Barry Behr is assistant professor of obstetrics and gynecology and director of the in vitro fertilization (IVF) laboratories at Stanford University and consulting laboratory director for Huntington Reproductive Centers. He spoke with associate editor Sarah Graham. An edited transcript of their conversation follows:

SA: What is the process used for freezing embryos?

Cell freezing has been around for decades. The major concern in cryopreservation of cells is the possible formation of ice crystals inside in the cells. You can imagine they are like little razor blades inside a balloon: ice crystals have sharp corners and little points. Thus they can damage a cell's membrane or the membranes of various cell structures that exist inside the cytoplasm of cells. This is true for all cells, whether they be embryos, steak, fish, fruit or vegetables. In fact, most people can experience the effects of cryopreservation when they taste a piece of fish that's been frozen versus fresh fish. It's usually tougher, it doesn't taste as good and it just has some negative attributes that distinguish it as not fresh. And that is due to cell death. The natural structures that exist either between cells or within cells are damaged during the freeze-thaw process and hence the flavor or texture of the tissue is altered. It's for those reasons that skeptics such as myself have voiced a concern about whole organ freezing because it's very difficult to apply the protective material that cells need to survive to large tissues or large organs or body parts.

Which brings us to egg, embryo or sperm freezing. The most advanced embryo that we freeze is between 100 to 120 cells, which is known as a blastocyst. It is around 120 microns, or about a tenth of a millimeter, in size so it is very conducive to freezing and using cryoprotectants. Cryoprotectants are basically antifreeze that we add to the solutions in which the cells are being frozen in to protect them from membrane damage and ice crystal damage. They are designed to both permeate the cells, meaning to get inside the cell, and to displace water to prevent intracellular ice crystal formation. They have a second function of stabilizing the membrane and protecting it from damage during cryopreservation. And thirdly, they provide a "hyper-osmotic" environment that helps the process of dehydration, which draws the water out of the cells. This process is accomplished by using cryoprotectants composed of typically large sugar molecules that make a more concentrated solution around the cells, which by osmosis and diffusion causes water to move out. Simultaneously, cryoprotectants that are made up of smaller molecules such as ethylene glycol or glycerol are able to permeate the cell so it doesn't shrink up like a raisin. Instead it can maintain its three dimensional structure to a degree but not be filled with water. Because our cells are made up of mostly water, if we dehydrated them to the point of having no fluid, they would be damaged by so-called solution effects, which result from the cell being too concentrated and not having its water replaced by some other compound.

Dehydration is one of the key steps in cryopreservation. It is important to keep in mind that why this process is so successful for embryos and sperm is because of their size and their relative low number of cells. A liver or kidney has billions of cells and a large three dimensional structure. A tenth of a millimeter is relatively small, in contrast, although it's still big on a microscopic level.

There is an emerging technology called vitrification that differs from the traditional slow cooling or slow freezing cryopreservation. The traditional method requires the sequential addition of cryoprotectants over a series of 10 to 20 minutes and then an approximately two hour process that cools the cells at about 0.3 to two degrees a minute down to -196 degrees Celsius, which is the temperature of liquid nitrogen. The reason the cooling is done slowly is to allow the permeation of the cryoprotectants and the appropriate dehydration of the cells to occur in a manner so that no intracellular ice crystals form.

Vitrification, however, is a technique that uses much, much higher concentrations of the cryoprotectants--about three to four times higher. But the cooling rate is 10,000 times faster. So we are cooling things at a rate of 10,000 to 20,000 degrees a minute by essentially plunging them directly into liquid nitrogen. The rapid cooling is necessary to do two things: prevent the toxicity of the high levels of cryoprotectants at room temperature and achieve vitrification, which means glass formation. By definition, vitrification is solidification through increased viscosity. So instead of going through a phase change from liquid to solid, glass formation occurs, which allows cells to be preserved in their existing state without going through the complete dehydration that you achieve in slow cooling. But it does have some problems. For one, there is concern about the high concentration of cryoprotectants needed. Most importantly, the only way to get the cooling rate required is by having the cells in direct contact with liquid nitrogen and liquid nitrogen can be a vehicle for disease transmission. Vitrification is used quite a bit in Japan and by a few places in the U.S. but there are currently no FDA-approved vitrification systems or media in this country. In contrast, there are many traditional cryoprotectant solutions that you can buy from vendors that are FDA-cleared. So although vitrification is a bit problematic and hasn't yet provided superior outcomes, it is much superior on the time frame. It also doesn't require expensive equipment that can cool chambers at a third of a degree a minute. You need a $10,000 - $20,000 dollar machine to do that versus what is essentially a Styrofoam bucket that you can pour liquid nitrogen into. It has some theoretical advantages but it's not out of the woods yet.

SA: How long does the traditional cryopreservation process take?

The standard method takes about two hours. Negative 196 degrees C. is close to absolute zero, but it's not absolute zero, which is the theoretical temperature at which no biological activity occurs. It is certainly close enough to stop biological activity and allow babies being born from frozen embryos that have been frozen for up to 14 years. Fourteen years is the longest--that I am aware of--for an embryo being frozen that resulted in a live birth.

SA: Are there any limits on how long embryos can be stored?

It is theoretically much longer than [14 years]. It should be a century or two, even, though it's not unreasonable to think there would be some degradation over time. We've learned a lot about cryopreservation since the first embryo freezing around 20 years ago. The embryos that were frozen back then could have all been used up--people don't usually wait 14 years to have kids. The nature of reproduction and building families does not really foster an environment that would allow you to wait 50 years before you thaw your embryos out. Unless your daughter wanted to carry her sibling, for example, which in theory is possible: a person born from IVF may still have sibling embryos frozen and when they reach 30 and are infertile could technically thaw out the embryos that were created at the same time they were created, gestate them and deliver their sibling. That hasn't happened yet but it could be possible.

SA: How are the embryos treated prior to use or implantation?

It's basically the reverse process. The key factors that must be achieved on thawing are rehydration of the cell without blowing it up. We warm them in air and in a water bath instead of in a machine and the warming rate is less precise and much faster than the cooling rate. Embryos are moved through a number of steps to slowly dilute out the cryoprotectants that were added when they were frozen. It's a stepwise dilution of the cryoprotectant as the cells are warmed first to room temperature and then to body temperature. They are then basically allowed to rest and are usually implanted within two to four hours of reaching body temperature.

As a note, successful sperm freezing was done long before any other type of cell freezing. Sperm can be frozen easily because they have no cytoplasm and therefore aren't susceptible to all the types of damage previously mentioned.

SA: What are the numbers on success rates of IVF with frozen embryos?

The first baby from a cryopreserved embryo was born in 1986. The success rate of IVF with frozen embryos is about two thirds of what it would be with fresh embryos. It's difficult to give a definitive number, because it depends on the raw materials you are dealing with. The pregnancy rate is determined primarily by age so if you're 40 your chances are much worse that what they were at 30, for example.