diffuse large B-cell lymphoma

[SillyStudent]

---

Members don't see this ad.

DLBCLs are derived from peripheral mature B cells at the GC or post-GC stage of differentiation. Now in this scheme I see DLBCLs can also be derived from B-immunoblasts.
In what group are DLBCLs derived from B-immunoblasts classified according to the Hans model of cell origin; GCB or non-GCB?

My first thought was non-GCB but how can these cells express post-GC markers if they didn't go through a germinal center reaction?

 

[cmz]

The so-called "immunoblastic" morphology goes along with having a possible non-GC immunophenotype (in most cases), so I think you're first inclination is correct. I'm not going to give you the direct answer to your question, but I would crack open an immunology text book or google and get a feel for how B-cells mature normally. Then I would think about their neoplastic counterparts ... (in a way, you already answered your question).

 

[Sulfinator]

Isn't immunoblastic simply a morphologic descriptor? I don't think saying a tumor is the immunoblastic variant of DLBCL implies that it was derived from an immunoblast. And I would expect that most immunoblastic variants would classify as the ABC subtype, as defined by gene expression profiling (which is the gold standard). Forget the Hans algorithm . . . it is sufficiently crappy and can be forgotten.

 

[SillyStudent]

The immunoblastic type is indeed a morphological type. I'm confused because I also found multiple text books stating that DLBCL can be derived from immunoblasts. I thought a DLBCL was always derived from cells at germinal center or post-germinal center differentiation but an immunoblast doesn't participate in a germinal center reaction.

 

[cmz]

Isn't immunoblastic simply a morphologic descriptor? I don't think saying a tumor is the immunoblastic variant of DLBCL implies that it was derived from an immunoblast. And I would expect that most immunoblastic variants would classify as the ABC subtype, as defined by gene expression profiling (which is the gold standard). Forget the Hans algorithm . . . it is sufficiently crappy and can be forgotten.

Correct, it is a morphologic descriptor. That is why I said "in most cases". The Hans algorithm is nice and all and may accurately profile ~90% of cases. The other 10%, of course, don't quite fit the gene expression profiling to the tee. Hence, there are other algorithms like the Tally method, Choi, etc. It never ends.

In private practice, the oncologists don't seem to care. It seems like if you do more than a CD20 and CD3 and maybe a Ki67, you're going overboard. Everyone seemed to care more before Rituximab, though 🙂 I always try to at least make a best effort to sub-classify any DLBCL I get because I know that a good portion of my cases inevitably end up going to MD Anderson for review.

As for the statement above, "I thought a DLBCL was always derived from cells at germinal center or post-germinal center differentiation but an immunoblast doesn't participate in a germinal center reaction"... you have to remember that DLBCL, by definition, is simply "a neoplasm of large B lymphoid cells with nuclear size equal to or exceeding normal macrophage nuclei or more than twice the size of a normal lymphocyte." If you read the 2008 WHO you'll notice that they expand on a few different subtypes of DLBCL that were never fully classified before in the 2001 WHO and so forth. So, in essence, when you say "DLBCL", you really are just talking about a morphologic definition.

And about immunoblasts... I consider immunoblasts (and so do most immunologists, I hope) to be activated B or T-cells. Everything I say from this point forward is related to B-cell immunoblasts only, but these cells are mature and have undergone V(D)J rearrangement. So, I think that your picture has to be at least somewhat flawed or is missing a few arrows. B-Immunoblasts have the potential to give rise to plasma cells. We all know what plasma cells do, right? They make tons and tons of Ig, which also has to be somewhat specific. How did they get to be so specific in the first place? Think about early B-cell maturation in the germinal center and the process known as negative selection.

I know I made a weird circular argument, but I haven't had my coffee yet 🙂

 

[pathstudent]

I still do the Hans Classifier as that is how I trained, but I don't put it in the top line. Whether is GC-DLBLC or Non-GC-DLBCL, the patient is going to the same treatment and followed the same way and they will be cured or they will relapse and if they relapse, second line therapies will be employed.

I think much more important is picking up the double hit lymphomas as well as correctly diagnosing the BCLUWFIBDLBCLABL as many oncologists will treat those more aggressively right from the gate. I had one case that looked like an overall garden variety GC-DLBCL with a Ki-67 of 70%. It was a lymph node found incidentally during abdominal surgery in one of my physician colleagues. This was back in 2010 before I routinely did FISH on all DBLCL, but I decided to do it on him because I had plenty of tissue that I sent for flow and karyotype. It came back with a triple hit (BCL-2, BCL-6 and MYC rearrangement). He went to the lymphoma experts in Rochester and to MDA and had an allogeneic BMT locally. He never cleared the disease and was dead in 9 months from the time of diagnosis. It was a realy eye-opener, because as I said, it looked like your every day GC-DLBCL. So now I get the FISH on every single DLBCL if possible.

 

[Over9000]

I still do the Hans Classifier as that is how I trained, but I don't put it in the top line. Whether is GC-DLBLC or Non-GC-DLBCL, the patient is going to the same treatment and followed the same way and they will be cured or they will relapse and if they relapse, second line therapies will be employed.

I think much more important is picking up the double hit lymphomas as well as correctly diagnosing the BCLUWFIBDLBCLABL as many oncologists will treat those more aggressively right from the gate. I had one case that looked like an overall garden variety GC-DLBCL with a Ki-67 of 70%. It was a lymph node found incidentally during abdominal surgery in one of my physician colleagues. This was back in 2010 before I routinely did FISH on all DBLCL, but I decided to do it on him because I had plenty of tissue that I sent for flow and karyotype. It came back with a triple hit (BCL-2, BCL-6 and MYC rearrangement). He went to the lymphoma experts in Rochester and to MDA and had an allogeneic BMT locally. He never cleared the disease and was dead in 9 months from the time of diagnosis. It was a realy eye-opener, because as I said, it looked like your every day GC-DLBCL. So now I get the FISH on every single DLBCL if possible.

It's crazy how that acronym is actually an acronym.

 

[cmz]

I still do the Hans Classifier as that is how I trained, but I don't put it in the top line. Whether is GC-DLBLC or Non-GC-DLBCL, the patient is going to the same treatment and followed the same way and they will be cured or they will relapse and if they relapse, second line therapies will be employed.

I think much more important is picking up the double hit lymphomas as well as correctly diagnosing the BCLUWFIBDLBCLABL as many oncologists will treat those more aggressively right from the gate. I had one case that looked like an overall garden variety GC-DLBCL with a Ki-67 of 70%. It was a lymph node found incidentally during abdominal surgery in one of my physician colleagues. This was back in 2010 before I routinely did FISH on all DBLCL, but I decided to do it on him because I had plenty of tissue that I sent for flow and karyotype. It came back with a triple hit (BCL-2, BCL-6 and MYC rearrangement). He went to the lymphoma experts in Rochester and to MDA and had an allogeneic BMT locally. He never cleared the disease and was dead in 9 months from the time of diagnosis. It was a realy eye-opener, because as I said, it looked like your every day GC-DLBCL. So now I get the FISH on every single DLBCL if possible.

I usually reserve ordering FISH for BCL2, BCL6, and MYC for cases that have a really high proliferative index (>85%+) and/or show other high grade features (e.g. Burkitt-like morphology). Double-hit lymphomas behave pretty aggressively and already carry a worse prognosis than your garden-variety DLBCL-NOS, and are usually treated with aggressive chemo + BMT upfront. I wonder if there are any studies out there that specifically address the utilization of FISH in all DLBCL. It seems like it's a waste of resources to order them for every single case.

 

[cmz]

BCLUWFIBDLBCLABL is just an awful acronym! ... but I suppose it's pretty accurate 🙂

 

[pathstudent]

I usually reserve ordering FISH for BCL2, BCL6, and MYC for cases that have a really high proliferative index (>85%+) and/or show other high grade features (e.g. Burkitt-like morphology). Double-hit lymphomas behave pretty aggressively and already carry a worse prognosis than your garden-variety DLBCL-NOS, and are usually treated with aggressive chemo + BMT upfront. I wonder if there are any studies out there that specifically address the utilization of FISH in all DLBCL. It seems like it's a waste of resources to order them for every single case.

I have seen many double hit lymphomas with Ki-67 under 90%. It is anecdotal but I suggest doing it routinely. Compared to the cost of the numerous PET scans and chemo and radiation or whatever, it is a fricking drop in the bucket.

When I was a resident I went to the a lymphoma work shop and I remember hearing Nancy Harris say "Do not try to save money when working up a lymphoma". A few IHC stains or FISH probes that end up not contributing are irrelevant in the cost of the patient's care, but if relevant, and you didn't order them, then you are ****ed and so is the patient.

 

[cmz]

I have seen many double hit lymphomas with Ki-67 under 90%. It is anecdotal but I suggest doing it routinely. Compared to the cost of the numerous PET scans and chemo and radiation or whatever, it is a fricking drop in the bucket.

When I was a resident I went to the a lymphoma work shop and I remember hearing Nancy Harris say "Do not try to save money when working up a lymphoma". A few IHC stains or FISH probes that end up not contributing are irrelevant in the cost of the patient's care, but if relevant, and you didn't order them, then you are ****ed and so is the patient.

I'm not so sure that this is the standard of practice. It might be your standard, though. In fact, I would argue that most hematopathologists (including the WHO overlords) wouldn't order FISH unless the tumor showed high-grade features (among other things, including flow and the clinical picture). Only a small fraction of DLBCL are actually "double hit lymphomas," so routine FISH testing on every case will likely yield an inordinate amount of negative results. Seems like a waste.

If the lymphoma comes back (often a poor prognostic sign for a DLBCL), the patient will probably end up getting a BMT anyway.

 

[Sulfinator]

BCLUWFIBDLBCLABL is just an awful acronym! ... but I suppose it's pretty accurate 🙂

Interestingly, my source who was present at the recent SH/EAHP that met at the beginning of USCAP tells me that the new term for the forthcoming WHO (expected to be out in 2016) will simply be "B-cell lymphoma, unclassifiable" (BCL-U). The between DLBCL and BL part will just be implied.

Not sure about the one that is between DLBCL and cHL.

 

[pathstudent]

Interestingly, my source who was present at the recent SH/EAHP that met at the beginning of USCAP tells me that the new term for the forthcoming WHO (expected to be out in 2016) will simply be "B-cell lymphoma, unclassifiable" (BCL-U). The between DLBCL and BL part will just be implied.

Not sure about the one that is between DLBCL and cHL.

Thank God a new one is coming out. The 2008 one is so so so out of date

 

[pathstudent]

I'm not so sure that this is the standard of practice. It might be your standard, though. In fact, I would argue that most hematopathologists (including the WHO overlords) wouldn't order FISH unless the tumor showed high-grade features (among other things, including flow and the clinical picture). Only a small fraction of DLBCL are actually "double hit lymphomas," so routine FISH testing on every case will likely yield an inordinate amount of negative results. Seems like a waste.

If the lymphoma comes back (often a poor prognostic sign for a DLBCL), the patient will probably end up getting a BMT anyway.

Trust me. If you had a dlbcl you would want to know if it had rearrangements. And how exactly do you grade dlbcl? if it is not high grade, that would imply it is a low grade dlbcl. That seems odd.

I prefer to break B cell NHL into two categories indolent (generally incurable but won't kill you for awhile) and aggressive (will kill you relatively soon but potentially curable).

I would group dlbcl, b lymphoblastic, burkitt and the intermediates as aggressive. I don't know criteria to group them into low and high grade.

And like I said. I had a case that you would consider low grade dlbcl (no starry sky ki67 around 70%) and he had a bloody triple hit. Totally lethal. And my stuff was reviewed at MDA and Rochester. Both signed it out exactly as me. Dlbcl with a triple hit.

 

[cmz]

Trust me. If you had a dlbcl you would want to know if it had rearrangements. And how exactly do you grade dlbcl? if it is not high grade, that would imply it is a low grade dlbcl. That seems odd.

I prefer to break B cell NHL into two categories indolent (generally incurable but won't kill you for awhile) and aggressive (will kill you relatively soon but potentially curable).

I would group dlbcl, b lymphoblastic, burkitt and the intermediates as aggressive. I don't know criteria to group them into low and high grade.

And like I said. I had a case that you would consider low grade dlbcl (no starry sky ki67 around 70%) and he had a bloody triple hit. Totally lethal. And my stuff was reviewed at MDA and Rochester. Both signed it out exactly as me. Dlbcl with a triple hit.

There's no such thing as "high" vs "low" grade DLBCL. I don't "grade" DLBCL. No one does, as far as I know. I just report out the Ki-67 index in a comment along with any unusual morphologic features that would warrant further testing. I generally diagnose B-cell NHL the same way you do, I assume. You're just quoting one case that happened to be triple hit. ONE case. If I had your triple hit DLBCL with all the ancillary tests/studies (FISH, Flow, etc), I would sign the thing out in exactly the same fashion as how you and the consultants signed it out. Of course, you're also leaving a huge part of the story out, namely the clinical history and presentation.

Speaking about the Ki-67 being "only 70%"... this is also kind of arbitrary unless you used a graticule or used a sophisticated computer program to come up with this number 😉 We all know that Ki-67 as interpreted with the human eye has excellent reproducibility 🙂

 

[pathstudent]

There's no such thing as "high" vs "low" grade DLBCL. I don't "grade" DLBCL. No one does, as far as I know. I just report out the Ki-67 index in a comment along with any unusual morphologic features that would warrant further testing. I generally diagnose B-cell NHL the same way you do, I assume. You're just quoting one case that happened to be triple hit. ONE case. If I had your triple hit DLBCL with all the ancillary tests/studies (FISH, Flow, etc), I would sign the thing out in exactly the same fashion as how you and the consultants signed it out. Of course, you're also leaving a huge part of the story out, namely the clinical history and presentation.

Speaking about the Ki-67 being "only 70%"... this is also kind of arbitrary unless you used a graticule or used a sophisticated computer program to come up with this number 😉 We all know that Ki-67 as interpreted with the human eye has excellent reproducibility 🙂

No you said you only order FISH it if has "high-grade" features. If it doesn't have "high-grade" features, presumably you would say that it has "low-grade" features.

I just never heard anyone refer to DLBCL as being high grade until I met you.

Yeah and my Ki-67 is reproducible. I count 1000 tumor cells with my bone marrow counter and mark if they are positive or negative.

 

[cmz]

No you said you only order FISH it if has "high-grade" features. If it doesn't have "high-grade" features, presumably you would say that it has "low-grade" features.

I just never heard anyone refer to DLBCL as being high grade until I met you.

Yeah and my Ki-67 is reproducible. I count 1000 tumor cells with my bone marrow counter and mark if they are positive or negative.

You haven't met me, so there's that...

...and in my first post, I clarified what I meant by saying "with high grade features" by adding the caveat, "(e.g. Burkitt-like morphology)." Notice how I didn't say just, "high grade"? It's a matter of style as to how anyone signs out their reports.

How much of a troll are you? If you routinely interpret all of your Ki-67 stains by counting 1000 tumor cells with your bone marrow counter, then I am clearly dealing with someone very, very special. I am done with this thread. Cheers!

 

[SillyStudent]

There is evidence indicating that ABC-DLBCL may develop from extrafollicular B-cells with high levels of AID.

Cattoretti G, Shaknovich R, Smith PM, Jäck HM, Murty VV, Alobeid B. Stages of germinal center transit are defined by B cell transcription factor coexpression and relative abundance. J Immunol. 2006;177:6930–9.

 

[Sulfinator]

There is evidence indicating that ABC-DLBCL may develop from extrafollicular B-cells with high levels of AID.

Cattoretti G, Shaknovich R, Smith PM, Jäck HM, Murty VV, Alobeid B. Stages of germinal center transit are defined by B cell transcription factor coexpression and relative abundance. J Immunol. 2006;177:6930–9.

Sure, sure. Nevertheless, DLBCL remains a heterogeneous group of lymphomas that we have yet to discover the best ways to parse out. ABC vs. GBC is a step in the right direction, though there's not a great facile way (that has sufficient reproducibility) to identify which is which in routine clinical practice. We're just not that sophisticated yet, but we are getting there.

 

[univlad]

Can you look for double hit lymphomas using IHC for BCl-6 or myc at all or is FISH required?

 

[Ziehl-Neelsen]

Can you look for double hit lymphomas using IHC for BCl-6 or myc at all or is FISH required?

There are a number of papers that suggest scoring systems utilizing IHC that have increased PPV with regard to identifying the cases that will be positive for the MYC/BCL-2 double hit by FISH.

http://jco.ascopubs.org/content/30/28/3460.full.pdf

http://jco.ascopubs.org/content/30/28/3452.full.pdf

http://jco.ascopubs.org/content/30/28/3452.full.pdf

What's more, the first 2 papers suggest that an increased IHC score is a independent predictor of poorer survival regardless of the results of the FISH studies, though I've yet to meet an oncologist who will recommend against first line R-CHOP in the absence of FISH confirmed rearrangement of the genes. In the end it's oncologist dependent. I've shown this evidence to some oncologists who are ok with not sending for FISH if the score is low and everything jives. I work with other oncologists who want FISH on every DLBCL regardless, so I don't do the stains on their cases as it doesn't influence them.

 

[Sulfinator]

You can't call something a double hit lymphoma unless you do FISH to demonstrate the translocations. Many people try to use IHC results to help them decide whether or not FISH is worth pursuing in a given case (i.e., they look for certain IHC characteristics that might make it more likely that the lymphoma is a double hit lymphoma), but there is no consensus about how best to do this. The IHC characteristics of double hit lymphomas are variable enough that any attempt to use IHC to decide which cases to FISH and which not to FISH will miss some double hit cases. I have heard that some labs do FISH on every would-be DLBCL case, but that seems to be too burdensome to me (in terms of turn around time and use of healthcare dollars).


Sent from my iPhone using SDN mobile

 

[cmz]

I've had cases where I thought that this will be a double-hit for sure based purely on morphology AND some of the IHC characteristics (e.g. super high Ki-67, cMyc IHC in >60% of neoplastic cells, etc). It happens. I personally don't order triple FISH on every case unless I feel there is a need to. Or perhaps I should start ordering this on every single DLBCL... $$ for tech-only FISH right?