Warning! You won't be able to use the quotation basket until you enable cookies in your Web browser.
Warning! Your Web browser is no longer supported. Please upgrade to a modern browser.

Image Analysis Verification - A Process Approach

PACE credits are no longer available for webinars more than 6 months old.


This webinar will address a different approach to Image Analysis Verification, with a focus on the entire process to include slide preparation, scanning and analysis. Traditionally, Anatomic Pathology Laboratories have evaluated the accuracy of image analysis results by comparing the end point of the process, or the results of the image analysis algorithm, to other test methods and results from other laboratories.

The approach discussed during this webinar will focus in on the entire process. The first step in grasping the importance of this concept is understanding that the results of a standardized histology process can vary. So it is important to understand the sources of variation, the degree of variation and the impact of that variation on the image analysis algorithm. Procedures can be developed and tested to determine how variation in the histology and scanning process can affect image analysis results. This process will help determine and document the robustness of an image analysis algorithm so that the data or results generated are accurate and reproducible.

Learning Objectives:

  1. Define basic terminology.
  2. Discuss the sources, degree and impact of variation within the histology process on the robustness and accuracy of an image analysis algorithm.
  3. Demonstrate how to write and execute a verification protocol that can evaluate the entire histology, scanning and analysis process to include accuracy, precision and repeatability.

Webinar Transcription

LAURALEE MCMAHON:  Thank you it is my pleasure to be here and thank you, Leica, for sponsoring this.  I am happy to be here and thank you to everyone for taking time out of their busy day to join me.

     Just a little bit of a brief background of myself.  I have been a histo-technologist for 15, going on 16, years and a supervisor of the immunohistochemistry lab here at the University of Rochester for about 8 years.  Our lab performs general IHC for clinical use using automated equipment.  We also do HER2 FISH on breast and gastric specimens.  We also do a fair amount of in situ hybridization, kappa lambda EBERs on paraffin sections.  We also do some mouse and human tissue research for some of the researchers here at the University. 

     Today, I am going to talk staining other specimens and it has been a very hotly debated topic here at the University between the Histology IHC Department and the Cytology Department.  Nowadays, the large amount of testing that is done on the ever-growing smaller amounts of biopsy tissue we are being asked to do more staining with less tissue and we are oftentimes asked to stain cytological preparations or thin preps or maybe a frozen section or sometimes even de-stain an H&E and stain it over with an immuno.  The reason being is because the tissues are so small that maybe that particular section only contains the tumor cells of interest.

     So here are my objectives for today.  I want to outline what other specimens are, learn, maybe, an alternative way to handle decalcified specimens when dealing with IHC and especially FISH and ISH, in situ hybridization, review ways to validate IHC staining of cytological preparations, and a quick overview of disclaimers and their use.

     Here is a long list of acronyms that you might hear today, most of which you probably already are familiar with but they are here for your reference.

     Validation.  We hear lots of talk about it in the IHC lab and now more than ever regulatory bodies are making sure that we have done our job and validated all the antibodies before being used on a clinical case.  For a long time there was no guidance on how to validate and much of it was left up to the discretion of the Director of the lab.  In the past few years, in recent history, more guidance has come out as to when you need to validate, how you need to validate, how many cases you need to run, and other guidelines but we know that if you change antibodies or platforms you must validate.  If you change the clone of an antibody or the pretreatment or maybe the time that you incubate the antibody on the specimen you have to perform at least some kind of modified validation.  Therefore, it only makes sense that if you are using antibodies on something other than formalin fixed, paraffin-embedded tissues that you need to validate.  It is something that is outside the parameters of your original validation that you use to get your antibodies up and running. 

     So what are other specimens?  Well, in my lab, anything that is not formalin and fixed paraffin embedded.  We validated all of our antibodies using FFPE and so anything that is not that would be considered another specimen.  So cytology samples, you know, cell blocks or thin preps or cytospin smears, decalcified specimens, technically, are other specimens, as well, even though they are formalin fixed, paraffin embedded there is that decal stuff that takes place before they are paraffin embedded.  And maybe alcohol fixed or some alternative fixative.  I know that there are a lot of alternative fixatives out there on the market that are less hazardous than formalin and maybe being used in the labs.

     A lot of time when talking about validation with other techs and even with pathologists I hear this statement.  “I have already validated all of my antibodies and I know they work so, see, I have all of this documentation.  Why do I need to do this?”  Well, this may have been okay in the past but now we have all of these regulatory bodies that have something to say about testing that is done in the lab and something to say about IHC.  CLIA, CAP, the state you live in, could have rules but even the country you live in.  If you are outside of the USA there may be different rules than what we have here in the states.  Today, I am going to cover these regulatory bodies and what we do in the states and, hopefully, if you are listening from outside of this country that you could research any organizations that inspect your labs in those countries.  CMS and the FDA or a joint commission of hospitals can come in and inspect you so if your lab is inspected by any of these agencies you should really research and be aware of what they require for validation.

     So what are the rules?  Well, with all of these regulatory bodies having something to say about validation whose rules do you follow?  It can get very confusing and complicated.  The laboratory should always follow the most stringent guidelines.  That way you are covered.  I would rather be overdoing my validation than under-doing it because I don’t think that I know of anyone has been cited for doing too much validation.

     I am going to start going over some of these rules and I am going to start with the CAP.  The CAP seems to inspect a large percentage of the laboratories out there and they have really made a good attempt over the past few years to provide some guidelines and standardization to the labs for validation.  So in 2014 the CAP published a paper that was called “The Principles of Analytic Validation of Immunohistochemistry Assays” and it contained 14 recommendations that outlined how to validate antibodies and different specimens and different circumstances and very specific to class one, class two, predictive markers.  Two of the recommendations, recommendation number seven and recommendation number eight, speak to validating antibodies on cytological samples and on decalcified specimens.  Those recommendations kind of, you know, they tell you that you need to validate but they leave the discretion, the extent of the validation, up to the discretion of the Medical Director.

     The CAP checklist items that deal with validation there are several of them.  One is the ANP.22300.  It talks about specimen modification.  This checklist items says that if you modify your procedure to deal with other specimens that you need to have an SOP that describes the procedure.  So for example, in our lab, we sometimes are asked to perform IHC stains on frozen sections so frozen sections are not formalin fixed, they are fresh frozen, and they don’t require any type of heat-induced epitope retrieval so we perform the stain without it so it is different from what our initial validation was.  I have an SOP that speaks to this modification.  Well, it is great that you have an SOP but this, itself, is not a validation.  The CAP wants you to have a written procedure that speaks to the procedure but not, it doesn’t give you guidance on the validation. 

     So the next checklist item, ANP.22550, it is the QC of antibodies.  It says that positive tissue controls are used with each antibody and I am sure that all of you out there use positive tissue controls, either with each specimen that you run or as a batch control for a particular antibody.  But if you go down through the guideline and you read the note section it states that, ideally, the tissue control should be processed and fixed in the same manner as the patient specimen.  However, the CAP realizes that this is not possible for all types of specimens we receive in the lab, like cytological preparations or decalcified tissues.  It would be pretty difficult to keep this separate bank of control tissues to match your patient samples.  So the guideline recognizes this.  It says that you should perform a parallel test to make sure that the antibodies you are using perform the same in those other specimens as it does in your formalin fixed, paraffin embedded tissue.  I will talk a little bit about the parallel testing later on in the presentation.

     Two other CAP checklist items, ANP.22983 and ANP.22985, deal with ER, PR, HER2 or predictive marker testing.  One talks about fixation that you should validate your ER, PRs, and HER2s if your fixation is something other than 10% neutral buffered formalin because the CAP and most of the manufacturer’s guidelines state that you should use 10% neutral buffered formalin, but this may be out of your control if you are getting your tissues from an outside lab. 

     They also say that you should use a disclaimer if your tissues are decalcified and I will talk about disclaimers a little bit towards the end of the presentation.

     I happen to live in New York State.  I am not sure how many of the participants out there live in New York State but we have an inspection that comes from the New York State Department of Health.  They provide what they call General Systems Standards and I think there are other states out there that do lab inspections, as well.  They have developed this General Systems Standards and it was created for clinical labs like chemistry and automated labs and micro and all those other labs that are involved.  But they are making it fit surgical pathology and it has been kind of a struggle to make those guidelines fit what we do in the surgical pathology world because it is such a different animal than a chemistry lab.

     One of the guidelines says that the lab should follow manufacturer’s instructions for the instrument or test system except where there is a difference between the New York State requirements and the manufacturer’s requirements.  So I met with my Quality Department and I tried to figure out if this applied to me, if they thought it applied to me, and how I should handle it.  They said, “Well, you have data sheets and you should use those as your manufacturer’s requirements and that you also must follow the most stringent guidelines, whether it be New York State or the data sheet.  And if you modify any of that you need to validate it.”  But they really don’t provide any guidelines on how to validate it.

     So I investigated further into the New York State General Systems Standards and it talks about following manufacturer’s guidelines and the data sheet so our instructions for use or whatever the manufacturer calls them and that should be your guidance.  Most data sheets specifically state the type of specimens that their antibodies should be used on so if you look through different manufacturer’s data sheets you will see known application sections where it will say the immunohistochemistry works best on formalin fixed, paraffin embedded tissues or a specimen preparation section that says you should use it on formalin fixed, paraffin embedded tissue sections.  But you see there is a common theme.  If the antibody is not used on FFPE then you must perform a validation on whatever your specimen preparation might be.

     So backing up a minute just to make you feel the pain that I went through in going through the New York State guidelines and the Quality Department, New York State deems off-label use is a lab-developed test so I have an antibody and a data sheet and the data sheet says I should use it on formalin fixed, paraffin embedded tissue and I am trying to use it on cytological material so that is considered off-label use.  So our Quality department went through this lab-developed test and this chart has been proposed by the Department of Health in New York State to help you through deciding what your risks are when developing lab-developed tests. 

     So well-established methodology, which is what is at the top of this chart, says that if it is FDA cleared, which most IBD and ASR antibodies are, or it has been peer reviewed, meaning there has been a lot of publications about it, then you can say yes.  You can go down this side.  Most antibodies have been. 

     So then your next step is the key determinant.  Does this test system provide critical or essential information for diagnosis?  Yeah, I would say most antibodies do if it all just staining tissues with an antibody to see, you know, if it is a tumor of unknown origin or to get ER, PR status, for example, for chemotherapy.

     And then is this impacting?  Yes, of course it is impacting the patient so is it a low risk or a high risk impact?  It categorizes antibodies into moderate risk or low risk and our Quality Department decided that most of our antibodies were considered moderate risk.

     So back to the package insert, you are following the manufacturer’s guidelines, if you are doing what the package insert says, so in investigating a lot of the antibodies that we had in the lab I looked at some ASR antibodies, for example, the adenovirus.  It states that it is intended for qualified laboratories to quantitatively identify by light microscopy the presence of associated antigens in sections of formalin and fixed, paraffin embedded tissue sections.  So meaning they want you to use it on formalin and fixed, paraffin embedded tissue.  And the same for P63 antibody and I also found that cytokeratin 5/6 happens to work in frozen sections and cell preparations so that was neat.  You know, you are trying to decide what antibodies to test in the lab on cytological preparations and here I found an antibody that works well in cell preparation.

     So the FDA and CMS also have something to say about it and theirs is a little more vague when it comes to classification of antibodies.  They classify antibodies as class one or class two.  Class one antibodies have the lowest risk to patients, meaning they cause the least amount of harm and the majority of IVD and ASR antibodies are in this category.

     Class two antibodies provide a little more information for the patient and they guide treatment options so ER, PR, HER2, EGFR, Ckit, probably, PDL-1 is in this category, now, where it is helping a pathologist and an oncologist treat a patient.  So there is a little more risk to using this antibody in the lab if you don’t have it done properly you could inadvertently not get the right result to a patient and then this patient could not get or get therapy that they did not need. 

     Class three usually relates to medical devices and they are used to sustain human life and they really do cause the most damage to a patient if not used properly.

     Some antibodies can fall into both categories.  When you are evaluating these antibodies you should validate and bill based on how they are used.  There is a different CPT code for some of these class two antibodies.

     Class three antibodies don’t really fall into either category and are not used in pathological diagnosis.  RUO antibodies production is not regulated by the FDA and they really should not be used in a clinical lab, although, I have had many a heated argument with a pathologist who wanted to bring on an antibody with RUO.  CMS considers these antibodies as not medically necessary and if you bill for them they will not reimburse them.

     So now that you have a brief outline of all the rules and regulations and organization what do you do and where do you start?  Don’t scream.  It is extremely overwhelming but what we did was come up with a plan.  A plan of action.  You need to decide what you are going to do, what is the validation going to encompass?  What antibodies?  How many cases to test?  After looking at all those guidelines you can decide how many cases you are going to test for each antibody or maybe you are going to test for one slide that contains a TMA, tissue micro array, that has, you know, 40 to 50 or maybe even 100 different cases on it?  Meet with all the people involved, the pathologists, the technician, a data miner if you have to find cases in your LIS system to test.  Have your pathologists do some research if they will find some papers that maybe you need some papers to know types of cases that antibodies work well in.  They are a good source of information and if you can get a few good pathologists to work with you it will help.

     Write up your plan and delegate the duties to these people that are involved.  You need to write up your plan.  It keeps you on track.  It gives you a good recipe to follow to perform the validation.  Then, you are going to actually do what you wrote up.  You are going to perform the validation.  You are going to take all your antibodies and you are going to stain them on all the samples that you prepared.  Then, review the results and have the required people sign off on the validation.  This is important to have documentation of what you just did. 

     That is easy, right?  Take a look at your antibody list.  You may need to make it smaller.  Decide if you need to test all of your antibodies on cytology or decalcified tissues or maybe you don’t.  Maybe you don’t ever use a particular antibody on a decalcified piece of tissue.

     What types of specimens do you perform IHC on?  This is what needs to be validated.  Are you doing cell blocks, decal tissue, frozen tissue, all three?  Which antibodies are you going to test on these?  For example, you most likely will never be asked to do a beta amyloid or ubiquitin stain on a fine needle aspiration of the lung because these antibodies are used only on brain tissue so it is not necessary to validate them on lung fine-needle aspirates.

     When you meet and get everyone together get a many people involved as possible and spread out the work.  It is very important that everybody has a task and not one person is bogged down with performing the validation because it can be very overwhelming and it can take too long to do.

     When we did our parallel study in the lab we had two pathologists, two cyto techs because we had the cyto techs preparing our cell blocks, the growth room because the growth room was getting the tumor samples and the cytology scrapes, the Lab Director and myself and one of my techs so we had quite a few people involved. 

     We all had duties to perform and we had deadlines to meet and it helped it move along smoothly and in a timely fashion.

     Writing up the plan is important and realizing that it is not static is important because it will evolve as you go.  You may start to look at antibodies and say, oh, I forgot this one or maybe you couldn’t get a lot of samples so you had to research the antibody and see some other types of diseases that you could validate the antibody on. 

This will become your validation documentation so it is very important that you write everything down and it will help you keep on track.  Incorporating a timeline into this will help.  It will make sure that you all hit your marks and if you have any deviations from the plan you can document this, as well.

Performing the validation is the next step.  You want to perform what you have written down, what you have delegated.  This might be the most tedious part of the project but with a good written plan it should be the most straightforward.  It is what people do every day in the lab so it should be the easiest part for people to complete.

And then comes reviewing all of the data.  When you have performed all the tasks involved, look at all the data and is it acceptable?  If it is not acceptable why not?  Look for trends.  Maybe there is an antibody that just doesn’t work on cytology samples no matter how many times you try it.  At that point, you may want to tell the pathologist that this particular antibody does not work on cytology samples and you are not recommending it being used on patient samples.  Which sounds straightforward enough but I have had many an argument with a pathologist disagreeing with me.  Maybe there is an alternative antibody that you can use instead of this antibody that doesn’t work on cytology.  You could find a different antibody and, of course, validate that or maybe there is an alternate marker that works the same as the marker that you were trying to validate.

Sign-off.  Sign-off is after all of the documentation is completed, you have reviewed it, you have come up with your concordance rates and you have decided that you are finished.  This is very important.  This is what the inspectors want to see when they come to the lab.  Doing all of that hard work will be for nothing if you can't prove that you did it. 

Well, I am just going to use a disclaimer.  I don’t want to do all that work.  It’s too much.  I have heard this before but is this the best approach for the patient?  I mean if you think about if you were the patient would you want to hear that?  Well, I said in the report that I wasn’t sure that this worked in cytology samples so the pathologist should have known better.  I mean think about it from the patient perspective.  Is this the best approach? 

So disclaimers, should they be used?  Maybe you should ask your Legal Department, if you have one, but I think, at least, you should make the best attempt to make sure your antibodies work on as many different samples as possible and know when they don’t work.  That is very important.  Knowing when they don’t work so that they are not used on samples that they are not going to work on.

The CAP talks about disclaimers in two checklist items, ANP.22985 and ANP.12425.  The ASRs need a disclaimer because they don’t have a lot of guidance from the manufacturers so I don’t know if you have ever looked at an ASR data sheet from a manufacturer.  There is not any information there other than the clone and the species and they don’t tell you a starting dilution or any of that.  But the FDA does regulate the production of those ASR antibodies so we know that they are good quality, the lot-to-lot variation is very low, so the FDA has deemed them as class one so you can use them on clinical samples but I think the validation may have to be a little more robust to prove that this antibody works in your hands on certain tissue samples.

This first, ANP.22985, talks about predictive markers on decalcified specimens and I will talk a bit about this later.  If you are going to do predictive markers on decalcified tissue specimens then you need to have a disclaimer in there that says this was performed on a decalcified tissue.  We can't guarantee it works the way it should because that is not how it was validated.  It is maybe a good time to ask your Legal Department, if you have one.  They can kind of guide you on how to word the disclaimer properly.

So back to the decalcified specimens; we had a very big increase in the number of requests for HER2 IHC and FISH on bone biopsies.  If you did a little research on this you would see that 80% of breast cancer patients will develop a skeletal metastasis during their lifetime and the ASCO/CAP recommends repeat testing on all recurrences.  So if you have a breast cancer that metastasized to the bone they want you to repeat the ER, PR, HER2 on the recurrence to see that it has the same status as the original or if it differs, you know, what happened. 

So here is your problem.  We didn’t validate our IHC ER, PR, HER2 on bone tissue and the FISH doesn’t always work on decalcified tissue, depending on how long it was put in decal, what type of decal you are using and, again, it wasn’t validated on decalcified tissue.

So our solution was to separate the soft tissue, the soft clot tissue from the bone biopsy, and process that separately from the bone.  The clot piece now becomes a FFPE tissue, formalin fixed, paraffin embedded, and it falls into the same parameters as our original validation so now we are avoiding decal, we are not needing to validate it because we are using our original validation, and we don’t need to use a disclaimer so we have covered all of our bases.

So this is just a growth photograph of how we do it.  This specimen came as one piece and what we did was separate the core, which is on the left, the hard, bony tissue, from the clot, which is on the right, and we put the bone into decal and processed it like we do all of our other bones.  Then, we took the clot and put it into fixative and processed it routinely.  No decal. 

And this is what we got.  So as you can see from the H&E there is a lot of tumor in the sample and you can also see that we performed HER2 and our pathologist read this sample as equivocal and he wanted to FISH it, too, because he was curious to see if it would work.  We did.  We FISH-ed it and it was positive.  It is kind of hard to see in this photo but it was positive.  We attempted to FISH the decalcified bone core and we just couldn’t get a good result.  A lot of the tissue would fall off of the slide and the hybridization just did not work.  We didn’t get a lot of cells with the control with a lot of signals to count.

We also took a look at our cytology, our current practice.  So our current practice for cytology was to take the fine needle aspirate, the FNA, and place it into a container with saline.  We are not allowed to have formalin by the patient, in the OR or by the patient, so it is the responsibility of the doctor taking the FNA or the technician picking up the FNA to get the sample, get it into saline and get it back to the lab where one of the prep techs or the cytology techs can pour some formalin into the container. 

So then after the formalin is poured into the container the sample was spun down into a cell block and placed into a cassette and processed for routine histology, which sounds okay.  You are fixing it in formalin and you are processing it into paraffin but what we found out is that we had some concordance issues when we were running our parallel tests so two problems here.

One is that we were diluting the formalin so you took your sample, it was in saline, and you poured 10% neutral buffer formalin on top of it.  If it was equal parts you are now making a 5% concentration of NBF and I don’t think they were measuring it.  I think they were just pouring it in.  They weren’t intentionally doing anything wrong.  They were just following the protocol.  They knew it needed to get fixative on it but I don’t think they quite understood how it was affecting the tissue, the sample.  It could be less than 5% if there was more saline than NBF being poured on top of it.  Maybe the concentration of formalin was less than that, 2%, 4%, we didn’t really know.

The other problem was that the time we were recording was the time the cassette was placed in formalin, not the time that the specimen had formalin poured on to it.  So we investigated further into this and sometimes it seemed like the samples were sitting in saline for hours or even overnight, depending on what time it came into the cytology lab.  We all know that tissue that is not placed into formalin quickly degrades after being taken out of a patient so it kind of makes sense if an FNA, if it was left in saline overnight, it would also start to degrade.

So we decided to perform a parallel study to see exactly what was going on and if our antibodies were performing the same on tissue as they were on cytological samples.  So as tissues came into the growth room the PA would take a sample of tumor, place it in the fixative, and routinely process it.  They would also take a scalpel blade and they would scrape off some of that tumor and place it in the saline, just like they do the FNA.  They would bring this to cytology and ask them to create a cell block.  That was current practice.  No changes.  We just wanted to see what our results were going to be.  No changes to the procedure.

And here are our results.  Well, our concordance rate was kind of low, lower than we had hoped and expected, and what we found was that if you look at the date and time that these samples were collected, so here is the date the sample was collected, the time that it was collected, and now here is the time that the cell block was being fixed.  This first one, 4/21, collection time is 12:46; the cell block fixation time was the next day at 9:50 in the morning.  There are enough publications out there that tell you that ER really starts to degrade after one hour after not being placed in fixative.

So what is to say that doesn’t happen with other antibodies?  We were finding a trend.  The next day, this is later in the day, 2:00; it was not being put into fixative until the next day. 

What we did was take different tissue samples, we stained the solid tumor and the cytology sample with the same antibodies just to see if the results would be the same and we found that sometimes the antibodies didn’t work the way we expected them to. 

Here is a photo of ER stain of the cytology preparation, which was probably that same tissue from the previous slide.  You can see a couple of cells maybe stained in that cytology preparation but nowhere near the amount of staining that is in the ER of the solid tumor.  The solid tumor was place into fixative within an hour of receipt and the cytology preparation was put into saline and brought to the cytology lab and given fixation the next day.  So this is a big problem and we didn’t realize it until we started looking at all of our procedures and how we are processing these samples and how it affects the IHC.

Again, just to kind of review, why the low concordance?  Well, we were pouring 10% neutral buffer formalin into the sample, mix it with saline; it was diluting the NBF so how do we know that this is okay?  We don’t, really.  We could test this if we wanted to.  The collection time to time in fixation was variable.  Sometimes it was within a couple of hours, sometimes it was the next day, and the fixation was variable.  Maybe it was placed into fixative and it didn’t meet the cutoff time to process it in histology so you may have gotten 6 hours of fixation or maybe 20 hours of fixation or maybe it only got 3 hours of fixation so that was variable, as well.  And we know looking at the ASCO/CAP guidelines for ER, PR, and HER2 that it needs to have at least six hours of fixation time.

So some hurdles that everybody has to go through and some that we found.  Do you get your specimens from outside labs?  This is a big hurdle.  If you are getting your specimens from outside lab you have no idea how they fixed it, how they processed it.  Maybe they are using an alternative fixative and they don’t want to let you know what their processing protocols are.  This is kind of hard to fix unless you have a good working relationship with the lab and they can work with you to try to marry up your processing protocols and make sure that your validation is valid on the tissue they are sending you.

It really is in the patients’ best interest so if you don’t have a good relationship with that outside lab or they are not willing to share that information with you push them a little.  Tell them that this is about the patient and not about sharing protocols. 

In our case, we were working with another department so our Histology Department and our IHC Department and our Cytology Department are all run by different people and we didn’t realize until we started this parallel study that what we were doing was affecting the IHC results.  Right now, we are working on getting a different procedure in place so that it is possible to get that FNA processed as close to the time that the solid tumor was taken.  So rather than it going over to cytology and sitting maybe having the growth room personnel apply the fixative or maybe it is getting spun down immediately instead of waiting.  So a way to fix the process to fit the validation.

Last but not least, you really need your pathologists’ buy-in and support.  I have worked with a lot of pathologists in my career that do not understand the process of validation and they don’t understand why an antibody doesn’t stain their alcohol-fixed tissue the way it stains a formalin-fixed tissue.  It can be very frustrating as a tech and not beat up on pathologists but I have also worked with some pathologists that are very good at this and they recognize this and may have been the ones that have been our champion to try to get this completed.  It is very helpful if you have pathologists that will work with you to get through this.

So in summary, you know that you have to do some type of validation so look at your procedures to see if there is a way to modify them to fit your already-established validation, like taking out that cytology sample and getting it into fixative right away and processing it the same way that you process your tumor samples.  Or, avoiding decal so you don’t have to put your specimen through decal before you run any immunos on it.  Obviously, this is impossible if you are running bone marrows on, you know, for your Hematopathology Department.  You can't avoid decalcification so you do; in this case, need to do a parallel study to make sure that the antibodies that you use on bone marrow samples are working with that decalcification. 

From my experience is that most antibodies work very well with decal.  It wasn’t a big deal.  When we ran our bone marrows with all of our hematopathology markers on those decaled specimens they worked great.  It wasn’t a big deal.  It was just a matter of getting it completed.

Can you fix your cytology material differently?  Sure.  You could work with your Cytology Department or if you are in charge of your Cytology Department you can kind of try to make it match what you do for routine histology.

So hopefully, today, you have gained some knowledge to help you go back to your lab and help validate your other specimens or run a parallel study.  Staining those tiny, little specimens that have four or five tumor cells and just hopefully you gained something today that will help you go back to the lab and run one of these validation studies and thank you so much for joining me today.  It was a pleasure to be here.



MS. SANCHEZ:  So Lauralee, the first question, it is a two-parter from Tracy.  She is a Lab Director, the only one that can sign off on validation or can it be someone they designate?

MS. MCMAHON:  That’s a good question.  I think that there is differing opinions.  I know that in New York State the Lab Director is responsible for signing off of any validation or procedures, SOPs, that you come up with in the lab.  The Lab Director can designate someone to sign off on QA and daily, you know, steaming run logs and temperature sheets and things like that but I think when it comes to a validation or a protocol that your Lab Director needs to sign off on these.

MS. SANCHEZ:  On decal solution do you typically use, somebody is asking, and what is your full procedure, you know, like long time, you know, fixative and so on and so on?

MS. MCMAHON:  Well, we have had kind of an uproar in our decalcification process in the lab.  We have used acid decals for a long time, you know, Formical, the fast decal, and the techs would go in and depending on the size of the specimen they would test it.  If it was bendable or they felt it was decalcified enough then they would process it.  but now our Hematopathology Department is really pushing to switch to EDTA, which we know is a little slower but it does help the hematopathologist use this tissue for other tests, for enzyme histochemistry and things like that, so we kind of use two.  We use some EDTA solutions, depending on the type of sample, and then for all other things, femoral heads, things that we know aren’t going to get any enzyme histochemistry, we use the acid decal, the quicker decal.

MS. SANCHEZ:  Thank you.  On parallel testing what happens if the antibody does not work for the specific tissue that you are looking for?  Do you do something to edit your procedure to get this antibody to work?

MS. MCMAHON:  You could.  You could find another antibody, if you wanted to, or maybe you could modify the procedure because you know that formalin fixation crosslinks proteins and you have to do a heat-induced epitope retrieval to get these antibodies to work, for the most part.  So if you add decal to the mix you are adding another variable.  You may need to put your antibody on longer or maybe modify your pretreatment so I would definitely go ahead and edit the procedure for the antibody to work on that specimen.  If you can do it, maybe increase your concentration; extend your time, things like that. 

     Obviously, you would have to have an SOP that speaks to that, that you are modifying the procedure, and maybe a little parallel study that says you modified the antibody procedure on bone marrow biopsies and it works best at this dilution for this time with this detection system and it worked very well and you proved it. 

     So I think that, yes, you should, you know, try to make it work, if you can get it to work.

MS. SANCHEZ:  Trevor is asking you is there a special procedure that needs to be followed with regard of this staining an H&E to do immuno?

MS. MCMAHON:  Well, what we do is we obviously take the coverslip off after soaking it in xylene and there is not really a special procedure.  I have worked with some techs that say you have got to decolorize the slide completely, soak it in acid alcohol, or you are soaking it in just 95% alcohol to get the hematoxylin and the eosin out of the slide.  But now you are exposing the tissue to acid alcohol.  Is this affecting the way the antibody works?  I don’t know.  I haven’t gotten that deep into it.  I am trying not to re-stain H&Es if I can completely avoid it but, yeah, I would say there is not a real procedure other than getting that coverslip off and getting that hematoxylin, eosin off. 

     I have worked with other techs who have said the H&E doesn’t matter that the hematoxylin and eosin is there because you are going to be staining it with an antibody and you are just going to stain over the top of it.  It seems to work either way.

MS. SANCHEZ:  Okay.  Marion is asking what type of labs were the New York State Department of Health General Systems Standards written for?  Are they clinical or are they more research?  She is just asking that general question.

MS. MCMAHON:  Oh, yes, they are written for all clinical labs so any lab that is in our hospital, chemistry, automated, toxicology, virology, you know, all the labs in the hospital, have to follow these New York State Guidelines.

     And then the blood bank, you know, of course, has New York State that comes in and they also are inspected by a specific blood bank for the blood bank.

MS. SANCHEZ:  Thank you.  And Joanne is asking is there any special procedure for thin prep preparations?

MS. MCMAHON:  There are a couple of papers out there that have stated that if you take a thin prep that has been, because we are asked to stain thin preps quite frequently and that is kind of what brought about this is we found that one antibody didn’t stain particularly well in thin prep so working with our cytopathologist she had read a couple of papers that said they soak their thin prep slide after it is created.  They put it in formalin for 30 minutes and for some reason the antibody seems to work very well. 

     So we tested this.  We had some leftover cytology thin prep slides created and we soaked them in formalin and we stained them with the antibodies.  Obviously, we didn’t have a parallel solid tumor to compare it to but we did make an effort to stain the thin prep with the antibodies of interest.  A lot of times it is CK5, P63, TTF, you know, the things that they get, lung FNAs are big around here, so we stain those thin preps using this method and the samples seemed to stain quite well.

MS. SANCHEZ: You and I talked about this, the complexity of the different laws, whether they are FDA, CAP, New York State, or international laws, when it comes to validation.  A lot people don’t know, really, where to start.  Do you have a suggestion?

MS. MCMAHON:  Well, I think I would start with whomever is the next agency that is going to walk into your lab.  If the CAP is going to walk in a month or two you better go and follow their rules.  But if you have some time and you want to sit down and pull try to pull all of the regulatory agencies that can inspect your lab, you know, together and make a guess.  There is some leeway here.  If an agency comes in and you inspected based upon the CAP guidelines and New York State came in and said it is not really good but at least you made some sort of effort. 

     It is difficult when you start looking at all the rules.  Here, at the hospital, the hospital is inspected by the joint commission but the lab is not.  We are only under CAP and New York State.  We do, occasionally, get an FDA visit but that has to do with some clinical trials that we perform.  They don’t usually come in looking at our IHC protocols but I have heard that FDA and CMS can look at patient billing and maybe a patient, if you are over billing, they might come in and look at your protocols and say, hey, you guys are billing an exceeding amount of IHCs.  I want to see what antibodies you are using and they best all be class one or class two and they had better be IBDs.  They better not be research use only.

     Although, that being said, we do use some research-use only antibodies but we don’t bill for them and the pathologists are not supposed to report on them.  They are just supposed to use them as assisting their diagnosis of the H&E, which is kind of a slippery slope.

     But, yeah, I would say getting all of the guidelines together and kind of trying to figure out which one is the most stringent and going with that one.

Leica Biosystems webinars, training presentations and related materials provide general information regarding particular subjects and are not intended to be, and should not be construed as medical, regulatory or legal advice. The views and opinions expressed are the personal views and opinions of the speaker(s)/author(s) and do not necessarily represent or reflect the views or opinions of Leica Biosystems, its employees or agents.

For the use of any product, the product information guides, inserts and operation manuals of the various products and devices should be consulted. Leica Biosystems and the editors disclaim any liability arising directly or indirectly from the use of devices, techniques or procedures described in these materials.

Copyright © 2017 by Leica Biosystems Richmond Inc. All rights reserved.LEICA and the Leica Logo are registered trademarks of Leica Microsystems IR GmbH.