2016 Revision of WHO Classification of Tumors of Hematopoietic and Lymphoid Tissues, Non-Hematopathologist, (Part 1 of 2)
This presentation is a review of the most important changes to the WHO hematopathology classification system as part of the 2016 revision. Emphasis will be placed on presenting the information so that it is relevant and understandable by non-hematopathologists and laboratory staff. This will be a two-part presentation, with the first lecture covering lymphoid neoplasms and the second presentation covering myeloid and precursor neoplasms. November 17, 2016 is the part two of the series.
- Understand the need for the 2016 revision to the 2008 WHO hematopathology classification system.
- Review pertinent changes involving lymphoid neoplasms in a manner understandable by general pathologists and laboratory staff.
- Illustrate important situations where advanced testing (e.g. IHC, molecular, FISH) may be helpful in further distinguishing hematopoietic neoplasms, which had been difficult to classify or confirm with certainty in the past.
PHILIP FERGUSON, MD:
The WHO classification for hematopathology was first published in 2002. For hematopathology it served as a unifying language that allowed pathologists and treating physicians to communicate on a much better level than before. Prior to the WHO classification molecular diagnostics, pathology, and oncology were much more siloed in practice and communication. Much of this was due to ignorance of the underlying mechanisms of disease and appropriate treatment. Diagnosis were less specific and drug regimens were broad in scope.
As understanding of disease pathogenesis has progressively increased and more effective personalized drug regimens available, there is a great need for molecular diagnostics to come together with pathology and morphology along with clinical oncologists to develop a classification system that integrated all three aspects. This was really the founding principle behind the 2002 WHO Hematopathology Classification System. Many people call this the hem path bible. It was not perfect, it was never intended to be perfect but rather a more precise scaffolding where knowledge could be deposited and diagnoses and classifications refined.
In 2008, the WHO pathology classification system was updated with refined diagnoses and new entities, some provisional. Again, this has served medicine well but knowledge increases and treatment options emerge and become more refined to help patients. It is therefore necessary to once again update the WHO book.
In 2016, we have a revision of the WHO Hematopathology Classification as opposed to a new edition. The International Agency for Research on Cancer, the IARC, oversees the WHO Classification System. Most of you probably know these as the Blue Books. Since this is a revision some provisional entities were promoted to definite entities and a few provisional entities were added but no new definite entities were permitted according to the IARC Rules. In 2014, the Clinical Advisory Committee met in Chicago, Illinois to develop this revision. The committee was composed of pathologists, hematologists, oncologists, and geneticists from around the world. There were actually two committees; one for lymphoid neoplasms and a second for myeloid neoplasms. It is the integration and consensus of all these specialties that makes the WHO classification so useful for patient care.
This presentation will be the first of two talks. Like the WHO Committee we will divide the presentation to cover changes to the lymphoid neoplasms today and in November we will examine changes to the myeloid disease categories. This talk is specifically targeted to non-hematopathologists and laboratory scientists. This is not meant to cover every change to the classification but rather highlight some of the most important changes which would likely affect a community laboratory practicing general pathology.
I would recommend any laboratory or pathologists who signs out hematopathology to download and read a fairly comprehensive review written by Dr. Swerdlow published this past May in the journal Blood. I believe that this is a free download from the Blood journal website.
[As you can see, there are numerous changes to the classification, mostly minor, but many significant. Today I really want to focus on a few changes to either more common entities that you will see in your lab or less common entities with significant changes that could really help pathologists on a day-to-day basis in their practice.
Hairy Cell Leukemia
Most patients with hairy cell leukemia present with splenomegaly, pancytopenia, and significantly, monocytopenia. Classically the bone marrow biopsy will be a dry tap because of reticulin fibrosis induced by the hairy cell leukemia. One will see a few circulating neoplastic cells that often will have hairy projections in the peripheral blood.
Hairy cell leukemia is an indolent B cell lymphoma that is characterized by hairy projections on the peripheral blood smear samples and it is an uncommon disease representing only about 2% of lymphoid leukemias. Patients typically have bone marrow and spleen involvement with a small number of circulating cells. Flow cytometry has a characteristic pattern where tumor cells express CD20, CD11c, CD103, CD22, and significantly, CD25. Immunohistochemistry for TRAP is usually positive and cyclin D1 can also be weakly positive. It’s important to remember that cyclin D1 expression can be seen not only we think about it in mantle cell lymphoma but also plasma cell myeloma and hairy cell leukemia. The difficulty with diagnosing hairy cell leukemia has often been differentiating it with hairy cell variant.
Here are a few images from a case I recently had in the peripheral blood and in fact, the clinician was tricked and not thinking it might be hairy cell leukemia because the patient had monocytes on their CBC report, but actually what was happening is that the hairy cells were so large they were registering in the monocyte gate so the patient actually did not have any monocytes in the peripheral blood. You can see in the bottom right corner where there is significant replacement in the bone marrow biopsy sample. Further pictures you can see in the top left how the marrow is completely packed and then the bottom right you had the CD20 antigen highlighting the vast replacement by the hairy cell leukemia. The reticulin stain was incredibly strong and diffuse throughout the bone marrow and the specimen was a dry tap.
This is a classic flow cytometry sample you can expect to see in a patient with hairy cell leukemia. Typically, you will have a CD10 negative, CD5 negative monotypic lymphoid population that has pretty bright CD11c expression and you can see that across the top and in the bottom right corner.
On the follow-up panel, specific for hairy cell leukemia, you can see how in the top right you have your co-expression of B cell marker with CD103 and that when you examine those they co-express CD22 and CD25. Further analysis at the reference lab showed that this case did in fact, have a BRAF mutation.
The BRAF V600E mutation has been found in nearly all of cases of hairy cell leukemia. This is first described back in 2011 by Tiacci and colleagues and published in the New England Journal of Medicine. BRAF testing in hairy cell has really reduced the difficulty of excluding similar entities, particularly one that’s called hairy cell leukemia variant. Testing can be done by a PCR or by immunohistochemistry now. There is a VE1 clone that is very good. PCR can have a very low level of detections, it’s really great especially when you’re looking for residual disease or upfront diagnosis at the beginning when the marrow is packed.
But more recently studies have been done looking at hairy cell leukemia with immunohistochemistry specific BRAF V600E immunostain and it showed excellent results with PCR. And you can see here how the hairy cells stain with the VE1 clone.
One of the big difficulties is how do you differentiate hairy cell leukemia from other B cell lymphomas which express CD103. Really, we think about three entities. Hairy cell leukemia is kind of the classic CD103 positive B cell lymphoproliferative disorder but in the back of your mind you always have overlap with other entities specifically splenic marginal zone lymphoma and hairy cell leukemia variant. Hairy cell leukemia variant is not related to hairy cell leukemia biologically. It’s just the morphology and immunophenotype can be very similar.
With the BRAF mutation this helps us to separate those. Hairy cell leukemia variant has also been found to have MAP2K1 mutations in about 50% of the cases and splenic marginal zone lymphomas have also been found to have NOTCH1 and 2 mutated pathways along with NFK beta pathway mutations in some cases.
Hairy cell leukemia variant you often will be suspicious for these cases because when you do the flow cytometry CD25 will be negative. This is a case recently of a suspected hairy cell leukemia variant and you’ll notice a couple of things; one is the lymphocyte gate is much tighter and with less side scatter because the cells tend to be a little bit smaller. Secondly clinically when you do the bone marrow biopsy they often will not have a dry tap and so you’ll have aspirate smears and then when you look at the biopsy the marrow won’t be packed out but you’ll just have an interstitial and sometimes sinusoidal pattern of infiltration by the monoclonal cells. When you evaluate these by flow cytometry it can look pretty similar to hairy cell leukemia but importantly in the bottom right you’ll see that there is lack of co-expression of CD25 on the CD103 positive cells and almost all cases of hairy cell leukemia will have CD25 expression so this can be a very important tipoff.
Most of the time hairy cell leukemia variant cases will be CD103 positive and splenic margin zone cases are typically CD103 negative but that is not the case all the time so it can still be difficult to differentiate between hairy cell leukemia variant and splenic marginal zone lymphoma. But with classic hairy cell leukemia the presence of the BRAF mutation really adds to the confidence in separating that entity out from other B cell lymphomas, hairy cell leukemia variant and splenic marginal zone lymphoma.
You’re probably asking yourself why is it so critically important to separate this one entity out? The classical hairy cell leukemia has an excellent prognosis when treated with purine analog drugs and hairy cell leukemia variant, splenic marginal zone lymphoma those cases do not respond typically to purine analog medications. To be able to correctly identify and categorize hairy cell leukemia is critical because it is an excellent prognosis, greater than 90% cure rate, a relatively short and easy treatment course for those patients. Now with the identification of the BRAF mutation that’s just an additional level of confidence really making a firm diagnosis of hairy cell leukemia.
Next, we’re going to examine lymphoplasmacytic lymphoma. Lymphoplasmacytic lymphoma occurs in adults with a median age often in their sixties and has about a 20% rate of familial predisposition. The interesting thing with lymphoplasmacytic lymphoma and the diagnostically challenging thing is that it can have morphologic overlap with so many other lymphomas that have plasmacytic differentiation.
The description of small lymphocytes plasmacytoid lymphocytes in plasma cells is not specific to LPL and therefore is a diagnosis of exclusion in the WHO Classification. It’s important to note that many of these people fulfill criteria for Waldenstrom’s macroglobulinemia and also that most cases have an IgM paraprotein.
There are several excellent articles written on the challenge and how to differentiate LPL from other marginal zone lymphomas with plasmacytic differentiation. This article here is an example of a good one written back in 2011 and those are very important and are still very relevant. In 2012, an article published in the New England Journal of Medicine identified the MYD88 l265P mutation in a significant number of lymphoplasmacytic lymphomas. This has really been an important part in the evaluation of these lymphomas. The 2016 Revision emphasizes that the MYD88 mutation is present in a vast majority of the cases, about 90%. Also, interestingly is a better clarification of this idea of IgM MGUS cases are more closely related to lymphoplasmacytic lymphoma than they are to myeloma and I think most people with experience in hematopathology who recognize that IgM myelomas are a very rare subtype of myeloma to begin with.
The MYD88 mutation is good and very helpful but it’s not perfect. Even though 90% of cases of lymphoplasmacytic lymphoma will have this mutation it can also be found in other lymphomas including the three listed here which can have significant morphologic overlap with lymphoplasmacytic lymphoma. Up to 15% of splenic marginal zones, 7% of extranodal marginal zones and even 2 to 3% of CLL SLL cases can also have this MYD88 mutation. Even though we’ve taken several great steps in the clarification of this diagnosis of exclusion MYD88 is not a silver bullet and you can’t use it isolated alone and by itself with regards to making this diagnosis, but it can go a long way in helping us have significantly more confidence in the diagnosis of lymphoplasmacytic lymphoma.
In addition, we also have new information is shown that the CXCR4 mutation is present in about 25% of cases of Waldenstrom’s and that these patients are resistant to targeted therapy. IgM MGUS cases about 50 to 80% of these cases will have the MYD88 mutation and as we stated earlier this appears very closely related to lymphoplasmacytic lymphoma and is probably a precursor as opposed to being related to myeloma. Gamma heavy chain disease lacks MYD88 mutation and it is not considered lymphoplasmacytic lymphoma.
Next, we’re going to look at chronic lymphocytic leukemia/small lymphocytic lymphoma and also its counterpart monoclonal B cell lymphocytosis. Monoclonal B cell lymphocytosis was defined as less than 5,000 circulating monoclonal lymphocytes or B cells in the peripheral blood and at the time in 2008 it was unknown if this monoclonal B cell lymphocytosis was a precursor to CLL SLL or not. Interestingly up to 12% of healthy individuals have been found to have monoclonal B cells in their peripheral blood. A lot of times these are very small populations but some individuals would have an actual lymphocytosis. The immunophenotype varied but some of the individuals had an immunophenotype-like CLL.
Now with the 2016 Revision we further divide the monoclonal B cell lymphocytosis into low and high-count cases because we now know that monoclonal B cell lymphocytosis precedes almost all cases of CLL SLL but in the individuals who have a count of less than 500 that there’s really very little chance if any that they will progress to CLL SLL. At this time with what we know currently those individuals do not need routine follow up.
The high count monoclonal B cell lymphocytosis cases those cases that have between 0.5 and 5 or 500 to 5,000 circulating monoclonal B cells in the peripheral blood those individuals need to be monitored on annual basis for progression. Additionally, in the 2016 Revision it eliminates the option to diagnoses CLL with a count of less than 5,000 monoclonal B cells in the peripheral blood even if those patients have cytopenias or disease related symptoms so unless there’s evidence of extramedullary disease these patients would be limited to the category of monoclonal B cell lymphocytosis.
On the lymphoma side of the diagnosis with small lymphocytic lymphoma there’s been interest in the significance of proliferation centers and this article from 2012 demonstrated that many of the cases with proliferation centers can express cyclin D1 and how that should not be a trap to misdiagnose it as a case of mantle cell lymphoma but we know now that large and confluent and/or highly proliferative proliferation centers are an adverse prognostic indicator and we need to recognize that and to assess that in our report.
On the other end of the spectrum, so to speak, the question arises is there an equivalent for small lymphocytic lymphoma compared to CLL with monoclonal B cell lymphocytosis and it appears that there may be monoclonal B cell lymphocytosis equivalent in SLL. There is emerging data that shows that when you have lymph nodes less than 1½ centimeters and those nodes do not have proliferation centers present that they behave very indolently and there’s no evidence of progression, at least in the studies that have been performed to date. There may very well be an equivalent or at least a very good low stage prognosis for these individuals.
One of the most exciting areas in pathology in the last five years has been the use of next generation sequencing and data is emerging with regards to CLL SLL from NGS that p53 NOTCH1, SF3B1, and BIRC3 mutations are associated with a worse prognosis. The literature is not entirely consistent on the clinical significance of this so at this time there is no predictive use I should say, of the information on a day-to-day routine practice basis but this is really something just to be aware of that in the near future may become impactful clinically and certainly there’s a lot of interest and a lot of research right now going into next generation sequencing in lymphomas. That’s something to stay tuned for in the coming months to years. I anticipate that you will see a large amount of data published on this and there’s probably excellent chance that this will become in the near future clinically impactful. Stay tuned on NGS.
Follicular lymphoma has historically been thought of as a largely monolithic entity and in my mind when I think about it I think of it as occurring in older patients and being either low-grade or intermediate-grade in process. It’s the second most common lymphoma behind diffuse large B cell lymphoma and is in the figure earlier shown it usually occurs in older adults and we’ll grade that in either the low-grade category, Grade 1 or Grade 2, or intermediate-grade, Grade 3. This would be based on the number of large cells in the follicle center areas per high power field. Now we’re recognizing that follicular lymphoma is much more heterogeneous than originally thought and we’re going to cover some of these special divisions we now recognize in follicular lymphoma which have significant impact in how you treat and prognosis of patients.
Pediatric follicular lymphoma is a special division of follicular lymphoma that occurs obviously by the name, in younger populations and characteristically has an intermediate-grade morphology. Its promoted from a provisional entity to a definite entity in the 2016 Revision. We typically think of Grade 3 morphology as having a more aggressive course in our classic follicular lymphoma paradigm but in pediatric lymphoma this has an excellent prognosis. It occurs in the head and neck region of children and there’s a very high male predominance of patients.
To make the diagnosis of pediatric follicular lymphoma it must lack the 14;18 translocation which is typically associated with the classic follicular lymphoma encountered in older adults. It also must lack an IRF4 gene rearrangement because that’s a separate entity we’re going to talk about in a little bit.
Oftentimes these cases will have mutations or deletions in the TNFRSH14 gene. There’s also a testicular follicular lymphoma that occurs in the pediatric population with similar morphologic features and outcomes to pediatric lymphoma. Often the treatment is by conservative excision alone without any further treatment.
As I alluded to earlier there is another specific mutation in follicular lymphoma that is a special category and this is the IRF4 translocation. It’s associated with the MUM1 gene. These translocations were identified in a subset of patients typically children or young adults and its associated with an excellent prognosis. You can see here in the bottom left where they tested by FISH on the IRF4 gene to look for the break apart. These cases have strong MUM1 expression and on the bottom right you can see how the age is usually in the young, less than 20-years of age group.
This is a distinct new provisional entity in the WHO 2016 Revision. These cases typically occur in Waldeyer’s ring or in cervical lymph nodes. Patients are low stage and the growth pattern is either Grade 3b follicular lymphoma or diffuse large B cell lymphoma. This will be discussed more in the diffuse large B cell lymphoma section.
We also have an interesting subtype of primary intestinal follicular lymphoma. This is a localized indolent proliferation that has the cytogenetic and immunophenotypic features of classic follicular lymphoma. The morphologic pattern in these are a low-grade and they typically occur in the second part of the duodenum. These cases are very indolent and have a good prognosis.
Also in the similar morphologic category is a new subtype of follicular lymphoma with a diffuse pattern that’s associated with deletion of 1p36. This was identified in a subset of follicular lymphomas that do not have the 14;18 translocation. These lymphomas lack the 14;18 translocation present as a localized mass often in the inguinal region and have an immunophenotype very similar to follicular lymphoma but also co-express CD23 and by definition, these lymphomas have deletions of chromosome 1p36. The deletion of 1p36 is not specific because classic follicular lymphomas can also have these deletions and in that setting, are associated with a more aggressive course and poorer outcome. These cases have diffuse pattern, they lack the BCL-2 gene rearrangement and they present as a localized mass often in the inguinal region.
Diffuse Large B Cell Lymphoma
Next, we’re going to examine the category of diffuse large B cell lymphoma. This is the most common lymphoma diagnosis made in the United States today and its generally considered to be an intermediate-grade process. We know historically that this is really a heterogeneous population of B cell lymphoma disorders that are really collected together under the auspices of morphology.
One of the new provisional entities in the 2016 Revision alludes back to our discussion under follicular lymphoma is this large B cell lymphoma with the IRF4 gene rearrangement. This is a provisional entity. It’s usually found in children or young adults in the head and neck region classically Waldeyer’s ring and the morphology is that of either a diffuse large B cell lymphoma or a Grade 3 follicular lymphoma.
As you can see, we’ve shown this slide earlier from the 2011 Blood article shows in the bottom left the break apart probe demonstrating the IRF4 rearrangement. These cases have strong MUM1 expression and are usually found in younger patients. I will say that if you have a CD10 positive strong MUM1 positive case regardless of age you really may want to consider testing for this rearrangement because you’ll note there are some even at older ages who are found to have this IRF4 gene rearrangement. These are generally localized disease so the patients are not presenting with widespread adenopathy.
To reiterate what we said, these are CD10 positive BCL-2 positive cases, they lack the BCL-2 gene rearrangement. That is very important. Really, the difficulty that you want to be careful about is to distinguish these from the classic follicular diffuse large B cell lymphoma in older patients and that goes back to this idea of whether CD10 is expressed on these MUM1 positive lymphomas because otherwise you’re going to consider this as a non-germinal center or activated phenotype which would be associated with a more aggressive course yet we know that these tend to do very well overall without significant treatment.
other issue that has become quite popular ad is mentioned in the 2008 WHO Classification is distinguishing between the germinal center immunophenotype or an activated or non-germinal center phenotype. The work was done using gene expression profiling showing that diffuse large B cell lymphomas really fall into one of these two categories and that there’s prognostic difference between these, specifically that the germinal center phenotypes behave less aggressively than the activator non-germinal center phenotypes.
Time has moved on and now as part of the 2016 Revision it will require that we subtype or subgroup these lymphomas and immunohistochemistry will be adequate to do that. The reason why it’s now required is because there’s a treatment difference between an activated and a germinal center phenotype so they’ve really done a lot of work to improve the survival probability in those more aggressive lymphomas and because of that work, now we need to make sure that we accurately identify the cell of origin. Again, when you go back and you look at the 2008 WHO we know that the original work was done on gene expression profiling but we also know that gene expression profiling is not routinely available in clinical practice so immunohistochemistry even though that we know that it’s not completely accurate and may mis-categorize probably in the range of 15% of cases, it’s still the best option we have readily available to pathology practices and it doesn’t specify which algorithm you have to use. There have been multiple algorithms developed from different groups. Probably the most common is the Hans classification. You can see there the Blood journal article from 2004 describing this. I personally like the Hans classification because it uses three immunostains which are very commonly found in most pathology laboratories. It takes CD10, if it’s positive it’s a germinal center phenotype. If it’s negative you’ll hit BCL-6, if it’s negative then it’s a non-germinal center phenotype. If BCL-6 is positive you examine the MUM1 expression and if its positive then it’s a non-germinal center phenotype and if it’s negative its germinal center phenotype. These three stains I find very useful and using a Hans classification and that’s what we use in our laboratory today currently.
Again, to emphasize that why it is important to subtype the cell of origin is that we know that the survival difference is significant and now we also know that they’re going to treat these cases differently. It now has predictive information. I will also mention that the cutoff typically used for these immunostains is 30%.
A new area that’s received a lot of attention in the last several years has been examining the protein expression of Myc and also BCL-2 in these diffuse large B cell lymphomas. We know between BCL-2 and Myc that 20 to 50% of lymphoma cases will express one, the other or both. The definition of what is considered positive in these cases, generally Myc you look for greater than 40 or 50% of the cells expressing Myc and greater than 50% for BCL-2. These cases that express both by immunohistochemistry are called double expresser lymphoma.
You can see here in this picture from Dr. Swerdlow’s ASH article showing the strong Myc expression and strong BCL-2 expression in a diffuse large B cell lymphoma. This is also to emphasize that measuring protein expression is not reflective of gene translocations and in fact, that we know most of these cases do not harbor Myc or BCL-2 translocations. However, they are more aggressive and they really kind of are in between diffuse large B cell lymphoma and a high-grade B cell lymphoma with regards to their aggressiveness. This should be considered a prognostic indicator. It is not a separate category however, for diffuse large B cell lymphoma. As an aside and caveat, our lab does not personally do the testing currently it’s not really been requested by our clinicians and it’s a little bit controversial whether or not this should be routinely performed. My personal opinion would be that you use the information in conjunction with your clinicians locally and how they’re going to treat patients clinically.
EBV positivity; in the 2008 WHO there is a new provisional entity titled EBV positive diffuse large B cell lymphoma of the elderly. Probably the most controversial part of this is what they define as elderly meaning that you’re greater than 50 years of age and pretty much everybody that I’ve talked to about this takes great offense in using the term elderly and the cutoff of 50 years. What we do know is that those cases have a worse prognosis. In the 2016 WHO the entity has really been renamed EBV positive diffuse large B cell lymphoma NOS because we know that it can occur in younger patients.
What we want to highlight here this is a nice article from 2013 in Blood, is that from a gene expression profiling standpoint kind of analogous to the gene expression profiling studies that we did with germinal center and activated phenotypes with diffuse large B cell lymphoma is that the EBV positive lymphomas are distinct and you can see that in the bottom left where you see all of the red gathered around these genes and they’re really multiple components of the NF-kappa B pathway. It’s important to remember that the EBV positive cases have a unique gene expression profile and also that they’re being recognized in younger patients.
There's a bit of controversy with regard to the prognosis particularly in younger patients. The literature is a bit conflicting but we know that some cases are not as aggressive as others. The NOS, not otherwise specified, designates that such cases cannot be classified as another EBV-specific disorder and the new provisional entity described is the EBV positive muco-cutaneous ulcer and these are secondary to iatrogenic immunosuppression typically. This is a nice article from American Journal of Surgical Pathology in 2010 that highlights this disorder. Remember that these cases are not going to be classified in another EBV positive disorder. Lymphomatoid granulomatosis would be another example.
As a closing remark with regards to diffuse large B cell lymphoma I would just point out an aside with regards to CD30 expression. There are a lot of lymphoma cases or subsets of lymphomas in other malignancies, not limited to lymphoma, they express CD30. Classically we think of anaplastic large cell lymphoma or Hodgkin lymphoma but other lymphomas can acquire CD30 expression. CD30 in and of itself is not a specific marker but there is a targeted therapy brentuximab which is an anti-CD30 agent so there is interest from clinicians typically when patients have failed standard therapy to see if a tumor expresses CD30 to try that drug against them. Keep that in mind when considering particularly your relapsed cases.
High-Grade B Cell Lymphoma
Lastly today we’re going to discuss a category of high-grade B cell lymphoma. I can say personally with regards to the 2008 WHO Classification I thought one of the more confusing aspects of that was this interface between intermediate and high-grade lymphomas. Specifically, the lines became blurred between what used to be Burkitt lymphoma or lymphoblastic lymphoma/leukemia and an intermediate-grade process like diffuse large B cell lymphoma we began to realize that some of these subtypes of diffuse large B cell lymphoma behaved as high-grade B cell lymphomas. We introduced some categories that accounted for that but yet there was not a lot of confidence with regards to how to definitively treat these patients because we know that treatment between a high-grade B cell lymphoma and intermediate-grade lymphoma is completely different. I think one of the nice aspects of the 2016 Revision is that there’s much more clarity brought between these two categories.
The first category is high-grade B cell lymphoma with Myc and BCL-2 and/or BCL-6 translocations. This is a so-called double/triple hit B cell lymphoma. These cases exclude cases that would meet criteria for follicular lymphoma or lymphoblastic lymphoma. The 2008 Classification placed these in this category of B cell lymphoma unclassifiable with features intermediate between diffuse large B cell lymphoma and Burkitt lymphoma.
With regards to these it’s important to remind us that as stated earlier, immunohistochemistry expression of Myc and BCL-2 or BCL-6 are not surrogate markers for identification of the translocation. It’s a little bit controversial I suppose with regards to whether you should test all large B cell lymphomas for the double or triple hit. In our laboratory, we do test all diffuse large B cell lymphomas that has been requested by our clinicians. I think that many laboratories are, I would say the majority of laboratories are but some do not agree necessarily with this test all strategy. There’s still a little bit of controversy with regards to that I suppose but I would just advise any pathologist to work with your clinicians and understand what their needs are and how they’re going to treat their patients and base your testing on that.
We know that these double/triple hit lymphomas are very aggressive, more so than Burkitt lymphoma or diffuse large B cell lymphoma highlighted by the chart to the right. We also have the category of high-grade B cell lymphoma not otherwise specified. Previously the provisional category in the 2008 WHO was B cell lymphoma unclassifiable with features intermediate between diffuse large B cell lymphoma and Burkitt lymphoma and that’s been eliminated. The high-grade B cell lymphoma NOS category is going to include cases with blastic features are truly intermediate between Burkitt and diffuse large B cell lymphoma but they don’t have the Myc translocation with BCL-2 and/or BCL-6 translocations. Those cases are separate and will be handled separately. This is a bit, I hate to use the word trash bag, but the cases that don’t fit anywhere else fall into this category.
Burkitt lymphoma has remained relatively unchanged in the 2016 Revision although we do have more information about this entity specifically we’ve identified TCF3 and ID3 mutations in up to 70% of sporadic cases and 40% of endemic cases. If you have a Myc negative Burkitt lymphoma they may be associated with the 11q alteration and such testing would be indicated. This is a new provisional entity entitled Burkitt-like lymphoma with 11q aberration.
To sum up the high-grade B cell lymphoma category I think this figure out of the review article in Blood is very helpful because it shows your morphologies across the top of how they align and then the definitive classification in the 2016 Revision. You can see if you have a large B cell lymphoma and it has a double or triple hit it’s going to fall in that special category. Otherwise it remains intermediate-grade lymphoma. If you have a blastic morphology in a diffuse large B cell lymphoma then it may fall into one of several categories dependent upon the molecular abnormalities identified. If it fulfills the criteria for Burkitt lymphoma it may fall in there, double or triple hit. If it doesn’t fit any of the specific categories then it will fall into the category of a high-grade B cell lymphoma not otherwise specified.
In just a short summary with regards to immunohistochemistry I think it’s important to include a couple of caveats. This is largely targeted to intermediate and high-grade B cell lymphomas but some of this can be applied more broadly. The first is that I always do a CD3 and CD20 on all my cases of lymphoma. This is really to me an XY axis for evaluation relative to the other markers I’m going to use. CD20 is also important to document in patients with a B cell lymphoma who have been treated with rituximab because those patients may become resistant with downregulation of the CD20 receptor. CD5 is important to identify transformed mantle cell lymphomas. Those will express cyclin D1. It can also be a prognostic marker in diffuse large B cell lymphomas by themselves. The CD10, BCL-6 MUM1 combination, of course as we discussed earlier in the Hans classification to differentiate CD4ll of origin for diffuse large B cell lymphomas which has become very important from a treatment standpoint and it now required in the 2016 WHO Revision.
TDT, if you have a blastic cell population it’s important to exclude a lymphoblastic lymphoma TDT is very helpful. EBV, remember our category of EBV-related large B cell lymphomas. It is no longer the category of EBV in the elderly patients greater than 50, but all ages. Ki-67 can be a very helpful qualitative measurement and is prognostic in certain cases. BCL-2 and Myc are prognostic markers and can be used in the identification of this double expresser lymphoma.
Those are just some of the stains that one may find helpful. These are certainly not required in any way but are an important menu to have available so that one can decide on any particular case what is most important.
This concludes our presentation for today. I thank you for your attention. In November, we’ll be examining the myeloid portion of the 2016 WHO Revision. I look forward to speaking to you again