Clinical and molecular characteristics of XPO1 mutations in patients with chronic lymphocytic leukemia

To the Editor: The XPO1 (exportin) gene (also referred to as chromosome region maintenance 1; CRM1) is a karyopherin that exports proteins and RNA fragments from the nucleus into the cytoplasm [1,2]. The human XPO1 gene, located on chromosome 2 (2p15), is believed to encode an oncogenic protein since many of the molecules exported by XPO1 into the cytoplasm are associated with either known tumor-suppressor genes, such as TP53 or FOXO3A, or transcription factors that contribute to cell proliferation and survival such as IkB-a [3] or STAT3 [4] whose accumulation in the nucleus results in cell death. The binding of XPO1 to various proteins is mediated by recognizing the leucine-rich nuclear export signals (LR-NES) on the N-terminus of snurportin 1 (SNUPN) forming a nuclear pore complex or a cargo, thereby transporting proteins out of the nuclear membrane. Overexpression, deregulation, or dysfunction of XPO1 has been reported in various types of cancer [5]. XPO1 is a therapeutic target in CLL [6], and selective inhibitors of nuclear transport (SINE) such as selinexor are now being investigated in clinical trials in CLL. Next-generation sequencing (NGS)-based mutational studies in patients with chronic lymphocytic leukemia (CLL) have demonstrated that XPO1 is recurrently mutated in <5% of patients with CLL [7]. The integration of commonly occurring gene mutations with current prognostic factors including cytogenetic aberrations improves prognostication for CLL patients [8–10]. In this study, we investigated the clinical significance of XPO1 mutations in patients with CLL. Peripheral blood and/or bone marrow cells from 486 CLL patients were analyzed using an amplicon-based NGS panel of 53 cancer-related genes using MiSeq system (Illumina, San Diego, CA). Details of this panel are listed in Supporting Information Table S2. In 13 patients, we also performed NGS-based mutation analysis with a panel composed of CLL-related genes including NOTCH1, SF3B1, POT1, BTK, and BIRC3. The study was approved by Institutional Review Board, and written informed consent was obtained from the study patients. Conventional cytogenetic and fluorescence in situ hybridization (FISH) studies were performed. Survival analysis was conducted by using GraphPad Prism version 6.00 for Windows (GraphPad Software, La Jolla, CA). XPO1 mutations were detected in 38 of 486 (7.8%) CLL patients. Of these 38 patients, 16 were treatment-naive and 22 were previously treated. Clinical characteristics of the patients are summarized in Supporting Information Table S1. The median age of patients with XPO1 mutations was 59 years (range 37–81 years). Nine patients (24%) were 65 years of age. Among the 38 patients with XPO1 mutations, 28 of 38 (74%) had early Rai stage disease, whereas 10 (26%) had advanced disease. Fifteen patients (40%) had sole del(13q), 11 (29%) had del(11q) [10 of whom had coexisting deletion del(13q)], three had del(17p), and three had trisomy 12. Complex karyotype was detected in two patients. Thirty patients (79%) had unmutated immunoglobulin heavy chain variable region (IGHV) status and three were mutated. There was no predilection to a specific VH gene usage. The median follow-up duration of patients after the detection of XPO1 mutation was 13.5 months (range 0–74 months). At the last follow-up, one patient with relapsed/refractory CLL had died of disease progression and infection, and 37 patients were alive; the median overall survival duration of these patients was not reached (Supporting Information Fig. S1). Among the 16 previously untreated patients, seven required treatment, and nine did not. Among the 22 previously treated patients, 18 were pretreated with fludarabine-based therapies, and none were pretreated with ibrutinib. Twenty-one of the 22 patients required subsequent treatment, and one remains under observation. Of these 21 patients, 15 (71%) received ibrutinib-based therapy and 13(87%) responded. Overall, one patient transformed into Hodgkin’s lymphoma. The clinical and molecular characteristics of the patients with XPO1 mutations are detailed in Table I. All XPO1 mutations were missense mutations in exon 15 leading to protein alteration in codon 571. The most frequent XPO1 mutation was p.E571K (n5 29), followed by p.E571V, p.E571G and p.E571Q in 12 patients (three of whom had coexisting p.E571K with p.E571G, p.E571K, and p.E571V mutations). The median allelic frequency of XPO1 mutation was 20% (range 1.6–46.2%). Nine patients (23%) had a mutant allelic burden of 10%. Concurrent CLL specific mutations were detected in various genes in addition to XPO1 in 10/13 patients (Supporting Information Fig. S2), most frequently in NOTCH1, SF3B1, and TP53 (annotated in Table I). We compared patients who had XPO1-mutated CLL and unmutated IGHV gene to CLL patients with unmutated IGHV gene (n5 136) without XPO1 or TP53 mutations (control group). Supporting Information Fig. S1B,C shows that the patients with XPO1-mutated CLL and the patients in the control group had a similar TTFT and survival duration (P5NS). Although the leukemogenic mechanisms of mutations in patients with CLL are unclear, the recurrent nature of XPO1 mutation in CLL strongly supports its involvement in the pathogenesis of the disease. We found that XPO1 mutations are commonly found in patients with unmutated IGHV (79%), early Rai stage (73%), CD38 overexpression, ZAP-70 positivity, and del(13q). Few patients had coexisting TP53 mutations. During an approximate 14-month follow-up after testing, we observed that XPO1 mutations may not be associated with poor prognosis. Overall, our results agree with two previously published reports [7,8]. Jeromin et al. [8] found that among 33(3%) of 969 patients who had XPO1 mutations, the majority (30 patients [91%]) had unmutated IGHV status. We were unable to determine the mutations in NOTCH1 in our panel in all patients. However, using a different gene panel in 13 patients, we found that NOTCH1 mutations were seen in 5/10 (50%) patients with XPO1 mutations. On comparing the survival duration and TTFT in a matched control group to XPO1-mutated patients, we found that patients with XPO1 fared better. The majority of patients who needed treatment responded to ibrutinib-based therapy. This suggested that although patients with XPO1 mutations had markers of poor prognosis, the overall outcome was not poor in those receiving ibrutinib. Low frequency of XPO1 mutations and good response to ibrutinib suggest that the XPO1 mutation may not be a marker of poor prognosis in patients with CLL.