Mrinal M. Patnaik, MD, is assistant professor of medicine and assistant professor of oncology at the Mayo Clinic Division of Hematology in Rochester, MN. His primary research interests are myelodysplastic syndromes (MDS) and chronic myelomonocytic leukemia (CMML).
He has conducted extensive research into CMML, with more than 20 publications in the past 2 years regarding this rare and aggressive clonal bone marrow disorder of late adulthood. Dr Patnaik’s research focus is in epigenetic and chromatin dysregulation in CMML, and transcriptional abnormalities stemming from mutations in epigenetic-regulator genes like TET2, ASXL1, and EZH2.
In this question-and-answer session, Cancer Therapy Advisor asked Dr Patnaik about the evolving understanding of CMML’s molecular underpinnings, diagnosis, prognostication, and progression.
Cancer Therapy Advisor: CMML is categorized as an MDS/myeloproliferative neoplasm (MPN) “overlap syndrome,” like juvenile myelomonocytic leukemia or atypical chronic myeloid leukemia, exhibiting features of MDS and MPN.1 Does the distinction between MDS- and MPN-phenotype CMML reflect differences in the molecular underpinnings of the disease?
Dr Patnaik: Yes, the MDS/MPN overlap syndromes like CMML are unique entities. So for a long time, we have borrowed from either MPN or MDS strategies to manage them. With the advent of molecular testing and more awareness, we are now certain that these are unique entities and need to be recognized.
In the MDS/MPN overlap, in adults, the most common disease that we see is CMML. Within each disease there is a continuum of progression. Patients with CMML present with either MDS-like or MPN-like features. Based on the underlying disease biology and surrogate markers such as bone marrow blasts and cytogenetics, eventually 20% to 30% of patients progress to acute myeloid leukemia (AML).
This secondary AML is very resistant to treatment and is associated with poor patient outcomes.
Cancer Therapy Advisor: Why is that?
Dr Patnaik: The molecular signature of “de novo” AML is different from secondary AML. Secondary AML has accrued a lot more epigenetic dysregulation, splicing abnormalities, and a lot of times, the patients undergo cytogenetic clonal evolution. In many cases, they’ve selected out through receiving prior treatments, such as hypomethylating agents or other epigenetic modifiers, that make it harder to treat from a disease standpoint—and from a host perspective, these patients have already been through so much that their ability to withstand induction chemotherapy is poor. In addition, drug resistance mechanisms are also upregulated in patients with secondary AML.
Cancer Therapy Advisor: What recent clinical advances have occurred in CMML?
Dr Patnaik: We have made advances in molecular biology and have been able to demonstrate the mutational landscape and the impact of these mutations in patients with CMML.
Cancer Therapy Advisor: Which gene mutations are associated with CMML? Are there prognostic biomarkers?
Dr Patnaik: Numerous prognostic models have been developed for patients with CMML.1 The 2 contemporary molecularly integrated models include the Molecular Mayo Model and the GFM model.
This disease does have a unique molecular signature. Approximately 60% of cases have mutations in TET2, which controls methylation and hydroxymethylation.
About 40% to 50% have mutations in SRSF2, which is a splicing factor that has been shown to be deleterious in MPN and younger patients with CMML (age < 65 years).
RELATED: Older Patients With CML at Higher Risk for Mortality, Vascular Events
However, the worst mutation acquired is ASXL1, seen in approximately 40% of patients with CMML. Frame shift and nonsense ASXL1 mutations predict for a shortened overall survival. Additionally, the presence of these mutations potentially predict for poor responses to hypomethylating agents. RAS mutations are often associated with an MPN-like phenotype.1
Although univariate analysis studies with RAS mutations have found inferior outcomes in CMML, these findings have not been substantiated in multivariate models.