Testing Strategies for Identifying Eosinophilia in Patients With Myeloid/Lymphoid Neoplasms

In an interview with Targeted Oncology, Annette S. Kim, MD, PhD, discussed the updates to the NCCN guidelines on myeloid/lymphoid neoplasms with eosinophilia, along with how to bring new pathology strategies for these disease into clinical practice.

With the rise of tyrosine kinase (TK) inhibitors, identifying and diagnosing TK fusion genes can help to improve the diagnosis and treatment of patients with myeloid/lymphoid neoplasms with eosinophilia. This has promoted updates to the National Comprehensive Cancer Network (NCCN) clinical practice guidelines

According to Annette S. Kim, MD, PhD, an associate professor of pathology at Harvard Medical School and associate pathologist at Brigham and Women’s Hospital, secondary causes of eosinophilia must be ruled out. Karyotypes for both bone marrow and tissue biopsies.

After this is done, confirmatory testing such as fluorescence in situ hybridization (FISH) or nested reverse transcription polymerase chain reaction (RT-PCR) should be performed.

In an interview with Targeted OncologyTM, Kim discussed the updates to the NCCN guidelines on myeloid/lymphoid neoplasms with eosinophilia in greater detail, along with how to bring new pathology strategies for these disease into practice.

TARGETED ONCOLOGY™: Can you explain the diagnostic criteria provided in this guideline from the NCCN on myeloid lymphoid neoplasms?

KIM: The problem in making these diagnoses is that there's a very broad differential for eosinophilia, including some rare hereditary things, but also really primary and secondary or otherwise known as reactive causes for eosinophilia. And so, the most important thing in making the diagnosis is to have a clinical suspicion based on the peripheral blood or bone marrow eosinophilia, other CBC abnormalities or organ damage, lab values, such as elevated tryptase, B12, LFTs, LDH, or uric acid etc. The NCCN guidelines recommend a full evaluation for reactive or secondary causes of eosinophilia, including allergic disorders, infections, such as parasitic infections, immunodeficiency syndromes, and eosinophilia associated with other neoplasms including some advanced solid tumors, but especially T-cell lymphomas and less commonly B-cell lymphomas or even other myeloid neoplasms, where the eosinophilia are not of the abnormal clone, but the tumors producing some cytokines that produce the eosinophilia.

In my role as a pathologist, we really focus in on the diagnostic criteria for the identification of these primary causes of eosinophilia including the myeloid lymphoid neoplasms with eosinophilia, as well as chronic eosinophilic leukemia that occurs in the absence of these tyrosine kinase gene rearrangements. It's critical for the workup of these entities to do a bone marrow biopsy and to sample any potential extra medullary sites of disease. And of course, with the bone marrow biopsy or the tissue biopsies, you should absolutely do a karyotype.

In addition, there are some additional immunohistochemical stains you might like to perform on the bone marrow such as CD117, CD25, and tryptase, as well as reticulum and trichrome for fibrosis, as these entities can often occur with increased associated mass cells, even if they don't meet criteria for systemic mast cytosis, as well as fibrosis. Unlike many of the myeloid neoplasms, which are really complex diagnostic algorithms with major and minor criteria, really the main diagnostic criterion is simply the identification of one of these tyrosine kinase fusions. So, you do this very extensive workup to rule out a bunch of things. But the bottom line is to make this diagnosis, you have to identify the fusion.

How do you identify these areas for your arrangements in these patients based on this guideline?

That is actually the really tricky part. The NCCN guidelines recommend upfront karyotype type for all recurrent myeloid lymphoid neoplasms with eosinophilia. Many of this recurrent translocation will be abnormal by karyotype, but still should be followed up by some sort of confirmatory testing, whether it be FISH or RTPCR. The exception to this is the PDGFR alpha phip 1L1 rearrangement which is cryptic by karyotype in almost all cases and for which FISH is absolutely required.

The problem comes is that many of these tyrosine kinases are quite promiscuous and have a number of different partners. This is especially true for PDGFR beta FGFR1 and even JAK2, and therefore some of the translocations can be cryptic by karyotype as well. And so, due to the large number of potential partners and the breadth of different breakpoints in either fusion partner targeted molecular methods, such as RTPCR are not particularly useful for diagnostic screening purposes but can certainly be used for confirmatory purposes as well as monitoring disease. Therefore, if you don't want to miss these potentially cryptic rearrangements, the NCCN currently recommends performing a wide panel of sort of FISH break apart studies. However, I feel like that is not particularly cost effective. And I would recommend that going beyond just the NCCN guidelines which talk about FISH and potentially RTPCR and instead performing RNA or DNA based next generation sequencing designed to specifically to identify rearrangements.

In your institution, how are these patients diagnosed?

In my practice, we've seen many cases now where the myeloid lymphoid neoplasms with eosinophilia were missed by karyotype, especially when the level of suspicion was low, or the presentation was atypical. So, we've routinely started routing these cases for either DNA or RNA based NGS testing designed to detect rearrangements. We had a recent example of my institution, where a pathologist was absolutely convinced that this had to be some form of a myeloid lymphoid neoplasm with eosinophilia, but PDGFR alpha, beta, FGFR1 FISFH didn't pick it up. But next generation sequencing identified a novel kit rearrangement that appears to be functioned very similar to those of other entities or provisional entities within this category. So, the take home message really is that if you don't look for it, you can miss it. And so, this really requires a multidisciplinary conversation between the clinician and the pathologist to make sure both sides of the street are aware of the potential for the diagnosis and what the clinical suspicions are so that appropriate testing can be performed. In addition, since many of the morphologic features that would point you to the structure are really coming from the bone marrow or tissue biopsy and the pathology and pathologist interpretation thereof, it's really critical for the pathologist to be involved in sort of cytogenetic and molecular ordering, or at least to push for broad sequencing as needed.

What are your recommendations for ensuring adoption of best practices for diagnosis and community practices?

I think communication between the clinician and the pathologist is really critical. Many community oncology practices send their bone marrows out to large commercial pathology groups, often with minimal history. And this really ties the hands of the pathologists from pushing for additional testing if the pathology is worrisome because the pathologist may not know some of the other clinical features, for instance, whether there's an elevated B12, or whether there's a seismic mass. So, communication with your friendly neighborhood pathologist is perhaps one of the most important factors in making this diagnosis.

The second important factor is really to empower the pathologist to order or to recommend ordering some FISH or NGS panels that will allow for the detection of these rearrangements that were not detected by karyotype in the setting of a strong clinical pathologic suspicion. For instance, there are certain histologic features that characterize some of these entities, for instance, large aggregates of immature erythroid cells and PCM1-JAK2 fusions comes immediately to mind. And so, the pathologist is the one who's going to sort of first raise the suspicion based on some of these histologic features, if it wasn't clear from the clinical presentation. As I mentioned earlier, I think it's becoming increasingly more cost effective to consider NGS testing rather than broad FISH panels. And while NGS may not be locally available, there are certainly many commercial labs that do this type of testing, and it could be sent out on those cases. Now, there are some pros and cons that go with NGS as well. But I think that increasingly, it's becoming more effective at translocation identification.

The pathologist and the clinicians really have to therefore maintain this high level of suspicion for an MLNEO and seek out appropriate confirmatory cytogenetic and molecular testing. Since these diseases are really rare, the number of kinases and fusion partners are numerous. And since reactive eosinophilia is also very, very common, there's unfortunately a low pretest probability for any such testing. However, testing with a broad panel is really clinically indicated given the dramatic therapeutic responses to targeted kinase inhibitors in many of these cases and the extremely poor prognosis of some of these patients. The real take home message is, if you don't look for it, you're going to run the risk of missing it. You need to always maintain that very high level of suspicion for these entities.

REFERENCE:

Myeloid/lymphoid neoplasms with eosinophilia and TK fusion genes, Version 3.2021. J Natl Compr Canc Netw. 2020;18(9):1248–1269 doi: 10.6004/jnccn.2020.0042