is, the less we can predict about its safety. It might do something so extraordinary we simply never could have predicted it. A bizarre example of this is the increased risk of narcolepsy in children and young adults who received a specific flu vaccine in Europe in 2009.1 It’s not at all clear what has led to this association, and there’s no way anyone could have identified it as a risk before the vaccine was used in large numbers of people.

  But even for something as new as gene editing, researchers and regulators can take a logical approach to the risks they want to assess. Gene editing is essentially a way of introducing changes into DNA. The reason why scientists have embraced the version of gene editing that was first reported in 2012 so enthusiastically is that it is more precise than any other approach developed in the past.

  The whole field got a bit of a fright in 2017 when a team from Columbia University published a paper claiming that gene editing in mouse cell lines introduced hundreds, if not thousands, of unexpected mutations in addition to the intended one.2 This was hugely worrying, especially as the technology was moving closer to clinical applications. But within a year, the panic was over. Other researchers demonstrated that the original experiments had been poorly designed, and the conclusions were flawed.3 To their credit, the original team revisited their work and conceded that the criticisms were valid. The paper was subsequently retracted, although only two of the six authors agreed with this decision.

  There’s been quite a lot of criticism of the editors of the journal that published the original paper. Here, for example, is a scathing statemen
t from a professor at the Australian National University: ‘I find it absolutely astonishing this paper got published in Nature Methods. This is a terrible paper and as a reviewer I would have dismissed it from the first round of review. This is a worrying trend from “high impact” journals to promote the hype over good science. The publication of this paper is clearly a failure in the peer review process.’4

  You might wonder why scientists protested so much about the original paper and its erroneous conclusions. After all, the correction process that we associate with scientific research seemed to work. A paper was published, other researchers had concerns, the situation was rectified.

  But there are good reasons for researchers to be concerned at what they see as a drop in standards in the scientific literature. Some of the protests came from companies developing gene editing for creation of therapies. Because the original publication got a lot of exposure in the popular press, the share price of these companies took quite a hit. Companies working in new technologies are often at the vanguard of developments, so it’s deeply frustrating for them when their investment position is compromised over false issues.

  Another problem is that retracted papers don’t disappear. You can test this for yourself. Put the details of the offending reference into an online search engine and you’ll get lots of hits that refer you to the paper but where there is no mention of it having been retracted. So this kind of problematic publication continues to muddy the academic waters.

  The other problem is that research that gets the science badly wrong, but that somehow taps into a zeitgeist that distrusts technology, can be hugely damaging. In 1998 a scientist called Árpád Pusztai, working at a research facility in Scotland, claimed that rats which had been fed genetically modified (GM) potatoes were stunted and their immune systems were suppressed. He made this claim on a TV programme before any of his science had been peer-reviewed.