Ups and downs of CHIP

In this issue of Blood, Berger et al present a longitudinal study of the clonal trajectory of preleukemic mutations in patients who developed therapy-related myeloid neoplasms (t-MNs) after autologous stem-cell transplantation. The authors show that t-MN driver mutations are often detectable as clonal hematopoiesis of indeterminate potential (CHIP) many years before patients develop t-MNs, and they demonstrate complicated patterns of clonal expansion and evolution over time, leading to the development of t-MNs.¹ 

t-MNs are one of the most devastating complications from cytotoxic chemotherapy and ionizing radiation therapy. Although overall incidence of t-MNs is <10% among patients with cancer, t-MNs often end with fatal outcomes.²  When a patient develops a t-MN, we as oncologists struggle to share the bad news with the patient, who has already fought or been fighting the primary malignancy. It is hard to imagine how difficult it must be for the patient. Because current treatment modalities are ineffective in curing t-MNs, there is a real-world unmet need to predict and prevent t-MNs before their occurrence.

Recently, several studies identified that preleukemic mutations or chromosomal copy number alterations were detectable in the blood samples of patients with cancer before treatment.^3-7  Detection of preleukemic clonal hematopoiesis was associated with an increased risk of t-MNs. Similarly, preleukemic mutations were also detectable in autologous stem-cell apheresis samples, which was associated with significantly increased risk of t-MNs and non–t-MN–related mortality in patients with lymphoma.⁸  The study by Berger et al, which accompanies this commentary, provides data on how these preleukemic mutations evolve over time, particularly in response to cytotoxic chemotherapies or hematopoietic growth factors.

The authors describe clonal kinetics of preleukemic mutations in 7 patients who developed t-MNs by analyzing multiple sequential blood or marrow samples taken before the t-MN development. Although the number of patients studied here was too small to derive meaningful patterns, the heterogeneous clinical courses of the 7 patients generated interesting questions about how clonal hematopoiesis behaves in response to external agents. In 1 patient, clonal expansion of SMC1A along with an increase in mean corpuscular volume followed treatment with danazole and erythropoietin. What do we know about clonal hematopoiesis and response to hematopoietic growth factors? In another patient, a TP53-mutated clone gradually increased while acquiring another TP53 mutation during thalidomide and cyclophosphamide treatments. What is the response of clonal hematopoiesis to immunomodulatory imide drugs? In some patients, the authors observed clear expansion of preleukemic mutations under the selective pressure of chemotherapy, whereas there were small independent clones that remained stable. Why did some clones remain stable or disappear while others expanded under the pressure of chemotherapy? Ultimately, this leads to a more clinically relevant question: which clones have a high risk of developing into t-MNs, and which will remain stable for a long time? Longitudinal analyses of clonal hematopoiesis in a large number of patients as well as biological studies that address mechanisms of transformation from clonal hematopoiesis to t-MNs are needed to answer these questions.

There were several other findings that were noteworthy in the report by Berger et al. First, preleukemic mutations were also detectable in a T-cell fraction in 1 patient. Although this needs to be verified in larger cohort, it suggests that CHIP originates at early hematopoietic stem-cell or progenitor stages, which has been previously suggested by other studies.^9,10  Second, the authors concluded that lymphoma did not arise from CHIP, because the mutational landscape was completely different. Third, p53-overexpressed cells were detectable in marrow by immunohistochemistry in patients with TP53-mutated CHIP. Because mutations are detected “digitally,” if these cells truly represent TP53-mutated CHIP, immunohistochemistry may help visualization of CHIP in marrow. Lastly, the authors found that t-MNs had higher numbers of mutations compared with de novo MDS but did not have a specific mutation signature; however, interpretation of these findings requires caution, because the data were derived from whole-exome sequencing, and they contradict previous findings.⁷  Nonetheless, the study by Berger et al has helped advance our understanding of how CHIP behaves during chemotherapy and contributes to the development of t-MNs.