Our group (Welch, Cell 2012) previously showed that hematopoietic stem and progenitor cells (HSPCs) acquire somatic mutations with age. This produces a genetically heterogeneous HSPC population with each HSPC possessing its own unique set of mutations. Later work from our group (Xie, Nature Medicine 2014) and others (Genovese, Jaiswal, NEJM 2014) demonstrated that some these mutations may provide HSPCs with a fitness advantage, allowing them to clonally expand over time in healthy individuals. We recently published data (Wong, Nature 2015) suggesting that cytotoxic therapy can select for HSPC clones with TP53 mutations, resulting in their clonal expansion and contributing to the subsequent development of therapy-related AML/MDS. From these data, we hypothesized that the intensive cytoreductive chemotherapy used to treat AML poses a significant selection pressure on a patient's non-malignant HSPC population, favoring HSPCs with specific somatic mutations and potentially resulting in oligoclonal hematopoiesis even after elimination of the founding AML clone.

To test this hypothesis, we performed enhanced exome sequencing on cryopreserved bone marrow cells from 25 adult de novo AML patients (who received a "7+3" regimen for induction of remission) at time of their initial diagnosis, at first morphologic remission (~day 30), and at long-term follow up (at first relapse or during a prolonged first remission) (Klco, JAMA, in press). In 15 patients, we observed genetic clearance of the AML founding clone at the time of first morphologic remission (defined as all AML founding clone mutations declining to a variant allele frequency (VAF) < 2.5%). Surprisingly, in 5 of the 15 patients exhibiting clearance of their AML founding clone, we observed a concomitant expansion of a non-malignant clonal population during cytoreductive therapy, resulting in long-lived clonal hematopoiesis. Somatic mutations harbored by these expanding hematopoietic clones were validated with a high-coverage PCR-based sequencing approach. In contrast to the studies highlighting clonal hematopoiesis in individuals unexposed to chemotherapy, patients with evidence of persistent clonal hematopoiesis after cytoreductive therapy (median age = 52.2 years) were similar in age to patients without such evidence (median age = 54.1 years).

The majority of these "rising clones" harbored somatic mutations in genes frequently mutated in AML such as DNMT3A, TET2 and TP53. Using next-generation sequencing and droplet digital PCR, we determined that in all of the patients with an expanding non-malignant clone, the clone was, in fact, present in the initial AML diagnosis sample at very low VAFs (0.007-0.75%). These populations rapidly expanded with chemotherapy, comprising 13-57% of the total hematopoietic population upon its completion. In all 4 of cases with sample availability, these clones remained at an expanded level a year or more after initial chemotherapy exposure. These results suggest that certain non-malignant HSPCs, having previously acquired specific aging-related somatic mutations, may gain a competitive fitness advantage after cytoreductive therapy, expand, and persist long after the completion of chemotherapy.

Two of the five patients with clonal non-leukemic hematopoiesis post-chemotherapy relapsed. In both patients, the relapsed AML clone evolved from the original AML founding clone and did not involve the non-malignant clonal population, which also persisted at relapse. Both patients re-achieved morphologic remission with salvage therapy. A post-salvage therapy bone marrow sample was available in one of the cases. Interestingly, it showed that the patient's non-malignant clonal population expanded even further with salvage therapy, eventually comprising almost 80% of the total bone marrow cells.

These results show that non-malignant oligoclonal hematopoiesis is common in AML patients after cytoreductive chemotherapy, with non-malignant HSPCs carrying certain somatic mutations often gaining a fitness advantage and expanding. The long-term clinical consequences of oligoclonal hematopoiesis after cytoreductive chemotherapy are unknown but are likely to be different from oligoclonal hematopoiesis developing in healthy elderly individuals. Additional studies will be required to define the mechanisms by which certain HSPCs gain a fitness advantage after cytoreductive chemotherapy.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution