Thursday, March 28, 2013

DNA scissors promise permanent fixes (cures?)

Many people face a routine of pills, shots, ointments or eye drops that will go on and on for the rest of their lives. If they can remember and stand to keep taking them, such treatments make temporary, biochemical changes in cells to counter chronic diseases, but then wear off and must be taken again.

What if doctors could make one-time, permanent changes in patients' genes that had the same benefits as their pills?  Along with eliminating the "pill burden," one-time fixes would eliminate billions of dollars in healthcare costs by rendering unnecessary the most expensive and profitable kind of drug: one that must be taken daily for life.

The promise of permanent cures through gene therapy was first envisioned in the 1950s when Nobel laureate Arthur Kornberg and his team at Stanford discovered the enzymes that enabled researchers to manipulate DNA chains for the first time. That lead to a better understanding of how genes work, and of how to clone them.

Among these early discoveries was that DNA-editing enzymes called nucleases cut DNA chains at certain spots. Nucleases are in the spotlight again as part of "audacious" plans to permanently cure diseases. In combination with newer, more precise DNA editing systems like CRISPR, discovered in an effort to make a better yogurt, researchers are closing in on the ability to cut DNA exactly where they want to, say at either end of a disease-causing snippet of genetic code in need of removal.

Steven Pittler, Ph.D., director of the UAB Vision Science Research Center, sat down with The Mix to talk about his efforts to do just that in eye diseases with genetic components. He is among the winners of the National Eye Institute's recent "Audacious Goals Challenge," which challenged Americans (not just vision scientists) to come up with one-page research ideas that are not possible yet, but that could be realized with some planning. His winning idea: let's use molecular scissors, a funky term for nucleases, to edit genes inside eye cells and cure ocular diseases.

Show notes for the podcast:

0:58 The NEI challenge asked all comers to finish sentences like "if this idea were to come true, it would mean XX for the field of vision research."

1:52 As opposed the typical call for in-depth grant proposals from researchers, the challenge was blinded.  Those reviewing the proposals had no idea whether a given idea came from a scientist or a lay person. It turned out that all of the winning proposals came vision scientists, proof that they are a creative group.

3:00 Dr. Pittler's winning idea was to use "molecular scissors" to cut out disease-causing snippets within genes particular to the eyes combined with other steps that replace them with healthy DNA sequences. Such scissors represent a set of enzymes familiar to scientists called a DNA nucleases or restriction enzymes. Such enzymes cut DNA at certain sport in the DNA chain only where a certain sequence of bases, the units that make up DNA code, are present.

3:54 Doctor Pittler and others seek to build on the historic early use of nucleases to achieve highly precise cuts in DNA on the way to making precise adjustments in the code. Recently, the field has been exploring the use two kinds of nucleases, zinc fingers and TALENs, that enable much more precise editing and  replacement of disease-causing genes. Researcher are now coupling nucleases such that the combination targets only a specific sequence they want to cut, rather than making 100,000 cuts too many like older systems.

5:20  All nuclease systems are useful because they recognize a specific sequence of DNA code and cut the chain there. They all use an enzyme called FOK1, which can be split into two parts, and that only cuts the DNA chain when, under circumstances carefully controlled by researchers, the two parts come back together. That quality has allowed for the design of artificial nucleases that do specific jobs as part of gene therapy. Zinc fingers, TALEN nucleases and CRISPR all work on similar principles, and newer DNA editing technologies may make possible efforts to address diseases caused by multiple genes.

7:04 Viruses are designed perfectly by evolution to invade our cells and use our genetic machinery to make copies of themselves. Thus, an early approach to gene therapy involved taking a virus, stripping it of its disease-causing elements and changing it such that it delivered a useful gene sequence into cells to counter disease. The problem is that these changes are "extra-chromosomal," so disease-countering changes occur in one generation of cells, but not in their descendants. Thus, virus-delivered gene therapies may correct a genetic disease temporarily, the but the effect fades as the treated cells die and are replaced in the constant turnover underway in many tissues. Dr, Pittler seeks to develop a technique that would alter DNA at the chromosomal level to bring about permanent cures.

8:41 Eye diseases that Dr. Pittler would like to help cure include retinitis pigmentosa (RP), an inherited eye disease that causes severe vision loss in 100,000 people in the United States. Also of interest is macular degeneration, often seen with aging and causing vision loss in 1.75 million affected Americans. One of three people will have macular degeneration by the time they are 75.

9:15 Curing such diseases will involve changing genes and related proteins at work in the specialized cells that make vision possible. RP, for instance, involves problems with the retinal pigment epithelium, a layer of cells required for the function of the retina. One its most important functions is that it engulfs and destroys old photoreceptor cells as they are replaced by new ones in the turnover that maintains the integrity of vision.

10:20 Photoreceptor cells absorb light and process color via chemicals groups called chromophores within the protein rhodopsin. When light is absorbed by the chromophore, it changes shape, which changes the shape of rhodopsin, which triggers chemical signals that are processed into images by the brain.

11:16 Dr. Pittler attended a recent NEI meeting where eye researchers from lead institutions nationwide gave input into the plan that will shape federally funded eye research for the next ten years. The next step for the institute will be to formalize a plan and presumably, direct resources toward its goals. While he waits for that plan and the chance to compete for grants, Dr. Pittler is looking to create funding sources to keep the research moving toward clinically useful, genome-based medicine. One solution may be to found a company that uses nuclease technology to genetically engineer mice to the specifications of client labs nationally.

13:54 One step toward the founding of new field is developing mice with genetic changes that accurately represent human diseases. For instance, researchers have identified hundreds of small changes, called mutations, in the gene for rhodopsin. Only a few of them have been introduced into mice and studied to identify their contribution to healthy vision or disease. Pittler;s envisioned company may even be able to manufacture cassettes, entire replacement genes with only the mutations desired by each client present.


Friday, March 15, 2013

CMV: we all have it but some babies lose hearing

Most of us are infected with it at some point with no symptoms, but that's little comfort to the babies born with a symptomatic version, some of whom lose their hearing. It's cytomegalo virus or CMV, a common virus that infects people of all ages.

This relative of herpes viruses is mostly a problem for those whose immune systems are weak, including the old, people with HIV and those taking drugs that knock down immune defenses, as well as for fetuses and infants.

A mash up of studies puts the estimated number of babies born each year with a congenital infection between 20,000 and 40,000 (.5 percent of all births). Most children are fine, but a few lose their hearing, and even fewer suffer birth defects or die.

Shannon Ross, M.D., associate professor in the Department of Pediatrics within the UAB School of Medicine, last year won a grant from the National Institute on Deafness and Other Communication Disorders (NIDCD) to try to figure out which traits of which versions of CMV match up with symptoms. She talked with The Mix about how having a marker to tell which babies are most at risk would help researchers decide when treatment with antiviral drugs, which come with serious side effects, is worth it.

Show notes for the podcast:

1:19 Between 20,000 and 40,000 babies in the United States, .5 percent of all those born in a year, are estimated to be infected with CMV.  Many studies have looked at this number, but each was conducted in a small group with its own unique characteristics in terms of mother's average viral level, economic status and region.  So the range is a combination of those estimates.

2:10 Of that 20,000 to 40,000, about ten percent are born showing symptoms of their viral infection, that is obviously ill from it. The babies may have petechiae, small spots on their skin caused by broken blood vessels. They may have an enlarged liver or spleen thanks to the infection, or be relatively small for how long their mothers carried them.

2:45 The field does not know why some babies show symptoms at birth and other don't. One theory holds that it might have to with when the baby was infected during the pregnancy, early or late.

3:13 Of all babies born with congenital CMV, about 10 to 20 percent go suffer for some form of hearing loss, with the risk varying based on whether or not the baby was born symptoms. Of those born with symptoms, perhaps 40-50 percent will have hearing loss. For the majority that are born without symptoms, about 10 percent go on to have hearing loss.

3:57 Dr. Ross recently won an NIDCD grant to study whether or not certain characteristics of the CMV virus can be linked to hearing loss.  In particular, they are examining when any certain strain of the virus, or combination of strains, is more closely associated with hearing loss.

4:35 While the question requires further study, evidence suggest that infected children might have up to five different strains of the virus in one infection. This idea is based on the results of past, small studies, but bigger ones on strain variability are underway.

5:23 Dr. Ross is hoping a strain or combination of strains can be identified as a marker for both whether or not a baby will have any symptoms, and whether or not that child is at risk for hearing loss.

6:06 It may be that a certain gene or version of a gene in a certain viral strain will hold the clue for how CMV causes disease, offering new clues that guide future treatment.

6:42 There has been a lot of interest in screening all babies for the virus because it's a common cause of hearing loss, said Dr. Ross. But for the many children born without symptoms, there is currently no way to determine which is at risk. Physicians don't know which children in the non-symptom group need to be watched closely for an early intervention. They don't know which kids need to get hearing AIDS to preserve their ability to learn to speak

7:48 Complicating matters is the availability of an approved drug for the treatment of CMV called ganciclovir. The problem is that the drug comes with considerable side effects like neutropenia, a loss of white blood cells, so you can't just give it to every infected baby without thinking twice. It becomes a balancing act for doctors between a child's risk for hearing loss versus exposing them to the side effects that could increase their risk for other infections.  For that reason, the field has chosen to treat a subset of babies born with symptoms, but not those born without them. No one will even study the question of whether ganciclovir can preserve hearing in non-symptomatic children, said Dr. Ross, until a marker points to which ten percent of them are at greatest risk.

9:02 Theories for how CMV infection might cause hearing loss include that just the virus making copies of itself may damage the inner ear, or that our body's immune response to the CMV may be the culprit.

10:11 When children lose their hearing from CMV, it can happen at birth or between ages one and two. The latter, delayed onset hearing loss happens when children are learning to talk, and has an additional cost in terms of learning and developmental delays.

11:15 Unlike HIV for instance, there is no good animal model for the study of CMV. There are many animal versions of the virus, but they do not react to CMV like human cells do. Plus, the animal versions do not cause congenital infections like they do in humans. That has made it very hard to understand the basic mechanisms of the disease (e.g. how the virus reproduces or infects cells).

12:13 Dr. Ross and colleagues are now doing "deep sequencing" of the genes in each strain of CMV virus taken from blood or urine of babies with congenital CMV infection.  They can then examine the entire gene code of each "quasi-species" (strain) of virus looking for genetic variations between strains that may contribute to the presence of symptoms. The team is now sifting through large amounts of data looking for patterns that could establish a marker.

Monday, March 11, 2013

What others missed about heart failure

Every research medical center markets itself as interdisciplinary, with brochures proclaiming that University X is breaking down barriers between research specialties to solve medical mysteries.

Perhaps it started in 2002 when Elias A. Zerhouni, M.D., then newly appointed director of the National Institutes of Health (NIH), began crafting his "road map" for the agency. He concluded that the quickest way to deliver more breakthroughs would be to put multidisciplinary teams on the case, and with the institutions best suited to do so winning more grants. I could argue that UAB has done this better than most, but that's another story.

My point is that I have a high threshold for getting excited about the "interdisciplinary" aspect of anything. It has to be "real," and I think a study, led by UAB's Ali Ahmed, M.D., and making news today, fits the bill.

The study found that an old heart failure drug digoxin, used less and less since it “failed” its clinical trial 16 years ago, may do something no drug has achieved since: dramatically reduce the chances that heart failure patients check into the hospital within 30 days of first taking the drug.

Who cares you say?  The nation's hospitals and Medicare care a great deal. Medicare penalized thousands of hospitals millions of dollars last year for admitting too many heart failure patients too often. The agency leveled the penalties with the rationale that many admissions and readmissions are preventable if care is handled properly, and because frequent admissions cost the agency $17 billion each year. Digoxin is known is reduce scary symptoms of acute heart failure like shortness of breath that send people racing to emergency rooms.

The current study is a re-analysis of the 1997 DIG study in which digoxin failed to lower the risk of death (all-cause mortality) in patients with chronic heart failure. The treatment did reduce the risk for hospitalization, but that fact was largely overlooked. After it failed to reduce mortality, digoxin went from one of the most prescribed heart failure drugs to an afterthought. In the meantime, newer drug classes like beta blockers and aldosterone antagonists came into vogue after successfully reducing both mortality and long-term hospitalization in pivotal trials. And yet, they don't appear to fully address the acute symptoms driving readmissions.

The current analysis found that digoxin could reduce admission for heart failure within 30 days of first taking it by 34 percent. But what made Ahmed check on whether it could do this in the first place? Marketing effort aside, it looks like it was an interdisciplinary approach. Let's look at his titles.

Within the School of Medicine:
  • professor in the division of Gerontology, Geriatrics, & Palliative Care
  • profession in the division of Cardiovascular Disease
Within the School of Public Health 
  • professor in the Department of Epidemiology
  • director of the Advanced Illness, Multimorbidity and Heart Failure (AIM-HF) program in the UAB Center for Aging
Comparing problems faced by heart failure patients, understanding the mechanisms of drugs old and new,  and sifting through public health and policy trends, Dr. Ahmed and his team were able to frame a useful question, and answer it - potentially and in part, with a solution overlooked by the field.

Here is some media coverage of the digoxin ACC presentation:

Thursday, March 7, 2013

HIV "cure" might fill a gap

It's strange when the discovery of a "cure" points to another, some say even greater, achievement. For instance, the news this week was that researchers used aggressive drug treatment to "cure" a Mississippi baby born with the AIDS virus. We had a quick talk about the story with Michael Saag, M.D. co-director of the UAB Center for AIDS Research.

The baby's mother, living in a poor, rural community, had been infected and passed the virus to her fetus, but had received no prenatal care before she went into labor. So the baby was born before anyone knew it was infected. As in many stories of discovery, what happened next included some accidents.

The hospital was out of the usual treatment, so the baby was referred to doctors that decided to use a high-dose, three-drug combination for older children instead of the usual low-dose treatment recommended for babies. They also administered the drugs quickly after birth.

After 18 months on the treatment, the baby and its family dropped out of the system, and the child went on an unplanned "treatment holiday." When the family reappeared, doctors realized that the child had been off the medications for a year but had no signs of infection.

As amazing as this case is, what jumps out at me is its rarity. In nearly all cases, HIV-infected mothers in the United States are tested and treated months before giving birth; nearly all of their babies are born with undetectable viral levels (functionally cured). In that light, this week's cure, should it be confirmed, would apply to the very few HIV positive babies in the U.S. that are not treated before birth. Somewhat lost in the excitement over a cure for the few exceptions is the massive achievement by the U.S. system of virtually eliminating mother-to-child HIV transmission.

"This case goes to show that once drugs are invented, as complex as that is, the field then faces the even more daunting task off getting them to everyone who needs them, and using them in the best way for each group of patients," said Saag. "That said, the new technique may fill a vital gap if it reduces viral load to almost zero in babies born to mothers that have slipped through HIV surveillance systems."

This week's discovery could have the most meaning outside of the United States, said Saag. The poorest African countries, for instance, have high infection rates, and only about 60 percent of pregnant women get treatment that can keep them from passing the virus to their babies.

Moving forward, Saag said, it is important for the field to understand why the baby was cured. Early on in exposure to the HIV, the infection either takes or it does not. If it takes, the virus goes into a latent state where it hides and quietly multiplies indefinitely, weakening the immune system while it grows strong. Once a pool of latent virus is present, it is harder to kill enough virus with antiviral drugs to get a cure. In the Mississippi case, Saag said it looks like strong, quick treatment after the initial infection may have prevented the establishment of a latent pool of infected cells.

As explained in this article in USA Today, it will take some time before the global impact of this week's cure is understood. In the best case scenario, babies not treated during pregnancy may receive more aggressive treatment at birth to spare them the burden of taking so many medications for a lifetime.