Evolution of Clonal Hematopoiesis after Discontinuation of Therapy for Multiple

Introduction:

Survival outcomes have improved dramatically for patients (pts) with multiple myeloma (MM), though cumulative exposure to therapies is associated with increased risk for second hematologic malignancies. Clonal hematopoiesis (CH) can be a precursor lesion for MM therapy-related leukemia (Saygin, Blood Cancer Discov 2024). Incidence of CH in MM pts prior to autologous stem cell transplant (ASCT) was reported at 22% with an allelic frequency (AF) cut-off >1%. (Mouhieddine, Nature Commun 2020). It is unclear if cessation of therapy can reverse the progression of CH and reduce the risk for leukemia. We hypothesize that discontinuation of therapy can lead to regression of CH clones in MM, and aim to identify patients whose risk for secondary leukemia may be reduced by treatment discontinuation.

Methods:

We studied prospective samples from pts in the MRD2STOP trial, in which pts with MM who achieved flow cytometric and next-generation sequencing (NGS)-based measurable residual disease (MRD)-negativity discontinued maintenance therapy (Derman ASCO 2024). DNA was extracted from CD138-depleted bone marrow cells obtained at baseline (prior to discontinuation of MM therapy), and at 12- and 24-months follow-up. DNA was sequenced with a hybrid capture–based targeted gene panel at a high depth of 2000x coverage for 22 genes and 95% of CH mutations found in the general population. The assay reaches an AF sensitivity of 0.01%.

Results:

A total of 95 samples from 38 individual pts were analyzed. Median age was 65 years (range, 39-84), 50% were female. Patients were 71% White, 18% Black, 5% Hispanic, 3% Asian and 3% more than one race; 34% had high-risk MM by cytogenetics. 97% of patients received one prior line of therapy prior to discontinuation. Induction therapy was a quadruplet regimen in 45%, triplet regimen in 53%, and doublet regimen in 2%. After induction, 61% received ASCT, and 13% received multi-drug post-transplant consolidation prior to maintenance therapy. Most (95%) patients received lenalidomide maintenance, and average duration of consolidation and/or maintenance therapy prior to discontinuation was 44 months (range, 13-90).

Using an AF cut-off of 0.1%, we identified CH mutations in 97% of pts at baseline with the following distribution of mutated genes: TP53 (83%), DNMT3A (33%), SF3B1 (13%), IDH1 (13%), ASXL1 (11%), GNAS (11%), KRAS (11%), NRAS (8%), TET2 (5%), JAK2 (3%), PPM1D (3%), SRSF2 (3%). Among all pts, 29% had one, 39% had two, 24% had three and 3% had four mutations in CH-associated genes in their marrow.

Using an AF cut-off of 1%, 42% of pts had CH clones. The frequency of CH with TP53 clones at AF >1% was 21% (8 out of 38). Therefore, the median AF of TP53 clones (0.18%, range 0.1%-14.5%) was significantly smaller than DNMT3A (0.33%, range 0.1%-13.8%) and ASXL1 (6.4%, range 0.1%-23.6%) clones, but comparable to other mutated genes. There were no significant correlations between age, gender, ASCT status, or duration of treatment and the frequency of mutated genes.

In assessment of serial samples with a baseline AF cutoff of 0.1%, the size of TP53-mutant CH clones was stable after treatment discontinuation with median AF of 0.18% at baseline, 0.17% at 12 months (p=0.40), and 0.2% at 24 months post-discontinuation (p=0. 79). Four patients with TP53-mutant clones AF >1% at baseline demonstrated clonal regression over time, and one demonstrated progression. Other CH clones showed similar patterns of stability over 24 months, but clonal regressions (DNMT3A, ASXL1, KRAS in 1 pt each) and expansions (ASXL1 in 2 pts) were seen in some cases.

One patient in this cohort developed therapy-related acute lymphoblastic leukemia two years after lenalidomide withdrawal, but no driver mutations were identified in the ALL clone.

Conclusions:

In summary, we demonstrated a high prevalence of CH in MM patients prior to therapy discontinuation. This included a disproportionately high frequency of TP53-mutant CH, but not the traditional age-related CH (i.e., DNMT3A, TET2, ASXL1), in MM pts who were exposed to current standard of care therapies. These CH clones remained stable even after treatment was discontinued, and more patients will be added to further assess the clonal evolution of TP53 CH. Longer follow up is needed to determine the full impact of these findings on risk for therapy related leukemia.