IHC and Special Stains to Identify Pathogens

Loralee McMahon, MS HTL (ASCP):

I am going to talk to you about using IHC and special stains to identify pathogens. Just to go over some of the objectives I've laid out for today, what I'm going to do is give you a brief overview of the reasons for the increase in bacterial and fungal infections. I'm going to identify some of the invasive bacterial, fungal, and viral infections that we encounter in the pathology lab, and specifically in surgical pathology. Lastly, I'm going to discuss how good histology, special stains, and IHC can help ID these pathogens.

Reasons for increased infections

Pathogens can attack all parts of the body. They know no boundaries. They can infect humans and animals and plants throughout the entire world. So, what are pathogens exactly? Well, they're bacteria. Fungus and yeast are in this category, and viruses and protozoans. If you were to open up any news website or read an actual newspaper, you would see lots of stories about pathogens. The big ones are E. Coli. That one appears in the news several times a year, especially scare because it comes from contaminated food sources, and a lot of times they don't find out that the food is contaminated until someone dies or gets sick, and they have to make massive recalls of the contaminated food. It's a little bit on the scary side when you think about it. Another hot one this summer was the Zika virus that was traveling from the tropical climate into the States, and it was very scary for pregnant women especially. Just recently there were two cases of leprosy that were diagnosed in some school children in California. So, it really does still appear.

Animals are infected by pathogens, and some of the common ones are listed here. Some of these can transfer to humans and some cannot, but they're also susceptible to pathogens. Even plants are susceptible to pathogens. Mostly they get fungus and they can be also susceptible to virus and bacterial infections.

Humans, however, have over 250 types of pathogens that can infect them. This list is just a small list of the more common ones. Some of these are deadlier than others, like the hantavirus, and some we can vaccinate against, like the chicken pox.

With all of these pathogens everywhere, it seems like a silly question to ask, but why do we care? Well, firstly, it can be very devastating to livestock. If it's an animal pathogen, it can cause entire farms to be shut down, and euthanizing all of these animals, while it's a massive loss of life, it's also a massive loss of money for these farms. Pathogens can spread rapidly, especially in developing countries. If you were to pick up the newspaper and look at what's happening in Haiti right now, the cholera is starting to run rampant after this hurricane that devastated the southern part of Haiti. Especially in these countries where the living conditions are poor to begin with, lack of sanitation, the water supply is poor, and there's little access to treatment, and there's most likely little access to labs that can help ID these pathogens. Pathogens can run rampant through dormitories or daycare centers or nursing homes. They have lots of people in close contact. We even had a case a couple of years ago in Rochester of measles. There was a little bit of a measles scare.

With the increase in world travel, people can just hop on a plane and head to a different country, there's this increase in the appearance of what we thought were extinct pathogens, like leprosy, for example. People didn't think in a nation such as ours or places where we have good sanitation and good water that these things would appear, but they can appear. There's also the increase in antibiotic resistance that comes from these pathogens. There's also a large amount of people that are walking around that are immunocompromised. Due to advances in medicine, people are able to leave the hospital when they're undergoing treatment for transplants or undergoing treatments for cancer, or HIV-positive patients are living longer. They're still immunocompromised, so they're an at-risk population. There's also a growing trend, whether you believe it or not, of the anti-vaccination crowd, where you are not vaccinating against these pathogens, so you may increase your chances of getting them.

Some statistics. The World Health Organization estimates that the 50,000 people die daily from infections. I thought this was an astonishing number, and this number might be higher due to the reportability from countries, and especially the Third World countries, where someone is infected. They don't have access to a lab. They don't know what they died from. Furthermore, they may have died from a pathogen, and it’s just not reported.

There's also been an increase in diseases. There's been 30 new diseases, that have been classified in the past 20 years.

Humans are colonized by thousands, if not more, bacteria, and yeast are in this category. Most are harmless and can even be beneficial. Your GI/GU tract contains a lot of these bacteria that maintain a healthy environment.

Types of infections

These are some of the types of bacteria. The shape is important to the pathologists when they're looking at these organisms under a microscope trying to decide what they are. The shape often helps them decide what type of bacteria they're looking at.

Fungal infections can be superficial, cutaneous, subcutaneous, systemic, and opportunistic. A lot of fungal infections aren't that serious, but they can be serious if you get into the systemic version of a fungal infection, especially after a surgery or if you're immunocompromised.

Here are your viruses. Viruses are also classified as one of the deadliest pathogen. They're very hard to fight. There are not a lot of things we can do to fight them, other than vaccinate, and there are some antivirals out there that can help.

Protozoa are also on the list of pathogens. These can be very deadly as well. They're usually found in tropical areas, and they can be very dangerous when they infect water supplies. There were cases of water supplies, like swimming areas, where people dove into water and got a protozoon that came into their body and they ended up dying because of the infection.

So, how do you treat patients that have become infected with pathogens? Well, for fungal infections, you can treat with antifungal medication. Bacteria and protozoon are often treated with antibiotics. Viral infections can be treated with antivirals if they're available. The common theme for all the treatments is that it has to happen quickly, so proper ID of the pathogen and speedy ID of the pathogen is needed.

How do you get these pathogens? Well, you have to have some sort of contact with it. You can breathe it in. You can touch it. You can eat it, if it's a contaminated food source. They can spread from animals to humans or human to human. A very big source of pathogens is mosquitoes and ticks.

We know that pathogens can be deadly, and complicating the matter is that we have a lot of people out there that are immunocompromised due to cancer treatments or transplants or HIV infection. We also have an increase in pathogens that are resistant to antibiotics. So, again, correct diagnosis and a fast diagnosis helps combat this.

The gold standard for identifying pathogens is microbiology, microbiology culture. It needs to happen quickly for the best patient outcome. If you're going to treat the patient, you need to find out what the causative agent is, so you can treat it properly, so it needs to happen quickly. Histology can diagnose an infection when the pathogen cannot be grown in culture or it's difficult to grow or it takes a long time to grow. If it takes more than a day or two, or up to a week, the patient may not have that much time, or the infection may get ahead of the medication, so that the medication isn't helping.

What happens when it takes a culture several days or a week to grow? What do you do? What happens if the ID of the pathogen needs to happen on the operating room table? Well, here is where histology can come into play. Just a couple examples of where histology can help, pneumocystis is notoriously difficult to grow in culture. You can do a diff quick on a bronchial washing, or you can do GMS on the tissue that comes into the lab. If the patient is on the operating room table, and the surgeon suspects that the patient has some sort of necrotizing fasciitis, which can be very deadly, the surgeon can leave the patient's surgical wound open to the air, which will help it heal, rather than close the wound. So, you can do a diff quick on the sample from the operating room within 20 minutes, and then the patient can either have a closed surgical wound and be treated for some other pathogen or leave the wound open, so it can heal.

Some of the reasons that we are seeing an increase in the incidence of pathogens in pathology, and I touched on this before, was treatment of patients with cancer, treatments for organ transplant, HIV-positive patients. More and more of these patients are coming in and having biopsies to find out if there's something wrong with the transplant or if they're being treated for cancer and they present to the office with some sort of skin rash or some other symptom that they may become into contact with a pathogen that is not normally seen in the human population. Like I said before, leprosy has kind of re-appeared because of world travel, and tuberculosis, again, also reappearing due to a lot of world travel. Again, there's this controversy on vaccinations. Should you or shouldn't you? It's causing some of these once thought eradicated diseases to be seen again in the labs. Pathologists need to be aware of this, and the lab needs to brush up on those stains that will prove that those pathogens are infecting the patient, so that they can give a positive identification.

Histology of pathogens – H&E

There are some pros and cons for both culture of a pathogen and histology of a pathogen. Culture is relatively non-invasive. You can get a specimen from blood, phlegm, anything that can be swabbed off of a patient. Histology is more invasive. You usually need some sort of surgical biopsy or a fine-needle aspiration to remove the sample from the patient. The turnaround time for microbial culture can be one to fourteen days. I've read recently where there is some automation coming to the microbiology lab, and this can take it from a day to a number of hours. Not being a microbiologist, I'm not totally abreast of those technologies, but histology has a relatively quick turnaround time, depending on the size of the specimen. If you get a small biopsy into the lab, you can usually turn that specimen around within eight hours. If you have a larger specimen, it might take overnight. A failed culture in micro means you've missed some time. If you had to culture a specimen overnight or two days or four days or fourteen days, you've lost that culture, and you've lost all that time, and you also need to go back to the patient and retrieve a sample. A failed stain in histology just requires you to re-cut and re-stain, so as long as there's tissue available in the block, you can still get another sample. One of the cons for processing histology is some of those organisms may not survive. Leprosy is an example. Leprosy will survive the processing steps, but when you go to stain the leprosy organisms you have to pretreat it with a very specific solution.

Fungal and bacterial infections in tissue are often detected with a good quality H&E. Frozen sections you can do a diff quick to get a rapid diagnosis. Infectious agents can be classified by special stains. You can also use IHC. IHC can help you find out if there's a viral infection.

Histology of pathogens – Special Stains

What are some of the stains that you can use in the lab? Some of the common stains that are used, at least in my lab, are good old H&E. There are a lot of things that can be seen on an H&E. GMS, Warthin-Starry, Giemsa, AFB or acid-fast, gram stain. IHC you can use. It's very specific to the pathogen and they're usually viral, although there are a few bacterial ones that are out there. There's also some NC2 hybridization that you can do on paraffin sections, such as CMV, cytomegalovirus, or Epstein-Barr, to find out if the patient is infected with one of those causative agents.

With all these stains to choose from, how do you make the decision? What stain do you use? This can be dependent on the pathologist. I've worked for a few different pathologists, and some pathologists prefer one stain over the other. It's something they're used to looking at and something that they prefer to be able to see that pathogen or that bacteria, something that they're used to looking at. Or, it can depend on the pathogen being detected. You need to know what you're looking for. If you're looking for a viral particle, you're not going to run a GMS. It's not going to show you that viral particle that is there. So, there are a lot of choices.

You should never underestimate the importance of a good quality H&E. This H&E in this picture is so dark that you can't see the H. Pylori organisms. They're there, and you can't see them. You can have drastic variability in your H&E stains if you don't have a good quality control program in place. This is another H&E of a gastric sample that should have some H. Pylori organisms floating around. You can see there's a giant fold in the tissue, and the stain itself looks kind of muddy, like maybe there was some water in the - -, or something is just not right. It's kind of out of focus and fuzzy. First line is to get a good quality H&E, a nice flat section, and very standard quality H&E on a daily basis.

Pitfalls when it comes to staining, whether you're doing and H&E or you're doing a special stain, you can have variability of the staining run. You can have variability of the technicians, which creates variability of staining. One tech may start the timer when they put the slides in the solution, and one may start the timer after they put the slides in the solution, so you get a little bit of variability in time. Automation can help this, but as I'll talk in a few moments, it doesn't always help it, if it's available. The size of your sample could make a difference. If you have a very small sample, you may not be sampling the pathogen. You don't know, or maybe your pathogen is deeper into the block than what you see on your cut slide. The quality of the sample, if you've heard me talk before I harp on this all the time. The tissue needs to be fixed quickly once removed from the patient. It needs to be processed for an adequate amount of time. The quality of the sample is very, very important to get a good quality H&E.

It's also important to have the history of the patient available on hand. A lot of times in the pathology laboratory the patient history might just be a couple of sentences on a requisition. History of skin rash, things of that sort, where you're not getting a good history of the patient. Or, the patient's history may not be available at the time the biopsy was presented at the lab. They may have a long history. They may have a history of traveling to foreign countries. They may have a history of a transplant. These are the things that the pathologists need to know to help guide them to determine what the pathogen is that's infecting patient.

I worked in a dermatology lab, dermatopathology lab, prior to this job, and our two biggest stains were the GMS and the PAS, commonly used to stain fungus, and the pathologists would need to see the size, the shape, the thickness of the fungus, if it was budding or no budding. This would help them determine the extent of the infection. Many, many samples would come with a patient history of skin rash or itching, like every sample we got. So, it helps if the patient history is a little bit larger, a little bit more specific. But, the pathologist would order the PAS or the GMS to decide what type of fungus or if it was a fungal infection versus something else. Maybe it was just an immune reaction with the skin. So, the problems with these two stains is that you can mask the color of the fungus, because it's making the fungus either pink if it's a PAS or black if it's a GMS. So, you can mask the color of the fungus, which is important in determining if it's a pigmented or colorless fungus. Both are technically tricky. If you're doing a GMS by hand as a silver stain, silver stains are notoriously tricky. PASs are also tricky is you are doing them by hand as well. There's a differentiation step; if you over differentiate or under differentiate you might not be able to see the fungus that are there.

So, the question with this variability from staining run to staining run it should you automate or should you not automate. The idea of automating is attractive, because you can remove the tech-to-tech variability or the day-to-day variability. We have automated our H&E staining in our lab, and we still have variability of staining, not so much day-to-day, but if you run 700 slides through, your solutions are getting diluted as you're running these slides through, so many the first slide is perfect, but your 700th slide is not. It's automated, but you’re still having variability. Special stains have been automated, but not all special stains are automated, so you're still going to have some stains that you need to do manually.

Immunohistochemistry is available, and it helps reduce the inconsistency, but there are some pitfalls with immunohistochemistry that I'll talk about later. There are still some issues. Most importantly, it's very important for the techs to be well-trained. They need to know what a good stain looks like. So, if they pull a slide off of an automated platform, whether it be special stains or H&E, they need to know what it should look like. So, you need to have well-trained techs that can trouble shoot. You should also have good controls in the laboratory, whether you get them from your in house or you buy them or you work with another hospital to obtain good, positive controls.

Like I said, you need to train your techs to trouble shoot, because an over stained or and under stained slide can mask an organism and create background, which can make the organism hard to distinguish or create a false negative. We still do a lot of our special stains manually, and the techs spend a lot of time on a daily basis looking at the stains, making sure that they're all staining properly. They know what a good control should look like, and they know if they're case control came out properly.

There are still many choices for bacterial stains. This, like I said before, could be pathologist dependent, or it could be money dependent. Maybe it costs you less to run a PAS versus a GMS, so you choose that stain. It could be that you wanted to automate everything, and the stains that are available to you are a smaller list than the actual stains that are out there. There are a lot of things to take into consideration.

This is just an example of variability of staining. This is a Warthin-Starry stain. I teach a class over the summer to medical technology students, and they go into labs and they stain slides. I get these slides back, and a lot of times they're control tissue that the lab has on hand. I looked at these slides, and I said, wow, this is the same stain, and I can see there are lots of organisms present, but the staining intensity is, so vastly different. The one on the left is very dark. The one on the right is very light. Imagine if you were looking for one or two organisms in this tissue. It would probably be pretty hard to find them. So, again, going back to having well-trained technicians, good quality control, good quality control for your tissue.

Another example is Giemsa. I have a fairly lightly stained Giemsa here. You can see nuclei. You can see there may be some organisms located in here. Then, I have this very dark Giemsa here. So, they're both the same stain, but look at how vastly different they look. Again, very important.

Histology of pathogens – IHC

When we talk about IHCs, which is my specialty, for diagnosing pathogens, we have a very small list. It's important for the pathologists to take into the consideration the entire patient history, along with the results of the stain that they're looking at, to make sure that the stain that they see makes sense, so they can troubleshoot the stain, if it shouldn't be staining or it should be staining. That can help you trouble shoot. So, proper control tissue, again, is very vital. If you choose the correct antibody, the antibodies are very specific to the pathogen, so you need to make sure that you choose the correct antibody.

Making sure that your counter stain is consistent. As I said before with the H&E, you need to have a good consistency with your daily counter stain. If one day your counter stain is very dark, you could mask some of the IHC-positive staining. Antibodies are usually pretty difficult to validate. The CAP has established guidelines for validating antibodies, and it may be very hard for you to get 10 positive cases of cytomegalovirus. These are kind of hard to do. So, you should work with hospitals that are close to you, or maybe you know somebody that you meet through the NSH or someone you could work with to help you validate these. Although there is a caveat to validating, most of these antibodies are considered analyte specific reagents, and it is left up to the director of the laboratory as to what the extent of the validation is. It needs to make sense for the lab. It can't cost you tons and tons of money or take up lots of time. It's just not cost effective. Most of the time the pathologists will look at the H&E slide and they'll see some changes in the tissue that will cause them to order the IHC , and it's used often, especially in the cases of H. Pylori, when it's difficult for the pathologists to see the organism with an H&E.

Some of the common ones we use the lab are CMV, cytomegalovirus. It's a herpes virus. It's a big problem for pregnant women. I've seen recent studies that have said that there's a large percentage of the population that are already infected with this virus, but it usually comes into play if you're pregnant. It's important to diagnose this properly as well. Toxoplasma is another bacterium that can be demonstrated with IHC. It's often transmitted from cats to humans, again, a problem for pregnant women. T. pallidum is starting to come. I've seen this more and more being used in our lab in place of the Warthin-Starry for identifying this spirochete. H. pylori, there are many options for H. pylori. You can often see it on H&E. You can see it with a Giemsa stain, or in cases where the patient has been previously treated for H. pylori infection or they have a malt lymphoma, they still could be infected with H. pylori. They're very difficult to see. There may only be one or two visible on the plane of section, and they usually hide out deep within the crypt. They're difficult to see.

As with H&E and special stains, you're going to have some pitfalls. You're still going to run into variability of staining if you haven't gotten your counter stain to be very consistent and have a good quality control with your counter stain. Your control tissue is very important. You need to have positive control tissue that works for the pathogen that you're looking for. If you have slide debris, you could block the positive staining. Usually, when they're using IHC for pathogens, there's not going to be very many pathogens in the tissue, so there may only be a few visible, so if you have slide debris or an over stained counter stain, you could mask the pathogens that are there. Your sample size comes into play also. If you have a small biopsy that comes, and the patient gets and H&E and a couple of special stains, and then they get an IHC, maybe by the time you get through that tissue, you've cut through those pathogens, and they're not apparent on the plane of section that you have that you're staining. Again, quality of your tissue plays into this as well. If you have poorly fixed or poorly processed tissue, the IHC is not going to work very well.

This is just an example of some slide debris that I've come across in my lab. As you can see with the big red arrow, that's just some slide gunk that appeared on the slide, and I'm not sure if it was tissue sloughing off. It could possibly be a contaminant in our buffer. It could be a contaminant in the water bath at the end of the staining run. Whatever it was, it was masking any organisms that were there. You could see there were a couple of the H. pylori organisms here that I'm putting this arrow on, but imagine if they were in here, and they were underneath this slide debris. It would be difficult to see.

This is just a picture that one of the pathologists showed me of some pigment in the tissue. It could be formalin pigment. It could be any kind of pigment. This can create a problem. If you're doing an IHC and you're using a DAB as your chromagen and your stain is brown, is this brown pigment or brown positive staining? That's why a lot of times we use a red chromagen to decipher between pigment and the actual organism. You see this pigment sometimes in lung tissue with patients that have smoked or were exposed to something, or if you have a fixative that creates a pigment. If you have a formalin that is not pH buffered properly, then you could end up with pigment. But, it can, again, mask your organisms that you're looking for.

I have a case example. It was an interesting case that kind of was the start of the reason why I wanted to do a presentation on this. We had a pancreas specimen that we received for complications following a transplant. The patient was immunocompromised, because they were being treated for transplant. They give you those immunosuppressing drugs, so that you don't reject your transplant. The pathologist received the H&E, and the bowel was ischemic. It has a funny appearance. The CNB showed numerous positive-appearing areas, but as the pathologist looked at the patient's slide, it looked a little blurry, cloudy, and they were just suspect that the staining was true. At the time, the patient's history was not available. It was from another hospital, and the patient's history hadn't made it with the patient's sample. So, the pathologist ordered a repeat CNB stain, and it turned out to be completely negative. They thought, well, okay. A positive result makes sense, because the patient was immunocompromised, but they didn't have the patient's history, so they weren't sure if there was a possibility of a CMV infection, so the negative stain was kind of believable, too. So, here's a picture of it. Here's the first stain, CMV, you see some positive inclusions here. The repeat, completely negative. Well, the pathologist was very suspicious at this point. So, they went back and they looked at both the original CMV stain and the repeat, and he happened to look at the control slide, which may sound silly, but if your control didn't work, chances are your patient tissue didn't work. The control was negative. So, it's very important to check your control. Well, the third time we repeated the stain, the patient was positive, so it just goes to show you that taking in the patient history is very important. Making sure that your positive controls work, is very important, to get the proper diagnosis. If this had gone out as a negative, then the patient would not have been treated properly.

Toxoplasma is one we see in the lab. We test for it in the lab quite a bit. We don't often see positive samples, but IHC demonstrates this very nicely, because you can't see it on H&E. You can see these big cells, but you need the stain to positively identify it. T. pallidum is just starting to sneak up on my radar. We've seen quite a bit of it lately. IHC also demonstrates this quite nicely. This is just a picture of an IHC, toxoplasma. You can see there are these large cells. You can see these on H&E, but they could be something else. Not being a pathologist, I don't know. They always ask for the toxoplasma stain to positively identify it. They're suspicious of toxoplasma, but they use the stain to properly identify it. T. pallidum, which is super tiny, and I hope you can see. This is a red chromagen. I put some arrows on it to see another immunostain that works quite well.

This is just an example that one of the pathologists gave me of an esophagus. I was not aware that Zoster could infect your—I think of Zoster as a skin infection and a skin rash, but I noticed that we had an esophagus that came back with varicella positive. This is also a good example of when it's dangerous if you don't have a positive control. These positive controls are pretty hard to get ahold of, so if you know a dermatopathology lab that's in your area, where you can get some skin samples that are varicella positive, that's very beneficial to you.

H. pylori is another big one that's come up. It's important to exclude H. pylori infection when you have atrophic gastritis, because it may suggest autoimmune-type gastritis. So, there's a different treatment for these patients. H. pylori are difficult to detect when a patient has previously been treated for infection. There's not as many of them around, and they tend to get down in the crypts. Monoclonal antibody versus polyclonal antibody, you need to make sure you're using the right one to identify the right bacteria. This is just a photograph of using H. pylori antibody with a red chromagen here on the left, and with a brown chromagen here on the right, just to give you an example.

Our pathologists have gone back and forth between wanting the red chromagen or the brown chromagen, especially when there are just a few of these little H. pylori floating around. Some of them prefer to have the red chromagen and some of them are indifferent, so they'll take the DAB chromagen either way.

So, just a short summary. There are many choices for detecting pathogens in surgical specimens, so a lot of things to consider. You need to know the patient history, or the pathologist needs to know the patient history. It's important to have a good quality H&E and a good day-to-day quality of your H&E, and good quality control in your lab. Have a quality control procedure or daily check of your H&E to make sure that the staining intensity is good. Maybe spot check some of the stains that come off your H&E stainer throughout the day. If you have a batch-type stainer where you're staining 100s of slides before you change your solution, you may want to spot check during the day to see if maybe you could change your solutions a little sooner, or even stretch it out. If you have an automated stainer that has solutions that are per slide, you may not run into this issue, but that's something to take into consideration.

It's important to have good quality control tissues for known positive samples. I've harped on this a lot. You need to know that your stains are working, so have good quality control and good quality control patient samples.

You also need to have well-trained technicians. They need not only know how to cut and how to run through the procedures of the stains, but they need to know what the stains should look like, not that it should just be brown or green or blue or pink. They need to know where the pathogens are located in your control tissue, what the staining intensity should look like. Maybe have a perfect set of quality stains next to where you check your controls and have the tech check his daily staining run versus your good quality control tissues that you've run. It helps out in the long run. You don't have pathologists coming back to you saying this stain didn't work right, please repeat it, if you have good, well-trained technicians that can trouble shoot. They also need to be trained on automated platforms. If you have automated platforms, they need to be able to trouble shoot the equipment. Is there something wrong with the equipment? Did solutions get switched the last time we changed the solutions? They need to be able to detect these errors and trouble shoot the equipment.

Questions and Answers:

Can you comment on why it seems that there is a dearth of IHC for specific pathogens, especially fungi? I feel like in the oncology field there are always new IHC methods, but infection disease seems stalled.

MS. MCMAHON: You know, I don't know. I think that it is one of those questions of where if it's not broke, don't try to fix it. So, if these stains work good, and they're reliable, then great. Use them, and continue to use them. IHC can be a little more expensive to run than a special stain, depending on the kits that you're using or if it's a silver stain or a non-silver stain, and it just seems that they're also getting a lot smarter at coming out with—maybe smarter is not the best word, but—at coming out with very, very specific IHCs that detect these pathogens. Whereas, if you run a PAS or GMS or an acid-fast, you have a broader range of what you're detecting. So, it maybe 1 of 20; whereas, if you're running an IHC stain, you know exactly what you're staining.

How do you get finger or toenails to stay on the slide when doing stains like GMS?

MS. MCMAHON: You have to treat the fingernail to make sure that it's soft enough for you to cut. I think cutting it thin enough is part of it. Using a charged slide or a coated slide, something that has some grip to it that's going to keep that nail on the slide, and baking the slide for quite a length of time and making sure that tissue sample is super, super dry, and then putting it in the oven and baking it for a long time. When I worked in the dermatopathology lab, this was an issue. We would have these fall off quite regularly. But, I think cutting them relatively thin as soon as you can and getting it on a slide, getting it really, really dry, and then baking it. We used to bake them flat. A lot of times when you bake slides in the oven, you put them in those metal racks or tissue-type containers, and you're baking them upright, so that the paraffin can kind of drip off, but if the slide is nice and dry before you put it in the oven and it's flat, I think it helps. I'm not sure why. Another trick I've tried is soaking—deparing your slide and soaking it in formalin. I don't know why it works, but it does work sometimes. I read it in a paper while I was doing some research on something, and for some reason it adheres the tissue to the slide better.

Validation is a struggle for a lot of labs. Do you have any tips or strategies on how to get a validation completed?

MS. MCMAHON: Yes. In order to validate some of these IHC stains, if they’re tricky to find tissue for, we will often call other institutions. I'll go into the pathologist's office, and I'll say you want me to bring this stain up. How am I going to validate it? I need to get at least five to ten cases to stain properly, and they will sometimes call in favors to fellow pathologists or colleagues that they've worked with before, and see if we can exchange some tissue samples to help in the validation. As far as special stains, again, the same thing. You need to make sure that your tissue controls are appropriate, and then try to stain some cases. See if you can get a multitude of cases to stain properly. I've often heard of labs using micro, culturing some of these things that they're staining. Culture the tissue. Take a swab of the tissue, send it to micro for culture, and then process the tissue through to paraffin, cut the slide, do the stain, and see if you get the same results.

On CMV staining, how often do you see a positive ICH without seeing the H&E features?

MS. MCMAHON: Well, I don't know exactly. I'd have to default to a pathologist, but I know in conversations with pathologists, they will often say I'm suspicious based upon the H&E. So, I imagine that they're seeing some sort of lymphocyte invasion or some kind of inclusion on the H&E, and then they'll go back to the patient history, and they'll say, eh, this patient might be CMV infected, so they'll run the CMV. So, probably a lot of the time they're using that H&E to notice that something is going on, something is changing in that tissue.

A follow-up question to the question about finger and toenails. At what point do you soak in formalin?

MS. MCMAHON: After you've processed the tissue, you've cut your slide, and you've depared it. So, in order to do your PAS or your GMS, you have to depar the slide to do the stain. So, right before you start your stain, after you've deparaffinized it, soak it in formalin for 30 minutes. I don't know why it works, but it helps.

We sometimes have culture-proven microbacteria staining only with FITE and not AFB. Any suggestions? Positive controls are always fine.

MS. MCMAHON: It could be the processing. I know that the FITE stain, you have to make sure that when you deparaffinize the slide there's—I'm losing what, there's something you add to the Xylene to make sure that you're coating those organisms, so they're not lost when you're deparaffinizing the slide. That could be. I'm not 100% sure. I'd have to look at maybe your protocols.

Do you have any tips for a successful GMS?

MS. MCMAHON: We do a lot of special stains in the lab, and we do see variability in our GMS, and it is a little tricky, because it is a silver stain, and I think that not stain—if you're doing it manually, not staining too many slides at once. Because, if you get a bunch of slides, and you're having to transfer them manually, and you have solutions, you're going to run into issues. I think making your silver solutions fresh helps. Not that I want to push automation on everybody, but when I automated our GMS when I worked in that dermatopathology lab, it rarely failed.

Which H. pylori do you prefer, poly or monoclonal? Why choose one over the other?

MS. MCMAHON: The monoclonal is very specific to H. pylori, and the polyclonal will cross-react with heilmannii, which is another spiral corked shaped organism. So, the pathologists have talked about it with me extensively. We used to use a polyclonal, and then the company that we were getting it from discontinued it, so we just happened to switch to the monoclonal. I don't believe the treatment is different, so I don't think it matters, but if you have a control that is full of heilmannii, and not pylori, you could run into the problem where your control will work, but your patient tissue is not going to work. So, you just have to know what your antibody is going to stain, and if it cross-reacts with anything, and just be aware that if the patient is infected with heilmannii, and you're using a monoclonal antibody, it’s not going to pick it up.

How do you decide if you're going to automate your special stains or not?

MS. MCMAHON: We struggle with this, because, like I said, we still do them manually, and we run about 9,000 to 10,000 a year. So, we're at the point now where we're going to validate, because we're taking up an exorbitant amount of tech time to run these stains. If we automated a few of our stains, our tech could take half hour or forty-five minutes to get all the slides ready, load them on the machine, walk away and go cut some H&Es, get with the bulk of the work. Right now, we're devoting an entire day, one tech has an entire day in the special stains area to stain these by hand. That's our number one reason why we would automate.

Another thing you could look at is what's your volume. Could you justify automating a stain if you have three or four of these a year, and you're buying a kit that costs—I don't know if the kit costs $1,000, and you're only staining three or four or five slides a year, probably not worth it. But, if you're staining five or ten or twenty a day, it may be worth it. There are just a lot of things you have to take into consideration.

I have seen PCR used for microbiological samples. Would you compare the sensitivity and specificity of IHC and PCR?

MS. MCMAHON: I don't do a lot of—I don't do any PCR. I can't really compare the sensitivity, but I know that we could use it as a validation. If I'm running an ICH, and could send it out for PCR to help me validate that would be one way to do it. I imagine PCR is more sensitive, but not knowing that much about it, being able to compare it would be hard for me. But, I imagine it's more sensitive.

Is there a role for in-house development of IHC reagents versus commercially-available specimens?

MS. MCMAHON: Commercially available reagents?

MODERATOR: Well, it says specimens, but I think she may have meant reagents?

MS. MCMAHON: I guess if you're talking about creating your—I guess it would depend. You run into a problem if you're creating your own in-house reagents, because of the classification of IHC. So, if you're in a research lab, and you can do it, great. But, if you're in a clinical lab, and you're billing for these, you're going to run into some issues with billing for antibodies that aren't in vitro diagnostic or IDD or ASR status. You're just going to run into that issue. So, if you can create antibodies that are specific to these organisms, great. But, I would just be careful. I live in New York. We are inspected by the Department of Health of New York State, and we are not allowed to do any of that. It's considered fraud in the eyes of New York State for some reason. So, we just have to be careful of the regulatory, and the billing aspects of it. But, if you can do it, I don't see a problem with doing it.

Is there a life expectancy or expiration date for control slides? Do freshly-cut control blocks perform better than pre-made slides that have maybe been prepared years prior? Is there a difference? So, the first part was it there a life expectancy or expiration date for control slides?

MS. MCMAHON: Well, I think it depends on who you talk to. I've worked with quite a few pathologists who have said that, yes, there is an expectancy or life expectancy to these slides, because they're cut, they're exposed to oxidation, and having to do with the proteins. In the same sense, I have stained control slides that have been years old, and they've still worked. So, I think what you need to do is consider—do a test yourself. Cut one, stain it, and then cut another one, and just hide it somewhere for a year or so. Then, stain it again, and see if it works. But, if you do have a good set of control tissue that is stained that you can consult while you're looking at your controls on a daily basis, you can maybe see if the antibody specificity of the control tissue is starting to fall off. I know that there are a lot of this when you deal with some of the ERPRs, and HER-2s. They don't like the slides to be very old. Obviously, if you're doing any kind of in situ work you don't want your slides to be very old, because that RNA/DNA can degrade. But, I guess it depends on who you talk to. I've stained slides that have been very old, and they've worked just fine. It's probably antibody-specific, so that's another thing to consider, but again, have a good set of control slides that you can look at for reference when you're checking your daily control slides to see if maybe this stain isn't as dark as it used to be when I cut the slide fresh. You can see some trends, maybe some antibody specificity is dropping off.