Molecular Profiling of Myeloid Progenitor Cells in Multi-Mutated Advanced Systemic Mastocytosis Identifies KIT D816V As a Distin

Home , Mast cell leukemia, Myelodysplastic syndrome

ORIGINAL ARTICLE Molecular proﬁling of myeloid progenitor cells in multi-mutated advanced systemic mastocytosis identiﬁes KIT D816V as a distinct and late event

M Jawhar1,8, J Schwaab1,8, S Schnittger2, K Sotlar3, H-P Horny3, G Metzgeroth1, N Müller1, S Schneider4, N Naumann1, C Walz3, T Haferlach2, P Valent5, W-K Hofmann1, NCP Cross6,7, A Fabarius1 and A Reiter1

To explore the molecular profile and its prognostic implication in systemic mastocytosis (SM), we analyzed the mutation status of granulocyte–macrophage colony-forming progenitor cells (CFU-GM) in patients with KIT D816V+ indolent SM (ISM, n = 4), smoldering SM (SSM, n = 2), aggressive SM (ASM, n = 1), SM with associated clonal hematologic non-mast cell lineage disorder (SM-AHNMD, n = 5) and ASM-AHNMD (n = 7). All patients with (A)SM-AHNMD (n = 12) carried 1–4 (median 3) additional mutations in 11 genes tested, most frequently TET2, SRSF2, ASXL1, CBL and EZH2. In multi-mutated (A)SM-AHNMD, KIT D816V+ single-cell-derived CFU-GM colonies were identified in 8/12 patients (median 60%, range 0–95). Additional mutations were identified in CFU-GM colonies in all patients, and logical hierarchy analysis indicated that mutations in TET2, SRSF2 and ASXL1 preceded KIT D816V. In ISM/SSM, no additional mutations were detected and CFU-GM colonies were exclusively KIT D816V−. These data indicate that (a) (A)SM-AHNMD is a multi-mutated neoplasm, (b) mutations in TET2, SRSF2 or ASXL1 precede KIT D816V in ASM-AHNMD, (c) KIT D816V is thus a phenotype modifier toward SM and (d) KIT D816V or other mutations are rare in CFU-GM colonies of ISM/SSM patients, which might explain at least in part their better prognosis.

Leukemia (2015) 29, 1115–1122; doi:10.1038/leu.2015.4

INTRODUCTION syndromes.11–14 The current understanding of their molecular 11,13,15,16 Systemic mastocytosis (SM) is characterized by abnormal proli- significance has been thoroughly reviewed elsewhere. feration and accumulation of mast cells (MCs) in various tissues, Besides the classification of SM, the newly identified molecular predominantly skin, bone marrow (BM) and visceral organs. The lesions and aberration profiles may play a role in disease severity, 8–10,17,18 extent of organ infiltration and subsequent organ damage is prognosis and treatment responses in SM patients. the basis for the classification of SM into indolent SM (ISM), Despite such significant progress in our knowledge on the smoldering SM (SSM), SM with associated clonal hematologic non- molecular pathogenesis of myeloid neoplasms, it remains to be MC lineage disease (SM-AHNMD), aggressive SM (ASM) and MC elucidated how these mutations develop and accumulate in leukemia (MCL).1–3 In patients with SM-AHNMD, the SM compo- individual patients over time and finally contribute to disease nent can resemble ISM, ASM or even MCL. Depending on the evolution, phenotype, progression and prognosis. In multi- subtype of SM, cell source (BM or peripheral blood, PB) and assay mutated hematologic neoplasms, single-cell assays on colony- sensitivity, an acquired mutation in the receptor tyrosine kinase forming hematopoietic progenitor cells have recently revealed a KIT, usually KIT D816V, is detectable. Using PCR-based assays with complex clonal architecture and heterogeneous evolution includ- high sensitivity, this mutation is detectable in more than 80–90% ing acute myeloid leukemia (AML), MPN, MDS and MDS/ – of all SM patients.4–6 The multilineage involvement by KIT D816V MPN.14,19 24 In SM, several lines of evidence suggest that multi- and the KIT D816V allele burden have an important impact on mutated and thus potentially more aggressive subclones develop disease phenotype and prognosis.6,7 within the non-MC lineage compartment, which also explains why The presence of additional mutations in genes encoding for many of these patients develop an AHNMD or at least marked signaling molecules (CBL, JAK2, KRAS, NRAS), transcription factors myelodysplasia or signs of myeloproliferation. The aim of the (RUNX1), epigenetic regulators (ASXL1, DNMT3A, EZH2, TET2)or present study was to examine the mutational profile of colonies splicing factors (SRSF2, SF3B1, U2AF1) has recently been reported grown from granulocyte–macrophage colony-forming progenitor in KIT D816V+ SM patients in advanced disease.8–10 These cells (CFU-GM) and microdissected mature cells (CD117+, CD3+ or mutations are not specific for SM as they were also identified in CD15+) obtained from patients with various subtypes of SM. The other myeloid neoplasms, including myelodysplastic syndromes results of our study show that advanced SM with AHNMD is a (MDS), myeloproliferative neoplasms (MPN) or MDS/MPN overlap multi-mutated stem cell neoplasm with a phenotype modification

1Department of Hematology and Oncology, University Hospital Mannheim, Mannheim, Germany; 2Munich Leukemia Laboratory, Munich, Germany; 3Department of Pathology, Ludwig-Maximilians-University, Munich, Germany; 4Department of Clinical Chemistry, University Hospital Mannheim, Mannheim, Germany; 5Division of Hematology and Ludwig Boltzmann Cluster Oncology, Department of Internal Medicine I, Medical University of Vienna, Vienna, Austria; 6Wessex Regional Genetics Laboratory, Salisbury, UK and 7Faculty of Medicine, University of Southampton, Southampton, UK. Correspondence: Professor A Reiter, Department of Hematology and Oncology, University Hospital Mannheim, Theodor-Kutzer-Ufer 1-3, 68167 Mannheim, Germany. E-mail: [email protected] 8These authors contributed equally to this work. Received 24 July 2014; revised 17 October 2014; accepted 7 November 2014; accepted article preview online 8 January 2015; advance online publication, 30 January 2015 Molecular proﬁling in systemic mastocytosis M Jawhar et al 1116 toward SM because of a late acquisition of KIT D816V, whereas Mannheim, University of Heidelberg) as part of the ‘German Registry on ISM/SSM seems to be not or only rarely affected by mutations at Disorders of Eosinophils and Mast Cells’. All patients gave written informed the CFU-GM level. consent.

PATIENTS AND METHODS Qualitative and quantitative assessment of KIT D816V in BM and PB fi Diagnosis and classi cation of SM Qualitative and quantitative assessments of KIT D816V and KIT D816V allele Diagnosis of SM requires the presence of one major (multifocal dense burden at the RNA level (expressed allele burden) and DNA level in BM infiltrates of MC in BM biopsies and/or in sections of other extracutaneous (n = 16) and PB (n = 3) samples were performed using allele-specific organs) and at least one minor or the presence of at least three minor quantitative real-time PCR analyses as previously described.6,7 A previously criteria (425% atypical cells on BM smears or spindle-shaped MC published CMML cohort was additionally screened for the presence or infiltrates, KIT D816V point mutation in BM or other extracutaneous absence of KIT D816V mutations.13 organs, expression of CD2 and/or CD25 by MC in BM, PB or another extracutaneous organ and baseline serum tryptase concentration 420 μg/l).1,25–27 All BM biopsies were evaluated by two reference Targeted next-generation sequencing analysis pathologists of the ‘European Competence Network on Mastocytosis' Next-Generation Deep Amplicon Sequencing by 454 FLX amplicon (H-PH, KS). Diagnosis of ASM was based on the presence of one or more chemistry (Roche, Penzberg, Germany)11 was performed to investigate C-findings (cytopenia with an absolute neutrophil count o1×109 /l, 18 candidate genes at known mutational hotspot regions as previously hemoglobin o10.0 g/dl or platelets o100 × 109 /l, hepatomegaly with described. Detailed mutation analysis of patients #1, #6, #8 and #14 were impaired liver function, palpable splenomegaly with signs of hypersplenism, reported previously (Table 1).10 malabsorption with significant hypoalbuminemia and/or significant weight loss 410% over the last 6 months).1,25,27 In SM-AHNMD, BM morphology and Colony-forming unit granulocyte–macrophage (CFU-GM) assay PB counts revealed an associated myeloid neoplasm, for example, chronic myelomonocytic leukemia (SM-CMML), MDS/MPN unclassified (SM-MDS/ Methylcellulose (0.9%) was used as semi-solid matrix and was supple- MPNu) or chronic eosinophilic leukemia (SM-CEL).1,2,28 Diagnosis of SSM was mented with 30% fetal bovine serum albumin (FBS), 1% BS albumin, 0.1 M based on the presence of 2/3 diagnostic B-findings ((a) BM MCs 430% and 2-mercaptoethanol and recombinant human GM-CSF (100 ng/ml; Metho- serum tryptase 4200 μg/l, (b) organomegaly, (c) myeloproliferative or Cult, StemCell Technologies, Cologne, Germany). BM mononuclear cells + 5 myelodysplastic features in the BM), in the absence of C-findings.1,25 (MNC) and/or CD34 cells were seeded in the culture mixture (1 × 10 cells/ml or 5 × 103 cells/ml MethoCult) in 35-mm Petri dishes (10 per group) and were incubated at 37 °C in a humidified atmosphere with 5% CO2 Patients´ characteristics for 14 days. Colony counting was performed according to standard + Nineteen KIT D816V patients (male, n = 10; female, n = 9, median age 65 methodology under an inverted microscope. Single-cell-derived CFU-GM years, range 47–76 years) with World Health Organization-based colonies (100–300 cells per colony) were detached from the dishes and SM were evaluated. Clinical characteristics are summarized in Table 1. diluted in phosphate-buffered saline. Classification revealed ISM (n = 4), SSM (n = 2), ASM (n = 1) and (A)SM- AHNMD (n = 12; in detail: ASM-CMML, n = 4; ASM-MDS/MPNu, n = 2; ASM- MDS, n = 1; SM-CMML, n = 2; SM-MDS, n = 2; SM-polycythemia vera (n = 1). Genotyping of CFU-GM The study design adhered to the tenets of the Declaration of Helsinki and DNA extracted from single-cell-derived CFU-GM colonies was submitted was approved by the relevant institutional review board (Medical Faculty to whole-genome amplification (Repli-G, Qiagen, Hilden, Germany).

Table 1. Clinical characteristics and mutational proﬁle of 19 SM patients

Pat.# Age Sex Dx AHNMD A/T M/E Karyotype MC inﬁltration Serum KIT D816V TET2 SRSF2 ASXL1 CBL EZH2 Other (years) in BM (%) tryptase AB (%) mutationsa (μg/l)

21 56 F ISM −−/ −−/ − 46,XX 5 20 20 15 58 F ISM −−/ −−/ − 46,XX 5 36 15 18 58 M ISM −−/ −−/ − 46,XY 5 26 20 20 48 F ISM −−/ −−/ − 46,XX 10 54 26b 16 47 M SSM −−/ −−/ − 46,XY 40 240 11 17 61 M SSM −−/ −−/ − 46,XY 30 200 35 19 75 F ASM − +/ −−/ − 46,XX 60 489 51 13 53 M SM MDS − /+ − / − 46,XY 20 91 12b + 23 76 M SM MDS +/ −−/ − Complex 10 19 8 SF3B1 14c 70 F SM CMML − / − +/+ 46,XX 10 23 19 ++ + + 25 71 F SM PV − / −−/ − 46,XX 20 113 27 JAK2 12 74 F SM CMML − / − +/+ 46,XX 20 88 45 + + 7 70 M ASM CMML − /+ +/ − 46,XY 60 265 19b ++ + 1c 73 M ASM CMML +/+ +/ − 46,XY,ins(9) 30 44 45 + + + 8c 75 M ASM MDS − /+ − / − 46,XY,del(5q) 40 250 33 + + ETV6/U2AF1 3 65 F ASM CMML +/+ +/ − 46,XX 30 102 45 ++ + + 6c 63 F ASM CMML +/+ +/ − 46,XX 30 450 60 RUNX1 24 75 M ASM MDS/MPNu +/+ − / − 46,XY 50 743 37 + IDH2 27 65 M ASM MDS/MPNu +/ −−/ − 46,XY 30 204 51 + ++ Abbreviations: AB, allele burden; AHNMD, associated clonal hematologic non-mast cell lineage disease; ASM, aggressive SM; BM, bone marrow; CMML, chronic myelomonocytic leukemia; Dx, diagnosis; F, female; ISM, indolent SM; M, male; MC, mast cell; MDS, myelodysplastic syndrome; MPNu, myeloproliferative neoplasm unclassiﬁed; Pat., patient; PV, polycythemia vera; SM, systemic mastocytosis; SSM, smoldering SM. Serum tryptase in μg/l; A/T: anemia ⩽ 10.0 g/dl (+), 410.0 g/dl (− ), platelets ⩽ 100 ×109 /l (+), platelets 4100 × 109 /l (− ); M/E: monocytosis 41 ×10e9 /l (+), ⩽ 1×109 /l (− ), eosinophilia 41×109 /l (+), ⩽ 1×109 /l ( − ). aOther mutations, which were only detected once in individual patients. bAB in peripheral blood; ++ two different mutations. cPatients being described in our previous publication.10

Leukemia (2015) 1115 – 1122 © 2015 Macmillan Publishers Limited Molecular profiling in systemic mastocytosis M Jawhar et al 1117 Mutational status was analyzed by mutation-specific PCR followed by JAK2,RUNX1,SF3B1,SRSF2,TET2and U2AF1; Supplementary Table 1). Sanger Sequencing. As CFU-GM colonies are derived from a single myeloid The most frequently affected genes were TET2 (n = 6), SRSF2 progenitor cell, the colonies are expected to be either positive (50% in (n = 5), ASXL1 (n = 5), CBL (n = 4) and EZH2 (n = 3; Figure 1). An IDH2 case of heterozygosity and 100% in case of homozygosity) or negative. mutation in a patient with SM-AHNMD is the first to be reported in fi Quanti cation of the mutational burden is therefore not needed. SM. Seven patients (ISM n = 4; SSM n = 2; ASM n = 1) were negative for additional mutations (Table 1). The number of additional Cytogenetic analysis and fluorescence in-situ hybridization analysis mutations was 0 in ISM/SSM/ASM, whereas the median number Cytogenetic analyses of at least 20 Giemsa-banded BM metaphases was 1 (range 1–4) in SM-AHNMD and 3 (range 1–4) in ASM- (24 and/or 48 h culture) were analyzed and interpreted according to the AHNMD. Nine patients carried ⩾ 2 additional mutations. The most 29 International System for Human Cytogenetic Nomenclature. If necessary, frequent combinations of mutations were KIT-TET2-SRSF2 (n =3) fl chromosome banding analysis was combined with uorescence in-situ and KIT-TET2-ASXL1 (n = 3). hybridization analysis according to the manufacturer's instructions (Metasystems, Altlussheim, Germany).30 SRSF2, TET2 and ASXL1 mutations in T cells fl The four SRSF2 mutations were all Pro95His; the somatic nature of Mutation analysis in microdissected and uorescence activated 13 cell-sorted cells this mutation has been proven elsewhere. In order to determine the somatic nature of TET2 (n = 9 mutations in six patients) and Tissue processing and detection of mutations in immunostained cells (KIT fi D816V and TET2 in CD117+,CD3+ and CD15+ cells) microdissected from ASXL1 (n = 5 mutations in ve patients) mutations, screening of + paraffin-embedded tissue or sorted by fluorescence activated cell sorting five patients was performed in CD3 T cells following micro- (TET2 and ASXL1 in CD3+ T cells) was performed as previously described.31,32 dissection (n = 3) and/or fluorescence activated cell sorting (n =3) and additionally by cross-referencing with the COSMIC33 database. fi Statistical analyses T cells of one of ve patients were TET2 mutation positive, whereas no TET2 or ASXL1 mutations were identified in T cells of Statistical analyses were performed using Excel or GraphPad Prism Software (version 5, GraphPad, La Jolla, CA, USA). the remaining four patients. For the patients without isolated T cells, TET2 and ASXL1 mutations were somatic (n = 4) or not known (n = 2) according to COSMIC database (Supplementary RESULTS Table 1). Presence of KIT D816V and additional mutations in PB cells KIT D816V was detectable in all 19 patients with SM examined. In Detection of KIT D816V in CFU-GM colonies 11 out of 12 patients with (A)SM-AHNMD, a total of 28 different We initially analyzed 330 single-cell-derived CFU-GM colonies additional mutations were detected (ASXL1, CBL, ETV6, EZH2, IDH2, from 19 patients (median number of colonies per patient, n = 15;

Figure 1. Circos diagram: Pairwise co-occurrence of mutations in 11 affected genes in 12 (A)SM-AHNMD patients. The number of the ribbons in each gene corresponds to the number of pairwise co-occurrence of mutations. The width of the ribbon correlates to the relative frequency of concurrent presence of two mutations.

© 2015 Macmillan Publishers Limited Leukemia (2015) 1115 – 1122 Molecular profiling in systemic mastocytosis M Jawhar et al 1118 range 10–30) for the presence or absence of KIT D816V. In CFU-GM Presence of additional mutations in CFU-GM colonies colonies from eight ASM ± AHNMD patients (n = 125; median per Twenty-seven of the 31 (87%) identified additional mutations patient, n = 15; range 10–30), the proportion of KIT D816V+ CFU- were also detected in 224 investigated single-cell-derived CFU-GM GM colonies ranged between 20% and 95% with a median of 60%. colonies. TET2 (n = 9) and/or SRSF2 (n = 4) mutations (seven In four of five patients with SM-AHNMD (n = 130; median per patients) were identified in all investigated CFU-GM colonies − patient, n = 30), CFU-GM colonies were KIT D816V . The fifth (122/122, 100%, median per patient, n = 10, range 10–22). ASXL1 patient (patient #12) had 9 out of 10 (90%) KIT D816V+ CFU-GM mutations were identified in 52 of 63 colonies (83%) in four of five colonies. All CFU-GM colonies from four patients with ISM and two patients. In two of four patients less than 100% of CFU-GM patients with SSM (n = 75; median per patient, n = 10; range 10–15) colonies were ASXL1 positive (Tables 2 and 3). All the SRSF2 − were KIT D816V (Figure 2). mutations (4/4) and 8 of the 9 TET2 mutations tested were somatically acquired. Similarly, all of the ASXL1 mutations (4/4) examined in colonies were somatic (Supplementary Table 1). Additional mutations, which were identified in PB but not in CFU- GM colonies, included ASXL1 (1 of 5 patients, 1/5), CBL (2/2) and U2AF1 (1/1). All mutations were heterozygous.

Subclonal structure of individual patients In eight patients (#1, #3, #6, #7, #8, #12, #24, #27), different subclones were identiﬁed that allowed us to infer the order in which mutations were acquired. For example, in patient #1, most CFU-GM colonies harbored mutations in KIT, TET2, SRSF2 and ASXL1 (KIT/TET2/SRSF2/ASXL1) but minor populations were mutant for TET2/SRSF2/ASXL1 only or TET2/SRSF2 only. Similar patterns were seen in the other patients (Figure 3) and although it is conceivable that KIT D816V could have been acquired early and subsequently lost during clonal evolution, it seems much more likely that KIT D816V was acquired as a late event relative to other mutations in each case. In four SM-AHNMD patients (#13, #14, #23 and #25), all progenitors were KIT D816V− but positive for the known additional mutations. This was unexpected as the KIT D816V allele burden in the PB of these cases was relatively high (median 16%, range 8–27), that is, only marginally lower than the levels seen in ASM or SM-AHNMD (median allele burden 20% and 45%, respectively). ASXL1 (#13) mutations and TET2/SRSF2 (#14) mutations were positive in 100% of investigated CFU-GM colonies, whereas SF3B1 (#23) and JAK2 (#25) mutations were positive in 20% of CFU-GM colonies (Table 3). These results suggest either acquisition of KIT D816V in a more differentiated progenitor cell than CFU-GM or the presence of two independent clones (biclonality), with the KIT D816V+ clone showing no expansion at the CFU-GM level. Figure 2. (a) Relative proportion of KIT D816V+ progenitors in ISM, In addition, we have investigated the molecular proﬁle of single- SM-AHNMD and ASM/MCL-AHNMD. (b) Correlation between the cell-derived CFU-GM obtained from CD34+ cells (n =21) in three relative proportion of KIT D816V+ progenitors and the KIT D816V patients (#13, #14 and #27). CD34+ CFU-GM were entirely positive for allele burden in ASM-AHNMD (r = 0.8; Po0.011). additional mutations (#13: ASXL1; #14: SRSF2, TET2;#27:TET2),

Table 2. Mutation status in seven patients

# Diagnosis % Positive CFU-GM colonies (number of colonies) % Allele burden in PB or BMa

KIT D816V TET2 SRSF2 ASXL1 KIT D816V TET2 SRSF2 ASXL1

14 SM-CMML 0 (20) 100 (15) 100 (15) − 19a 42/45b 20 − 7 ASM-CMML 20 (10) 100 (10) − 80 (10) 19 32/33b − 30 1 ASM-CMML 80 (30) 100 (22) 100 (22) 90 (22) 45a 49 40 40 3 ASM-CMML 95 (20) 100 (10) 100 (10) 100 (10) 45a 46/42b 50 50 12 SM-CMML 90 (10) 100 (10) −−45a 38 −− 24 ASM-MDS/MPNu 40 (20) − 100 (20) − 37a − 40 − 27 ASM-MDS/MPNu 40 (10) 100 (10) −−51a 50 −− Abbreviations: ASM, aggressive SM; BM, bone marrow; CMML, chronic myelomonocytic leukemia; CFU-GM, granulocyte–macrophage colony-forming progenitor cell; MDS/MPNu, myelodyplastic/myeoloproliferative neoplasm unclassiﬁed; PB, peripheral blood; SM, systemic mastocytosis. Mutation status of CFU-GM colonies and allele burden of disparate mutations in PB in seven patients with multi-mutated (A)SM-AHNMD and mutations in KIT, TET2, SRSF2 and/or ASXL1. aAllele burden in BM. bPatients with two mutations in TET2.

Table 3. Mutation status of four patients

# Diagnosis % Positive CFU-GM colonies (number of colonies) % Allele burden in PB

KIT D816V TET2 SRSF2 ASXL1 SF3B1 JAK2 EZH2 KIT D816V TET2 SRSF2 Other mutations

14 SM-CMML 0 (20) 100 (15) 100 (15) −−−−19a 42/45b 20 − 13 SM-MDS 0 (15) −−100 (10) −−−12 −−ASXL1 (18) 23 SM-MDS 0 (15) −−−20 (10) −− 8a −−SF3B1 (41) 25 SM-PV 0 (25) −−−−20 (10) − 27a −− JAK2 (26) Abbreviations: CMML, chronic myelomonocytic leukemia; CFU-GM, granulocyte–macrophage colony-forming progenitor cell; MDS, myelodysplastic syndrome; PB, peripheral blood; PV, polycythemia vera; SM, systemic mastocytosis. Mutation status of CFU-GM colonies, entirely negative for KIT D816V but positive for additional mutations, and allele burden of disparate mutations in PB of four patients with multi-mutated KIT D816V+ SM-AHNMD. aAllele burden in BM. bPatient with two mutations in TET2.

Figure 3. Logical hierarchy of mutation acquisition as determined by genotyping individual CFU-GM colonies. The total number of colonies analyzed, patient number and genotyping results are indicated. For example, for patient #1, 30 colonies were analyzed and three types were detected: (i) 3 colonies were TET2 and SRSF2 mutant but were wild type for ASXL1 and KIT; (ii) 3 colonies were TET2, SRSF2 and ASXL1 mutant but wild type for KIT; (iii) 24 colonies were mutant for all 4 genes. The simplest explanation for this pattern is that the SRSF2 and TET2 mutations arose initially (it is not possible to determine which was ﬁrst since both mutations were present in all colonies tested), a subclone of the double mutant clone then acquired an ASXL1 mutation and then a subclone of the triple mutated clone acquired KIT D816V, as indicated by the blue arrow. In all cases, tested KIT D816V was acquired as a late event relative to other mutations.

© 2015 Macmillan Publishers Limited Leukemia (2015) 1115 – 1122 Molecular profiling in systemic mastocytosis M Jawhar et al 1120 whereas CD34+ CFU-GM cells were entirely negative (#13 and #14) profiles of single-cell-derived CFU-GM colonies in 19 KIT D816V+ or only partially positive (#27, 3/5 colonies) for KIT D816V. SM patients. The study included patients with ISM/SSM/ASM (n = 7; median additional mutations = 0), SM-AHNMD (n =4, Mutation analysis in microdissected cells median additional mutations = 1) and ASM-AHNMD (n =8, median additional mutations = 3). The data highlight several In three of three patients with ASM-AHNMD, microdissected MC new aspects in the pathophysiology and evolution of KIT D816V+ and CD15+ cells derived from BM biopsies were positive for KIT SM and confirm that mutational lesions in additional genes are of D816V as well as for mutations in TET2, which were initially also + prognostic significance. detected in the single-cell-derived CFU-GM colonies. CD3 cells KIT D816V+ CFU-GM could be identified in all eight patients with were positive for KIT D816V and TET2 mutations in one patient fi ASM ± AHNMD but in only 20% (1/5) of patients with SM-AHNMD, (#14) and negative in two other patients (#1, #3). These ndings whereas all CFU-GM colonies derived from CFU-GM of ISM indicate that both mutations must have occurred in a progenitor patients were completely KIT D816V−. However, CFU-GM from before the separation of the myeloid and the MC lineage. In ISM, + − individual ASM ± AHNMD patients were never entirely KIT D816V all CFU-GMs were KIT D816V , and consistent with this finding, all – + − (median 60%, range 0 95). In contrast to KIT D816V, additional CD15 cells were also KIT D816V . Because microdissected MCs fi + mutations were identi ed in CFU-GM from all 12 multi-mutated were KIT D816V , these results indicate that the mutation (A)SM-AHNMD patients and many of these mutations were expanded in or even occurred within the MC lineage. present in 100% of the CFU-GM colonies examined. These mutations included TET2 and SRSF2 mutations in six of six patients Association between mutation status of CFU-GM and disease and four of four patients with (A)SM-AHNMD (Tables 2 and 3). All burden mutations, except KIT D816V, were present concomitantly in The relative proportion of KIT D816V+ progenitors correlated well individual CFU-GM (Figure 3). with the KIT D816V allele burden in PB (r = 0.8; Po0.018; Figure 2). Similar approaches toward the identification of the clonal MC infiltration (in %), KIT D816V allele burden, relative number of architecture have already been applied in related myeloid 21 33 KIT D816V+ progenitors and the number of additional mutations neoplasms, such as AML, CMML and myelofibrosis 15,16,34,35 all showed a suggestive trend and were indicative of an advanced (MF). CFU-GM are relatively mature progenitors and it is 36,37 stage of SM (Table 4). likely that the MC lineage develops from an earlier cell type. In the present study, the major aim was to dissect the molecular aberration patterns and thus clonal evolution in myeloid-restricted Relative frequency of KIT D816V in CMML without known CFU, namely CFU-GM. Therefore, GM-CSF was used as single association of SM growth factor, whereas the use of SCF would have been required If appropriate staining (for example, CD25, tryptase, CD117) of BM to induce the growth of MC lineage progenitors. In this assay, the biopsies and serum tryptase levels are not available, the clinical molecular analysis of CFU-GM colonies allows the reconstitution of phenotype of CMML may be indistinguishable from SM-CMML. the sequential acquisition of mutations in myeloid progenitor cells Based on our results, the molecular phenotype of SM-CMML and in multi-mutated patients. In patients with TET2, SRSF2 and ASXL1 CMML may be distinguished by the presence (multi-mutated SM- mutations, our data suggest that these mutations are acquired CMML) or absence (multi-mutated CMML) of KIT D816V. We at an early stage (similar to CMML38 and other myeloid therefore screened for KIT D816V in 271 morphologically and neoplasms8,24,33,39–41) and that KIT D816V is acquired later. This partially molecularly characterized CMML cases. The relative hypothesis was supported by the observation that microdissected + frequency of KIT D816V cases was 4% (11/271 cases) with a CD15+ myeloid cells did not exhibit KIT D816V in all patients. In median allele burden of 4.4% (range 1.5–37.0) as assessed by cases of multi-mutated SM-AHNMD, in which we could only quantitative real-time PCR. The presence of KIT D816V may have identify additional mutations but not KIT D816V in CFU-GM been missed by conventional Sanger sequencing in 5 of colonies, it is possible that the disease is biclonal, with KIT D816V 11 (46%) patients because of an allele burden of less than 10%. present in the MC lineage while the additional mutations are only present in CFU-GM progenitors. Alternatively, it is perhaps more likely that KIT D816V specifically drives differentiation down the DISCUSSION MC lineage without expansion of mutant CFU-GM, a hypothesis We recently reported on the identification of multiple additional that would clearly explain the strong association between KIT mutations by targeted NGS in approximately 90% of patients with D816V and the phenotype of SM. KIT D816V+ advanced SM including SM-AHNMD, ASM ± AHNMD Similar to TET2 and SRSF2, ASXL1 mutations were detected in a and MCL ± AHNMD (Figure 1).10 The most frequently mutated significant number of patients and in a high proportion of genes were TET2, SRSF2, ASXL1, CBL and EZH2. In order to gain individual progenitors. If these mutations and KIT D816V were more insights into clonal evolution of myeloid progenitors in SM detected concomitantly in individual single-cell-derived CFU-GM with or without AHNMD, we explored the molecular aberration colonies, the overall frequency of additional mutation-positive

Table 4. Clinical characteristics of the analyzed patients

Median (range) ISM,n= 4 SSM,n= 2 SM-AHNMD,n= 5 ASM ± AHNMD, n =8

Serum tryptase (μg/l) 31 (20–54) 220 (200–240) 88 (19–113) 257 (44–743) MC inﬁltration in BM (%) 5 (5–10) 35 (30–40) 20 (10–20) 35 (30–60) KIT D816V allele burden (%) 20 (11–26) 23 (11–35) 19 (8–45) 45 (19–60) KIT D816V+ CFU-GM colonies (%) 0 0 0 (0–10) 60 (20–95) Additional mutations 0 0 1 (1–4) 3 (1–4) Abbreviations: ASM, aggressive SM; BM, bone marrow; CFU-GM, granulocyte–macrophage colony-forming progenitor cell; ISM, indolent SM; MC, mast cell; SM, systemic mastocytosis; SSM, smoldering SM. Serum tryptase, MC inﬁltration in BM, KIT D816V allele burden, KIT D816V+ CFU-GM colonies and number of additional mutations in ISM, SSM, SM-AHNMD and ASM ± AHNMD.

Leukemia (2015) 1115 – 1122 © 2015 Macmillan Publishers Limited Molecular profiling in systemic mastocytosis M Jawhar et al 1121 CFU-GM colonies was always higher than those that were KIT MPN or MDS/MPN11–13,15,16 in early KIT D816V − progenitors D816V+ (Tables 2 and 3), indicating that the KIT D816V mutation may contribute to the respective phenotype, which has was secondary. In contrast, ASXL1 mutations were not identified in been described as AHNMD by morphologists for so many all progenitors in 100% of the TET2/SRSF2-positive patients years.43 The presence of multi-mutated myeloid non-MC-lineage suggesting that they are likely to occur later than TET2 and progenitors of the CFU-GM-type suggests an initial clonal SRSF2. Of interest, SRSF2 and ASXL1 mutations are two of the expansion at an early stage of hematopoietic development four mutations indicating a high-molecular-risk cohort in with a subsequent phenotype modification toward SM because of myelofibrosis35 and CMML42 with a significant negative impact a later acquisition of KIT D816V. In contrast, ISM/SSM seems to be on overall survival. In addition, the number of mutations was not or only rarely affected by mutations at the CFU-GM level, recently established as an important prognostic marker for both which may at least in part explain its excellent prognosis. diseases and we could recently show that the presence of additional aberrations in SM is also associated with a poor CONFLICT OF INTEREST prognosis in a small group of patients.10 Our new data would also S Schnittger and TH have equity ownership of MLL Munich Leukemia Laboratory support the conclusion that the presence of these mutations is fi associated with a higher risk of progression, as most of them were GmbH. The remaining authors declare no competing nancial interests. detected in advanced SM. However, screening of larger cohorts of SM patients is warranted to confirm that not only the presence but ACKNOWLEDGEMENTS also the number and nature of specific mutations may have an This work was supported by the ‘Deutsche José Carreras Leukämie-Stiftung impact on prognosis in SM. Of interest, the allele burden of CBL e.V.’ (grant no. DJCLSR 11/03 and 13/05) and the Austrian Science Fund (FWF) grant mutations in two patients was low and all progenitors were SFB F4704-B20. The authors thank S Uhlig and B Schleider for excellent technical negative, suggesting that the pathogenetic and prognostic assistance. relevance of CBL mutations in SM is low. AHNMD most frequently presents as CMML and less frequently as a MDS/MPNu, MDSu, MPNu, AML or CEL. The recently AUTHOR CONTRIBUTIONS 38 described clonal diversity in CMML and other myeloid MJ, JS, S Schnittger, KS, H-PH, S Schneider, NN, TH, AF and AR performed the 8,24,33,39–41 neoplasms is similar to our findings in ASM-AHNMD. laboratory work for the study. NM, GM, W-KH, PV and AR provided patient The overall qualitative and quantitative presence of identical material and information. H-PH, KS and CW reviewed the bone marrow mutations in TET2 and SRSF2 in progenitors obtained from biopsies. MJ, JS, S Schnittger, KS, PV, W-KH, NCPC, AF and AR wrote the paper. patients with CMML and SM-CMML suggest that the clonal origin of CMML and SM-CMML may be identical, whereas the additional acquisition of KIT D816V acts as phenotype modifier toward SM- REFERENCES CMML. We found that 4% (11/271) of cases diagnosed as CMML 1 Valent P, Horny HP, Escribano L, Longley BJ, Li CY, Schwartz LB et al. Diagnostic were positive for KIT D816V (data not shown), suggesting that criteria and classification of mastocytosis: a consensus proposal. Leuk Res 2001; routine screening for KIT D816V in CMML might be useful. 25: 603–625. In contrast to advanced SM, all ISM patients were negative for 2 Pardanani A. Systemic mastocytosis in adults: 2012 Update on diagnosis, risk fi 87 – additional mutations. In these cases, also the CFU-GM colonies strati cation, and management. Am J Hematol 2012; :401 411. and even the microdissected CD15+ cells were entirely KIT D816V−, 3 Valent P, Akin C, Escribano L, Fodinger M, Hartmann K, Brockow K et al. Standards and standardization in mastocytosis: consensus statements on diagnostics, treat- highlighting that the KIT D816V mutation may be restricted to ment recommendations and response criteria. Eur J Clin Invest 2007; 37:435–453. other (probably later) stages of stem cell development in these 4 Kristensen T, Vestergaard H, Moller MB. Improved detection of the KIT D816V patients and possibly only to the MC lineage. In this regard, it is mutation in patients with systemic mastocytosis using a quantitative and highly noteworthy that the MC lineage is considered to be separate sensitive real-time qPCR assay. J Mol Diagn 2011; 13:180–188. from other myeloid lineages with development from a MC- 5 Garcia-Montero AC, Jara-Acevedo M, Teodosio C, Sanchez ML, Nunez R, Prados A restricted progenitor independent of CFU-granulocyte-erythroyte- et al. KIT mutation in mast cells and other bone marrow hematopoietic cell monocyte-megakaryocyte, CFU-GM and other classical myeloid lineages in systemic mast cell disorders: a prospective study of the Spanish 37 Network on Mastocytosis (REMA) in a series of 113 patients. Blood 2006; 108: progenitor cell subsets. Selective involvement of such MC- – restricted progenitors in ISM has been proposed and may explain 2366 2372. 6 Erben P, Schwaab J, Metzgeroth G, Horny HP, Jawhar M, Sotlar K et al. The KIT at least in part its excellent prognosis with low probability of D816V expressed allele burden for diagnosis and disease monitoring of systemic progression to advanced SM. Two of the reported indolent phase mastocytosis. Ann Hematol 2014; 93:81–88. patients (#16, #17) presented with a BM MC infiltration of 20% and 7 Sotlar K, Colak S, Bache A, Berezowska S, Krokowski M, Bultmann B et al. Variable 50%, respectively, a serum tryptase level of 167 and 166 μg/l, presence of KITD816V in clonal haematological non-mast cell lineage diseases respectively, and a KIT D816V allele burden of 11% and 35%, associated with systemic mastocytosis (SM-AHNMD). J Pathol 2010; 220:586–595. respectively, as markers for high disease burden. However, there 8 Soucie E, Hanssens K, Mercher T, Georgin-Lavialle S, Damaj G, Livideanu C et al. In were no C-findings and the clinical course was stable for the aggressive forms of mastocytosis, TET2 loss cooperates with c-KITD816V to 120 – previous 4 and 16 years, respectively. A third patient with ASM transform mast cells. Blood 2012; : 4846 4849. presented with 60% MC infiltration, a serum tryptase of 400 μg/l 9 Traina F, Visconte V, Jankowska AM, Makishima H, O'Keefe CL, Elson P et al. Single nucleotide polymorphism array lesions, TET2, DNMT3A, ASXL1 and CBL mutations and a KIT D816V allele burden of 50%. In these three patients, all 7 − are present in systemic mastocytosis. PLoS One 2012; : e43090. CFU-GM were KIT D816V and no additional mutations were found 10 Schwaab J, Schnittger S, Sotlar K, Walz C, Fabarius A, Pfirrmann M et al. Com- as possible explanation for the indolent clinical course even prehensive mutational profiling in advanced systemic mastocytosis. Blood 2013; in ASM. 122: 2460–2466. In general, the relative proportion of KIT D816V+ progenitors 11 Kohlmann A, Grossmann V, Klein HU, Schindela S, Weiss T, Kazak B et al. Next- correlated well with established parameters for quantification of generation sequencing technology reveals a characteristic pattern of molecular disease burden, for example, BM MC infiltration, serum tryptase mutations in 72.8% of chronic myelomonocytic leukemia by detecting frequent 28 – levels and KIT D816V allele burden in individual patients. alterations in TET2, CBL, RAS, and RUNX1. J Clin Oncol 2010; : 3858 3865. However, the presence of KIT D816V in non-MC lineages may 12 Haferlach T, Nagata Y, Grossmann V, Okuno Y, Bacher U, Nagae G et al. Landscape of genetic lesions in 944 patients with myelodysplastic syndromes. Leukemia cause a high KIT D816V allele burden and even C-findings, 28 – fi 2014; :241 247. whereas the extent of MC in ltration and serum tryptase levels is 13 Meggendorfer M, Roller A, Haferlach T, Eder C, Dicker F, Grossmann V et al. SRSF2 6 low. Furthermore, the presence of additional mutations, known mutations in 275 cases with chronic myelomonocytic leukemia (CMML). Blood to be present in a significant proportion of patients with MDS, 2012; 120: 3080–3088.

© 2015 Macmillan Publishers Limited Leukemia (2015) 1115 – 1122 Molecular profiling in systemic mastocytosis M Jawhar et al 1122 14 Lundberg P, Karow A, Nienhold R, Looser R, Hao-Shen H, Nissen I et al. 29 Simons A, Shaffer LG, Hastings RJ. Cytogenetic nomenclature: changes in the ISCN Clonal evolution and clinical correlates of somatic mutations in myeloproliferative 2013 compared to the 2009 Edition. Cytogenet Genome Res 2013; 141:1–6. neoplasms. Blood 2014; 123: 2220–2228. 30 Schoch C, Schnittger S, Bursch S, Gerstner D, Hochhaus A, Berger U et al. 15 Tefferi A. Novel mutations and their functional and clinical relevance in Comparison of chromosome banding analysis, interphase- and hypermetaphase- myeloproliferative neoplasms: JAK2, MPL, TET2, ASXL1, CBL, IDH and IKZF1. FISH, qualitative and quantitative PCR for diagnosis and for follow-up in chronic Leukemia 2010; 24: 1128–1138. myeloid leukemia: a study on 350 cases. Leukemia 2002; 16:53–59. 16 Vainchenker W, Delhommeau F, Constantinescu SN, Bernard OA. New mutations 31 Sotlar K. c-Kit mutational analysis in paraffin material. Methods Mol Biol 2013; 999: and pathogenesis of myeloproliferative neoplasms. Blood 2011; 118: 1723–1735. 59–78. 17 Gotlib J, Berube C, Growney JD, Chen CC, George TI, Williams C et al. Activity of 32 Sotlar K, Marafioti T, Griesser H, Theil J, Aepinus C, Jaussi R et al. Detection of c-kit the tyrosine kinase inhibitor PKC412 in a patient with mast cell leukemia with the mutation Asp 816 to Val in microdissected bone marrow infiltrates in a case of D816V KIT mutation. Blood 2005; 106: 2865–2870. systemic mastocytosis associated with chronic myelomonocytic leukaemia. J Clin 18 Bibi S, Langenfeld F, Jeanningros S, Brenet F, Soucie E, Hermine O et al. Molecular Pathol Mol Pathol 2000; 53:188–193. defects in mastocytosis: KIT and beyond KIT. Immunol Allergy Clin North Am 2014; 33 Itzykson R, Kosmider O, Renneville A, Morabito M, Preudhomme C, Berthon C et al. 34: 239–262. Clonal architecture of chronic myelomonocytic leukemias. Blood 2013; 121: 19 Hou Y, Song L, Zhu P, Zhang B, Tao Y, Xu X et al. Single-cell exome sequencing 2186–2198. and monoclonal evolution of a JAK2-negative myeloproliferative neoplasm. Cell 34 Kralovics R. Genetic complexity of myeloproliferative neoplasms. Leukemia 2008; 2012; 148: 873–885. 22: 1841–1848. 20 Bendall SC, Nolan GP. From single cells to deep phenotypes in cancer. Nat 35 Vannucchi AM, Lasho TL, Guglielmelli P, Biamonte F, Pardanani A, Pereira A et al. Biotechnol 2012; 30: 639–647. Mutations and prognosis in primary myelofibrosis. Leukemia 2013; 27: 21 Jan M, Snyder TM, Corces-Zimmerman MR, Vyas P, Weissman IL, Quake SR et al. 1861–1869. Clonal evolution of preleukemic hematopoietic stem cells precedes human acute 36 Agis H, Fureder W, Bankl HC, Kundi M, Sperr WR, Willheim M et al. Comparative myeloid leukemia. Sci Transl Med 2012; 4:118–149. immunophenotypic analysis of human mast cells, blood basophils and mono- 22 Landau DA, Carter SL, Stojanov P, McKenna A, Stevenson K, Lawrence MS et al. cytes. Immunology 1996; 87:535–543. Evolution and impact of subclonal mutations in chronic lymphocytic leukemia. 37 Agis H, Willheim M, Sperr WR, Wilfing A, Kromer E, Kabrna E et al. Monocytes do Cell 2013; 152:714–726. not make mast cells when cultured in the presence of SCF. Characterization of the 23 Melchor L, Brioli A, Wardell CP, Murison A, Potter NE, Kaiser MF et al. Single-cell circulating mast cell progenitor as a c-kit+, CD34+, Ly-, CD14-, CD17-, colony- genetic analysis reveals the composition of initiating clones and phylogenetic forming cell. J Immunol 1993; 151: 4221–4227. patterns of branching and parallel evolution in myeloma. Leukemia 2014; 28: 38 Grossmann V, Kohlmann A, Eder C, Haferlach C, Kern W, Cross NC et al. Molecular 1705–1715. profiling of chronic myelomonocytic leukemia reveals diverse mutations in 480% 24 Papaemmanuil E, Gerstung M, Malcovati L, Tauro S, Gundem G, Van Loo P et al. of patients with TET2 and EZH2 being of high prognostic relevance. Leukemia Clinical and biological implications of driver mutations in myelodysplastic syn- 2011; 25: 877–879. dromes. Blood 2013; 122: 3616–3627. 39 Smith AE, Mohamedali AM, Kulasekararaj A, Lim Z, Gaken J, Lea NC et al. Next- 25 Valent P HH, Li CY, Longley JB, Metcalfe DD, Parwaresch RM, Bennett JM. generation sequencing of the TET2 gene in 355 MDS and CMML patients reveals Mastocytosis. In: Elaine S, Jaffe NLH, Stein Harald, James WV (eds). Pathology low-abundance mutant clones with early origins, but indicates no definite and Genetics: Tumours of Haematopoietic and Lymphoid Tissues: World Health prognostic value. Blood 2010; 116:3923–3932. Organization (WHO) Classification of Tumours vol. 1. IARC Press: Lyon, France, 2001; 40 Mian SA, Smith AE, Kulasekararaj AG, Kizilors A, Mohamedali AM, Lea NC et al. pp 291–302. Spliceosome mutations exhibit specific associations with epigenetic modifiers 26 Horny HP, Valent P. Diagnosis of mastocytosis: general histopathological aspects, and proto-oncogenes mutated in myelodysplastic syndrome. Haematologica morphological criteria, and immunohistochemical findings. Leuk Res 2001; 25: 2013; 98: 1058–1066. 543–551. 41 Tefferi A, Lim KH, Levine R. Mutation in TET2 in myeloid cancers. N Engl J Med 27 Horny HP AC, Metcalfe DD, Swerdlow SH, Campo E, Harris NL et al. World Health 2009; 361:1117–1118. Organization (WHO) Classification of Tumours. Mastocytosis (Mast cell disease). 42 Itzykson R, Kosmider O, Renneville A, Gelsi-Boyer V, Meggendorfer M, Morabito M Pathology & Genetics. Tumours of Haematopoietic and Lymphoid Tissues vol. 2. et al. Prognostic score including gene mutations in chronic myelomonocytic Lyon, France: IARC Press, 2008; pp 54–63. leukemia. J Clin Oncol 2013; 31: 2428–2436. 28 Tefferi A, Thiele J, Vardiman JW. The 2008 World Health Organization classification 43 Horny HP, Sotlar K, Sperr WR, Valent P. Systemic mastocytosis with associated system for myeloproliferative neoplasms: order out of chaos. Cancer 2009; 115: clonal haematological non-mast cell lineage diseases: a histopathological 3842–3847. challenge. J Clin Pathol 2004; 57: 604–608.

Supplementary Information accompanies this paper on the Leukemia website (http://www.nature.com/leu)