Beware clonal haematopoiesis following chronic lymphocytic leukaemia treatment

Abstract

Clonal myeloid haematopoiesis co-existing with chronic lymphocytic leukaemia (CLL) is not rare and is associated with high risks for development of therapy-related myeloid neoplasms (t-MNs) after cytotoxic chemotherapy. This has serious implications for the use of these therapies in newly diagnosed CLL patients. Therapy-related myeloid neoplasms are devastating conditions. Their development following cytotoxic chemotherapy for various neoplasms often ultimately leads to treatment-refractory acute myeloid leukaemia and is usually lethal for afflicted patients. These neoplasms often develop following exposure to alkylating agents or anthracyclines,1 but they have been reported following single-agent fludarabine.2 T-MNs typically present after prior cytotoxic therapy given in an adjuvant context or for relatively indolent neoplasms, and thus their development is particularly distressing for many otherwise cancer-free or cancer-controlled patients. Effective long-term control of t-MNs is difficult to achieve, and at present always requires allogeneic haematopoietic transplant. The paper by Voso et al.3 regarding clonal haematopoiesis of indeterminant potential (CHIP) as a risk factor for t-MNs following chemotherapy for CLL is thus timely and worthy of attention from all physicians who manage CLL. CHIP has (unsurprisingly) been associated with increased risks of t-MN. The current report is, however, the first reported study of t-MN following therapy for CLL. It is worth noting that this experience is based upon 15 patients who were selected only for the availability of pretreatment genomic material. Amongst the relevant 13 patients whose t-MN followed the use of cytotoxic therapy, a prior mutation consistent with myeloid CHIP was present in 10, or 77%. Mutations compatible with CHIP were found in 34 of 285 cases of CLL not developing t-MN. The authors thus appropriately conclude that CHIP seems to be a significant risk factor for the development of t-MN following cytotoxic therapy for CLL. This work presents a detailed description of the mutations found in prechemotherapy haematopoietic material. The authors are conservative in terms of attributing these mutations to the CLL or the myeloid elements found in the available pretreatment cells. They correctly observe that mutations of TP53 and SF3B1 could in principle arise either from clonal CLL cells or from a myeloid CHIP clone, and thus cannot be firmly identified as having driven the development of the t-MN. Data found in their table S4, however, show that in two patients (#6 and #8), mutations in TP53 (two cases) and in SF3B1 (one case), were identical in pretreatment samples of CLL containing cells and in the t-MN material. The same mutated sequences were found in these two paired samples in both cases and for both genes. Thus, although we cannot be completely certain, it seems very likely that in fact the pretreatment mutation was found in myeloid cells that were subsequently transformed in response to the exposure to cytotoxic agents, leading to the t-MN. These findings have important implications for the management of CLL and other neoplasms. First, CHIP is reasonably common in patients who are candidates to receive cytotoxic agents. Pre-existing CHIP clearly poses very significant risks for the development of t-MN and most of these secondary neoplasms will prove to be lethal for affected patients. Thus, particularly in older patients with higher risks of CHIP, it seems reasonable to consider screening patients for myeloid clonal states prior to offering cytotoxic therapy for management. Notably, there are now alternative systemic therapies available for CLL (and many other neoplasms) which do not involve cytotoxic therapy and thus should not pose the same risk of potential secondary neoplasm. Another and potentially provocative conclusion is that mutations of TP53 or SF3B1 found in pretreatment blood or other tissues from CLL patients may not necessarily represent changes in the CLL clone but may be situated in associated myeloid cells. This issue in it itself may affect decisions concerning CLL management, as TP53 mutations are associated with poor prognosis and as this finding in a CLL patient may potentially increase the desire for aggressive therapy. Finally, in such patients the use of cytotoxic chemotherapy in disease management might have disastrous downstream consequences, leading to a lethal t-MN. The development of effective systemic targeted therapies for CLL not employing cytotoxic or mutagenic agents has revolutionized the therapy of CLL. The application of these agents has already greatly reduced the use of potentially leukaemogenic therapies. Recently, however, it has been asserted that “In fit patients with immunoglobulin heavy chain mutation, it is still acceptable to use the chemotherapy regimen with fludarabine, cyclophosphamide, and rituximab (FCR) and, in those who [are] not fit or are not IgVH-mutated, bendamustine-rituximab regimen”.4 Given the availability of effective non-cytotoxic agents for management of essentially all CLL patients, haematologists and oncologists managing this group of highly treatable but fragile patients should take pause, and note the findings described by Voso et al.