Once the piece of tissue has been fixed it is necessary to remove all of its water, a step referred to as "dehydration". The reasons for this are quite straightforward. The final goal is to have a piece of tissue with paraffin wax intimately surrounding the proteins, and paraffin wax and water do not mix in any proportion, so one of them has to go. That one is the water.
Note that it is not just a case of having the tissue encased in paraffin wax, the proteins themselves have to be both permeated with, and surrounded by, the wax. Wherever there were water molecules in the original fresh, unfixed tissue there must now be hardened paraffin wax, right down to the subcellular level.
The most efficent method of doing this would be to use a reagent that was miscible with both water and wax. Unfortunately, although a few such fluids do exist, they are not as effective as a two stage system. In the first stage water is removed with a dehydrant, in the second stage the dehydrant is replaced with a fluid miscible with wax. The reason for preferring a two stage process is primarily due to the presence of fats, triglycerides, which are miscible with paraffin wax but soften it. This softening causes difficulties with sectioning and has to be avoided, so the fats have to be removed. This is one of the functions of the post-dehydrating fluid.
Almost any water miscible, anhydrous fluid can be used as a dehydrant providing that it does not damage the tissue proteins and is also miscible with the fluids to be used subsequently. Cost may also be a factor.
All of the commonly used fluids are inflammable and suitable precautions must be taken. There should be fire fighting equipment readily at hand, and it should be placed in an area that is accessible should a fire happen. Equipment mounted on a wall just outside the room used for processing is a convenient place. The equipment should include a fire blanket and fire extinguishers suitable for use on a solvent fire. The advice of a fire marshall should be taken with regards to the type of equipment and its placement. If there is a fire that is beyond the capabilities of simple equipment, leave the area immediately and call the fire department. Do not take chances with your life.
In some jurisdictions it is possible to purchase recycling equipment for these reagents. These are fundamentally stills and the fluid is repurified by distillation. In the case of ethanol it is customary to consider the distillate as having 96% purity due to contamination by the water removed from the tissues during use. It can be used for most purposes, but may contain unidentified contaminants from the tissue and other sources, depending on what use was made of the ethanol initially.
Dehydration is accomplished by simply diluting out the water until it is reduced in amount to below the tolerance level of subsequent fluids, i.e. the clearing agent and the molten wax. There is little published data in this area, and the tolerance levels of the commonly used fluids are not readily available, although it is very low. It is also not an absolute, by which I mean that the presence of moisture above the tolerance levels does not stop clearing and infiltration altogether, it makes it less efficient and inhibits complete replacement of subsequent fluids. This, in turn, causes sectioning, block stability and block storage problems. In practical terms, it is advisable to dehydrate as thoroughly as possible, to the stage where water can't be detected at all, i.e. to the stage where there is no moisture present at all.
In the comments below I shall refer throughout to the use of ethanol as the dehydrant, mostly because it is by far the commonest in one form or another. The underlying points being made apply as well to other dehydrants.
During the last 20 years or so, particularly in North America, there has been a move to process tissues very rapidly so that same-day diagnoses may be given. The tissue is first given minimal and inadequate fixation in formalin, which may be warmed in an attempt to fix more rapidly. Then the tissue is immersed in ethanol, also often warmed, for the minimum time possible to remove sufficient water to get the tissues through the clearing agent and into wax. As the tissue was not properly fixed initially, this warm ethanol completes the fixation making the tissue somewhat brittle in the process, as ethanol fixation is prone to do. This inadequately fixed and dehydrated tissue is then placed into xylene or a xylene substitute, which often have lower clearing efficiency than xylene, and again left for the minimum time to remove enough ethanol for the tissue to go into wax. Once in the wax it is left for the minimum time necessary to permit a block to be made, barely.
This approach does not produce properly processed blocks. In many cases there is incomplete removal of reagents at each of the steps. While sufficient water may have been removed by the dehydrant for the clearant to penetrate, a small residual amount remains which is greater than the tolerance level of the clearant and embedding waxes. Due to this presence of moisture, clearing is not complete as it has been inhibited by the residual water, and has not been applied for long enough to remove all of the dehydrant. Similarly, the wax does not completely replace the clearant due to insufficient time of infiltration and could not replace the residual moisture and ethanol in any case. Consequently the wax block contains both moisture, ethanol and clearant, all which inhibit proper hardening of the wax within the tissue and affect sectioning characteristics.
The results of minimalist processing are difficult to section tissues that crack and shatter and which shrink back after a few days as the incompletely removed reagents evaporate. The solution often applied is to soak in water to moisten the tissue, or to reduce dehydration even further in an attempt to retain moisture. The true resolution is to properly fix the tissue in the first place, adequately dehydrate and clear, and infiltrate long enough that the tissue fibres are properly permeated with wax. Minimalist processing is a classic example of "more haste, less speed".
There are, of course, degrees of inadequacy in this kind of processing, which ranges from frankly raw tissue to almost properly processed tissues with just a slight opaqueness to them. It is justified by the argument that the need to provide a fast turn around time for the patient's benefit justifies processing minimally even if doing so produces substandard sections. The other side of that argument is that biopsy material is usually not replaceable so care has to be taken with it to ensure that the quality of the final sections is sufficient to enable a diagnosis to be made. Ultimately, in a diagnostic setting, the pathologists will determine what is adequate since they bear legal responsibility for the results. In a research or educational setting there is rarely any need for such speed and more conventional practices can be used.
It is inadvisable to transfer fixed tissues directly from water or aqueous fixative directly into absolute ethanol. Doing so causes a rapid removal of water which can distort the appearance of more delicate cells and structures. It is advisable to remove water gently and allow the tissue to slowly adjust to its removal. The more delicate the tissue, the more gently this should be done, but there is no hard and fast rule.
In a routine surgical pathology laboratory the initial ethanol is often about 60% or 70%, whereas more delicate embryonic tissues may be started at about 50% ethanol. It is unusual to begin at much less than that, although there would be no harm in doing so in a particular case if the morphology is a concern.
If using neutral buffered formalin as the fixative it is important to remember that the buffer salts may precipitate into ethanol if the initial concentration is too high. This causes difficulties with sectioning as gritty material may have been deposited within and around the tissue. The initial ethanol should be about 60% in such cases, and the retort of automatic processors must be flushed with plain water on a regular basis to avoid concentrating crystals.
In the case of non-aqueous fixatives or those containing ethanol, the initial concentration of ethanol should be not less than the concentration in the fixative, and perhaps 10% greater. It should not have a concentration less than that of the fixative. Tissues fixed with anhydrous fixatives should be placed directly into absolute ethanol. Breaking this rule usually causes no real harm, it is just a waste of time to dehydrate in a fixative then increase the water content with the dehydrating fluid. On the other hand, the possibility does exist that increasing the water content may cause some alcohol insoluble constituents to dissolve.
Number of changes
A very basic question is, "How many changes of ethanol should there be?" This depends to some extent on the amount of dehydrant available, but the answer is usually, "As many as possible."
In most processing setups, whether manual or automated, there is a limited volume of dehydrant available due to the container size etc, and a set volume of tissue slices, perhaps in cassettes which must be covered by the dehydrant. The most efficient use of dehydrants is to use repeated, relatively small changes rather than a single large volume. This is easily seen.
If tissue contains 10 mL of water and we place it into 100 mL absolute ethanol, after equilibriating the tissue would still contain 10/110 of the initial water, i.e. 0.9 mL or 9%. If we had split the ethanol in two and placed the same tissue into 50 mL ethanol, after the first change the water content would be about 1.7 mL or 16%, and after the second change it would be about 0.3 mL or 2.8%. That is, by using two changes the water content has been reduced more efficiently with the same volume of dehydrant. The comparable final amount of water present after 4 changes of 25 mL each is 0.07 mL or 0.2%. Larger volumes of dehydrant give correspondingly greater reductions in water content as percentages.
Based on the foregoing it should be clear that dehydration is best accomplished by multiple changes, as many as possible, preferably with relatively large volumes of dehydrant, so that we get both a simple volumetric and a compounding effect on the levels of water present. What has to be stressed is that sufficient time must be allowed for the dehydrant and the water contained in the tissue to equilibriate, to completely mix. If this is not done, then when the dehydrant is replaced by fresh dehydrant, there is some carryover of excess water in the tissues.
It has been a fairly common practice to have the first position of an automatic processor filled with fixative, usually neutral buffered formalin. Some technologists also had a fixation bath in the second position. The second fixative bath is redundant. To get the same effect, simply double the time in the first fixative station. This frees up an extra position for dehydrant, and the time for it can be obtained by reducing all the other dehydrant positions by 5 minutes. Overall the time remains the same, but the extra dehydrant bath helps a lot with completing dehydration. This is a hangover from many years ago when two baskets would be used, one behind the other, on a single level rotary processor rather than purchasing a dual level model.
Anyone who drinks tea or coffee, or makes lemonade from lemon juice is aware that stirring increases the speed of mixing. Not mixing and leaving the drinks to stand will eventually accomplish the same thing through natural currents, but mechanically stirring will speed up the process so that what naturally would take hours now takes a few seconds.
So it is with dehydration, although the results are not quite as dramatic because of the inhibiting effect of solid tissues. Nevertheless, mechanical agitation does increase the speed at which ethanol mixes with the water present and ensures that water leaving the tissue is immediately removed from the vicinity and replaced with fresher fluid, measurably reducing the time required for dehydration. It should not be too vigorous. Partly this is because violent mixing is not needed but also because it may cause damage to the tissues or, worse, cause fragments to break off and attach to other specimens leading to misdiagnoses. A gentle movement of the fluid is all that is required, or periodic removal and refilling of the container as some automatic processors do.
Some automatic processing machines permit gentle heat to be applied during dehydration. Although the use of heat is known to speed up fixation, albeit at the cost of some loss in quality, its advantages are less clear during dehydration. Some of the common dehydrants are fixatives in their own right (ethanol, methanol, acetone, propanol) and applying heat undoubtedly causes a greater degree of fixation if the tissues were improperly fixed initially. Since all these fluids are poor fixatives, the "parched earth" effect may be exaggerated. If the tissue has been completely fixed, however, the dehydrant should have little to no fixative effect if used at a moderately elevated temperature. Hot, as opposed to warm, ethanol may still cause damage.
Since warming fluids makes them less viscous, it is to be expected that doing so increases the effectiveness of dehydration by increasing the ability to penetrate the tissue. Many technologists will give anecdotal support to it, and it is a common practice.
There is no clear evidence that using ethanol under reduced pressure increases the speed of dehydration. It is often done because of the improvemts it brings about for infiltration, so it is believed that it may do the same during dehydration. At the very least, at the reduced pressures used, it causes no harm and since it is readily available is often used.
Long, slow dehydration still gives the best quality results. One would be hard pressed, though, to find a laboratory today that allowed a week or so for processing of blocks, particularly in a diagnostic setting. Even 48 hour processing is now a minority approach. Overnight processing is probably the most common, with significant numbers of laboratories using same day processing for small biopsies.
Often the amount of time available for processing is fixed, start at 4 p.m. and finish at 7 a.m., for instance. The best way to determine the time for dehydration is to remove the fixation, clearing and infiltration times from this, then divide the remaining time between the number of available dehydration stations, giving a little extra to the first dehydrant. Keep in mind that if fixation is complete then dehydration proceeds more efficiently, and if dehydration is complete then clearing is faster. The same applies to infiltration. Give as much time as possible to fixation and dehydration as the success of the final two steps depends on those being accomplished properly.
Dehydration, when done at room temperature, starting with moderate concentrations of ethanol, and applying them long enough between changes for the water in the tissue and the ethanol to equilibriate, causes only a little shrinkage even if it takes several days. Applying slight heat may cause a little more than seen at room temperature, but it is still barely noticeable.