Detection Of Low-level Mutations Research Articles

Ultra-Accurate Assessment of Pretreatment ABL1 Kinase Domain (KD) Mutations in Patients (pts) with Newly Diagnosed Philadelphia Chromosome-Positive Acute Lymphoblastic Leukemia (Ph+ ALL) Using Duplex Sequencing (DS)

Ultra-Accurate Assessment of Pretreatment ABL1 Kinase Domain (KD) Mutations in Patients (pts) with Newly Diagnosed Philadelphia Chromosome-Positive Acute Lymphoblastic Leukemia (Ph+ ALL) Using Duplex Sequencing (DS)

Clinical Impact of Clonal Hematopoiesis after Autologous Stem Cell Transplantation for Lymphoma: A National Population-Based Cohort Study

▪Background:Somatic driver mutations in hematopoietic cells may lead to clonal hematopoiesis of indeterminate potential (CHIP). In patients with lymphoma CHIP has been associated with increased risk of therapy-related myeloid neoplasms (tMN) and inferior survival after autologous stem cell transplantation as demonstrated in a large single center study and in a case-control study (Gibson CJ et al., JCO 2017 and Berger G et al., Blood 2018). Here, we investigated the clinical impact of clonal hematopoiesis in a nation-wide population-based cohort of Danish lymphoma patients undergoing autologous transplant with prospective data from four national patient registries.Methods:Patients with lymphoma who had undergone leukapheresis at all danish transplant centers from 2000 to 2012 were identified. DNA and RNA was extracted from mobilized peripheral blood products. Targeted sequencing of all samples was performed using an Illumina TruSeq Custom Amplicon panel (Illumina, San Diego, CA, USA) designed to cover >95% of mutations associated with CHIP (ASXL1, ASXL2, BCOR, BRCC3, CBL, CREBBP, DNMT3A, ETV6, GNB1, IDH1, IDH2, JAK2, KRAS, NRAS, PPM1D, RAD21, SF3B1, SRSF2, TET2, TP53). To allow detection of low-level mutations and secure variant calling, unique molecular identifiers (UMI's) were used. Filtering of variants was done by stringent criteria consistent with earlier studies. Assessment of mutations was performed blinded to the patients' clinical data.Prospective clinical patient data was obtained for all patients from four national registries, including the Danish Lymphoma Registry (diagnosis, involvement, lymphoma treatment, relapse and death), the Danish National Patient Registry (hospital admission diagnoses and treatments), the Danish Cancer Registry (primary and secondary cancer diagnoses) and the Danish Pathology Database (histopathological examinations and diagnoses), respectively.Results:Samples from 574 patients were included. The median age was 55.5 years (IQR: 45.3 - 62.2) and the median follow-up time for survivors was 9.2 years (IQR: 7.1 - 11.2). The lymphoma subtypes were typical of patients selected for autologous transplantation; diffuse large B-cell lymphoma (191 pts), follicular lymphoma (102 pts), mantle cell lymphoma (88 pts), Hodgkin's lymphoma (80 pts), peripheral T-cell lymphoma (77 pts) and other histologies (36 pts). Of the 574 patients analyzed, 191 (33.3%) of the patients had somatic mutations meeting CHIP criteria (total mutations called=210). The most commonly mutated genes were DNMT3A (n=59, 28%), TET2 (n=48, 23%), PPM1D (n=34, 16%), ASXL1 (n=21, 10%) and TP53 (n=18, 8%). As expected CHIP mutations were more frequent in patients above 60 years (p=0.002). Prevalence of CHIP was associated with an inferior overall survival (p=0.004) and event-free survival (p=0.03). It was also associated with increased risk of biopsy-confirmed tMN (p=0.03) and higher probability of receiving blood transfusions after autologous transplant (p=0.027).Especially patients with mutations in DNA damage response genes PPM1D and TP53 (found in 48 pts, 8.3%) had a significantly increased risk of adverse outcomes. Both overall survival and event-free survival were significantly poorer with the presence of DNA damage pathway mutations (p<0.0001 for both, Figure 1A), as well as risk of tMN (p=0.01). In addition, PPM1D/TP53 mutations were associated with increased rates of any secondary cancer (p=0.004), including non-hematological cancer, and hospital admissions with severe infections (p=0.01, Figure 1B).The impact of low-level mutations and statistical modelling of interactions between parallel outcomes will be presented at the meeting.Conclusion:To our knowledge this is the first population-based study of clonal hematopoiesis in patients with lymphoma. We find that CHIP and particularly mutations in DNA damage response genes (PPM1D/TP53) are associated with increased mortality, which confirms findings from single center studies. These data support the evaluation of CHIP for risk assessment in lymphoma patients before high-dose chemotherapy. Our study also identifies increased rates of several clinically relevant adverse outcomes (severe infections, blood transfusions and secondary cancers) in lymphoma patients with clonal hematopoiesis. DisclosuresGrønbæk:Janssen Pharma: Membership on an entity's Board of Directors or advisory committees; Celgene: Membership on an entity's Board of Directors or advisory committees; Otsuka Pharma: Membership on an entity's Board of Directors or advisory committees.

Abstract 941: Method for sensitive detection of tumor fingerprints in plasma

Abstract Presence of excess unaltered, wild-type DNA providing no information of biologic or clinical value may often mask rare alterations containing diagnostic or therapeutic clues for cancer. There is a growing demand for removing unaltered DNA over large pools-of-sequences, especially for applications in liquid biopsies with circulating DNA (cfDNA). We recently developed nuclease-assisted minor-allele enrichment with probe-overlap (NaME-PrO), a single-step approach that removes WT-DNA prior to DNA amplification, following which current genomic analysis processes remain unchanged. NaME-PrO employs a double-strand DNA-specific nuclease (DSN) and overlapping oligonucleotide-probes interrogating multiple DNA targets. Following genomic-DNA denaturation, the temperature is lowered and the probes form transient double-stranded regions with their targets, and DSN is added to provide nuclease digestion to the selected sites. Mutations create mismatches that inhibit DSN digestion; thus, subsequent amplification yields DNA with alterations enhanced at multiple targets. In this manner, WT DNA at hundreds or thousands of DNA regions can be digested simultaneously. We also show that addition of organic solvents like DMSO in the NaME-PrO reaction enables DNA denaturation at lower temperatures that do not inactivate the enzyme (DSN). This enables DSN addition from the beginning of the reaction, enabling homogeneous, closed-tube application of the approach and avoids contamination-prone sample manipulations due to open-tube addition of DSN enzyme after denaturation. Using NaME-PrO ± DMSO we demonstrate several-hundred-fold multiplexed mutation enrichment in diverse human samples on multiple clinically relevant targets in 50-plex reactions. Application with targeted resequencing of tumor-circulating DNA demonstrated detection of mutations at the 0.01-0.1% levels with few sequence reads. In ongoing studies aiming to detect minimal residual disease using cfDNA in late-stage breast cancer, we used tumor exome sequencing to identify multiple clonal mutations for individual patients. We then performed multiplexed NaME-PrO to trace the "tumor fingerprint" in plasma. In ~50-plex reactions we showed that the majority (&amp;gt;60%) of the tumor-specific mutations were also detected in plasma. To evaluate the ultimate sensitivity of the method to identify such "tumor fingerprints," we applied NaME-PrO in serial dilutions of cfDNA from breast cancer patients into cfDNA from cancer-free, normal patients. We demonstrated that dilutions down to 1,000-10,000-fold, corresponding to mutations at &amp;lt;0.01% allelic frequency, lead to identification of at least one tumor-specific mutation in plasma. NaME-PrO can be combined with platforms that utilize molecular barcodes to provide "maximum efficiency sequencing," enabling sensitive detection of low-level mutations without requiring "deep sequencing" and excessive sequence reads. Citation Format: Fangyan Yu, Mariana Fitarelli-Kiehl, Ka Wai Leong, Viktor Adalsteinsson, Heather Parsons, Mike G. Makrigiorgos. Method for sensitive detection of tumor fingerprints in plasma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 941.

Molecular monitoring in chronic myeloid leukemia-how low can you go?

Molecular monitoring of BCR-ABL1 transcripts for patients with chronic myeloid leukemia (CML) is now used to assess response to tyrosine kinase inhibitors (TKIs), including treatment failure that mandates a change of therapy. Therefore, many centers have adopted the molecular technique for measuring BCR-ABL1 and rely on conversion of values to the international reporting scale for appropriate clinical interpretation. However, the technique has a degree of inherent variability despite standardized procedures, which means care should be taken by the clinician when assessing response based on BCR-ABL1 cutoff limits. The last few years have witnessed the emergence of a new molecular response target, which is the achievement and maintenance of a deep molecular response. The ability to achieve treatment-free remission for some patients has shifted the relevant boundary for molecular response. However, the definitive safe BCR-ABL1 transcript level and length of the maintenance phase after which treatment cessation can be attempted has not yet been determined. For patients with TKI resistance, BCR-ABL1 kinase domain mutation analysis remains an essential assessment to guide therapy. Furthermore, low-level mutation detection is clinically relevant for response prediction to subsequent TKI therapy for some patients. Multiple low-level mutations may be a biomarker of a clonally diverse disease with the propensity for resistance evolution. Overall, molecular monitoring, including low-level monitoring is a fundamental component of management for patients with CML.

Clonal Diversity of Recurrently Mutated Genes in Myelodysplastic Syndromes

Myelodysplastic syndromes (MDS) are a group of chronic and often progressive myeloid neoplasms associated with remarkable heterogeneity in the degree of cytopenia, morphology and clinical course. Various somatic mutations are involved in the pathogenesis of MDS. The recent identification of acquired mutations in the key components of the spliceosome machinery strongly implicates abnormalities of mRNA splicing in the pathogenesis of myelodysplastic syndromes.In this study, we performed Targeted Massively Parallel Sequencing on a custom 29 myeloid gene panel (all coding regions) of myeloid genes using an AmpliSeq approach (Life Tech) on 148 MDS patient samples. Barcodes were added using NEB kit and libraries were prepared for Illumina HiSeq2500 sequencing.Results: The median age of patients was 72 (19-97) years and 65% patients were male. Majority of patients were RA-RS, RCMD, RCMD-RS, RAEB1 and RAEB2. Conventional bone marrow cytogenetic was normal in 80/148 (55%) of cases while sequencing of 29-gene panel revealed at least one mutation in 131/148 (88%) patients. Importantly a combination of cytogenetic and mutation screening detected genetic abnormalities in 141/148 (95%) of MDS patients, thus highlighting the importance of combining targeted mutation screening along with conventional karyotyping in MDS patients.The majority of patients had more ≥2 mutations 112/148 (75%) suggesting clonal heterogeneity. As shown in Figure (1A), the most frequently mutated genes were ASXL1, SF3B1, NOTCH1, TET2, SRSF2, RUNX1 and DNMT3A. Mutations in the spliceoosome complex genes were detected in 66/148 (44%) of patients with SF3B1 (30/148; 20%) more frequent than SRSF2 (24/148; 16%) and U2AF1 (12/148; 8%). These mutations were mutually exclusive except for one patient.SF3B1 mutations were detected in 95% (19/20) of RA-RS and RCMD-RS cases, while >50% of patients with SRSF2 mutations had RAEB1 or RAEB2. Mutations in the spliceosome complex co-occur with mutations in the epigenetic modifiers such as DNMT3A and TET2 mutations which is consistent with the previous report.We identified 47 patients with 2 (n=37) or three (n=10) TET2 mutations. Of these, 16 patients had mutations at significantly different allele frequencies suggesting that independent clones had arisen, raising the possibility that in these samples a pre-leukemic genetic or epigenetic cellular environment exists that selects for TET2 mutations. We also saw a similar trend, although to a lesser extent, for NOTCH1, RUNX1 and NRAS mutations.Conclusions: Combination of targeted sequencing of most commonly mutated genes and conventional cytogenetic can detect genetic abnormalities in 95% of MDS cases. This study shows that >75% of MDS cases have more than one mutations demonstrating clonal heterogeneity. SF3B1 mutations are detected in 95% of RA-RS and RCMD-RS cases and are frequently associated with mutations in epigenetic modifiers. A high level of mutation detection has been possible because detection of low level mutations due to deep sequencing. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.

Abstract 1892: Development of a clinical targeted next generation sequencing test for challenging formalin-fixed paraffin-embedded (FFPE) cancer samples

Abstract Introduction: There is an urgent need for deep targeted sequencing for detection of varieties of somatic aberrations, including single nucleotide variants (SNV), insertions/deletions (indels), loss of heterozygosity (LOH), and copy number variants (CNV) in heterogeneous FFPE cancer patient specimens. We have developed a Next Generation Sequencing (NGS)-based assay for deep sequencing of a number of selected oncogenes and tumor suppressor genes for testing of clinical breast cancer samples. The panel includes those genes, which are commonly mutated in cancers (e.g. breast and ovarian cancer), based on The Cancer Genome Atlas studies and are potentially indicative of responses to experimental compounds undergoing clinical trials. Importantly, the assay can be used in FFPE samples which pose challenges because of high variability in quality and often suboptimal quantities. Methods: To minimize FFPE sample input requirement, we optimized multiple assay steps. DNA was extracted from macrodissected FFPE tissue, and randomly fragmented. The DNA library was constructed and subject to target capture and enrichment before being sequenced on the Illumina MiSeq® systems. To analytically qualify the assays, we tested well-characterized cell lines and reference HapMap cell lines. Furthermore, to confirm the performance of the NGS results in clinical FFPE samples, we compared the results generated using orthogonal methodologies, e.g. SNVs using a custom Ion AmpliSeq™ panel (Life Technologies) on Ion Personal Genome Machine® (PGM), or CNVs using real-time PCR assays or custom Nanostring nCounter® CNV assays (NanoString Technologies). To establish the lower limit of detection, we used mixed DNA samples from FFPE cell lines and from HapMap DNA and further tested Horizon FFPE reference DNA containing mutations at defined allele frequencies. Results: For FFPE DNA extraction, we selected Cobas® FFPE kit (Roche), and Maxwell® CSC DNA FFPE kit (Promega) for higher yield and better purity. For library construction, we chose a KAPA streamlined library preparation kit. For target capture, we used custom xGEN® DNA probes (Integrated DNA Technologies) for improved flexibility in assay optimization, reduced reagent cost, and minimizing batch-to-batch variability. We were able to detect SNVs for &amp;gt;99% of exons selected down to a 5% sensitivity level using about 100 ng FFPE DNA input. Indels, CNV and LOH could also be detected reliably. Conclusions: We developed a targeted NGS assay that could be used in clinical trials and meets the dual challenge of limited DNA amounts and/or variable quality often associated with FFPE clinical specimens and the detection of low level mutations. Considering the recent recognition of Illumina NGS platform by FDA, this approach can represent a path forward to assist in cancer patient assessment, clinical trial recruitment and care. Citation Format: Wenge Shi, Christine Chin, Tingdong Tang, Loretta Hipolito, Preethi Srinivasan, Derek Chiang, David Peng, Emmanuelle Di Tomaso, Shabnam Tangri, Jelveh Lameh, Reinhold Pollner. Development of a clinical targeted next generation sequencing test for challenging formalin-fixed paraffin-embedded (FFPE) cancer samples. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 1892. doi:10.1158/1538-7445.AM2014-1892

Next-Generation Sequencing Of The BCR-ABL1 Kinase Domain May Be Beneficial In Decision Making Among Chronic Myeloid Leukemia Patients With Tyrosine Kinase Inhibitor Resistance

Background and AimBCR-ABL1 mutation testing is recommended for chronic myeloid leukemia (CML) patients who have suboptimal response and/or treatment failure with tyrosine kinase inhibitor (TKI) therapy. BCR-ABL1 mutations in the kinase domain (KD) of ABL1 account for at least 40-50% of all TKI resistant cases. Thus, detection of low-level mutations after development of resistance may offer critical information to guide subsequent therapy selection. The current gold standard for BCR-ABL1 mutation detection is Sanger sequencing (SS), which has an analytical sensitivity of approximately 10-20%. In this study, our aim was to detect low level BCR-ABL1 variants in follow up samples of CML patients with TKI resistance using next-generation sequencing (NGS) approach. MethodsEight patients with CML who were resistant to imatinib had been routinely sequenced with SS for BCR-ABL1 KD mutations between December 2009 and December 2012. We then retrospectively analyzed these samples with NGS. RT and long range PCR was performed to amplify BCR-ABL1 fusion transcripts and the PCR products sequenced bidirectional after library preparation. We performed a fusion transcript based BCR-ABL1 mutation assay using Roche 454 amplicon deep-sequencing technology that is suited for detecting low level variants in pooled amplicon samples. Sequencing data was analyzed with GS Amplicon Variant Analyzer (AVA) software, and the variant frequency cut-off was adjusted to 1%. ResultsClinical features, sequencing results, and the outcomes of the patients were summarized in Table 1. Four patients were male, and the median age was 37 years (range, 20-60 years). The patients were all in chronic phase at the time of the diagnosis. After imatinib resistance, 4 patients had received dasatinib (DAS), and 2 were given nilotinib (NIL) as second line TKI treatment. The remaining two patients had both received DAS and NIL (Table 1). In a set of 20 clinical samples, at different time points, NGS not only identified all the mutations detected by SS, but additionally identified low level variants present between 1 – 28.12 %. T315I and E255K/V were the most common mutations, which were detected in four patients, both by SS and NGS at the same time points (Table 1). Two patients (patient #1 and #4) had T315I, and they both progressed to blastic phase and died. E255K was detected in patients #2 and #3, and patient #2 had achieved and maintained complete cytogenetic and major molecular responses with 100 mg daily DAS, whereas patient #3 had received both NIL and DAS, but she was deceased due to myeloid blastic crisis. Among 4 patients (patients #5, #6, #7, and #8), mutation analysis was performed at eleven different time points, and these patients were wild-type with SS. We also did not detect any clinically significant mutations in these patients by NGS. Most probably mechanisms other than KD mutations were responsible for the TKI resistance among these four patients. ConclusionsPolyclonal mutations in BCR-ABL1 KD are commonly identified in TKI resistant patients. Thus, detection of low-level mutations after development of resistance offers critical information to guide subsequent therapy selection. An inappropriate kinase inhibitor selection could highly increase the risk of treatment failure with clonal expansion of the resistant mutant. In our imatinib resistant cohort, we detected low level variants accompany to known mutations which may constitute background genetic variations. Although we had expected to detect mutations earlier by NGS (i.e. before these mutations can be detected by SS), we did not observe such finding in our patients. The patients' samples may not show a stable mutation spectrum between time points. Hence, it is not always possible to spot a mutation before patients show resistance to therapy. Regular NGS analysis might detect these mutations in earlier phases, which might help clinicians to choose the most suitable individual treatment modality for the patients. AcknowledgmentThe authors would like to thank the Interlaboratory Robustness of Next-generation sequencing (IRON) Phase II study group members, especially to Simona Soverini and Alexander Kohlmann who designed BCR-ABL primers and plates. We also would like to thank the Research Fund of the Istanbul University (Project no. 24244) and Turkish Society of Hematology for supporting the study.▪ Disclosures:Sayitoglu:Roche Diagnostics: Research Support Other.

Abstract 2113: A wild-type-blocking reference sequence enhances COLD-PCR and enables fast amplification and high enrichment of all types of low-prevalence unknown mutations

Abstract Background: Molecular profiling of somatic mutations in cancer often requires the identification of low-prevalence DNA mutations in an excess of wild-type (WT) DNA; however, the method selectivity and sensitivity are often limiting factors. COLD-PCR (CO-amplification at Lower Denaturation Temperature) resolves several limitations in low-level mutation detection by using critical denaturation temperatures to enrich mutant-containing amplicons during PCR. However, certain mutation types are not enriched by some COLD-PCR formats. We report a novel enhancement in COLD-PCR that enables an Improved and Complete Enrichment (ice-COLD-PCR) for all mutation types. Methods: A Reference Sequence (RS) that preferably forms double-stranded structures with WT sequences, but not with mutant sequences, was added to COLD-PCR. A RS was designed to bind to one WT-sequence strand, avoid primer binding and prevent polymerase extension. To validate the use of a RS in COLD-PCR, we evaluated segments of TP53 exon 8. Serial dilutions of mutant cell-line DNA, or human lung tumor DNA, in WT DNA were analyzed. Mutations that increase, decrease, or retain the amplicon melting temperature were tested. Following conventional-PCR, COLD-PCR (full or fast COLD-PCR formats), and ice-COLD-PCR methods amplicons were sequenced and the degree of mutation enrichment was compared. Several clinical lung adenocarcinoma samples with known low-level mutations were also tested with ice-COLD-PCR. Results: ice-COLD-PCR yielded ∼13-fold enrichment for Tm-increasing and Tm-equivalent mutations, and ∼15-fold enrichment for Tm-reducing mutations. In contrast, Full-COLD-PCR demonstrated ∼5-fold enrichment for all mutations. Further, fast-COLD-PCR, which can only enrich Tm-reducing mutations, exhibited ∼17-fold enrichment for these types of mutations, while the Tm-increasing and Tm-equivalent mutations remained undetectable. Regardless of mutant type and position, after ice-COLD-PCR amplification, all mutation types are strongly enriched and can be reliably sequenced down to a level of 1-3%. Ice-COLD-PCR duration is ∼1hour, compared to several hours for full-COLD-PCR. Conclusions: The inclusion of appropriately designed RS within COLD-PCR selectively inhibits WT amplification throughout PCR, while preferentially enriching mutants and reducing time-intensive hybridization times. Ice-COLD-PCR combines high sensitivity, speed, and low-cost, and facilitates direct sequencing for all types of unknown low-prevalence mutations in clinical cancer samples. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 2113.

Patients with Imatinib Resistance Harbour Low Level Mutations of the BCR-ABL Kinase Domain Predominantly in the CD34+ Cells.

Abstract 3267Poster Board III-1 Objectives:BCR-ABL kinase domain (KD) mutations can be detected at a low level prior to the start of imatinib (IM) in patients with advanced phase chronic myeloid leukemia (CML) and the presence of such mutations in CD34+ cells of CML patients in complete cytogenetic response is thought to underlie disease persistence on IM. Since new tyrosine kinase inhibitor (TKI) specific mutations have been shown to arise on nilotinib or dasatinib treatment, we have asked in this analysis whether patients with resistance to TKI may harbour additional, low-level BCR-ABL KD compound mutations in the early progenitor cells. Patients and Methods:Using a MACS® indirect CD34 Micro Bit Kit we selected a minimum of 1×104 CD34+ and CD34- cells from 16 patients at the time of TKI resistance. The median purity of the CD34+ and CD34- cell fractions was 93% (range 87 to 98%) and 95% (range 58 to 100%). Ten of these 16 patients (56%) had BCR-ABL KD mutations detectable by direct sequencing (DS M244Vx2, G250E, V299Lx2, F317I/Lx4, F359Vx2) defined as high level mutations. Complementary DNA prepared from total white cells, CD34+ and CD34- cells was used for the amplification of BCR-ABL and the sensitive detection of 8 specific BCR-ABL KD mutations (F317L, F359V, T315I, E255K, E255V, Y253H, G250E, Q252H) by quantitative Ligation-PCR, which yields a reproducible sensitivity of 0.5% BCR-ABLmutant/BCR-ABLtotal, enabling the quantitative detection of low level mutations which are not detectable by direct sequencing. Results:Of the 10 patients carrying high level mutations, 9 (90%) were represented within our specific ligation PCR-panel. In these cases, the total mutated BCR-ABL level was comparable between the total white cells (median 99%, range 14 to 100%), CD34+ (median 100, range 18.02 to 100%) and CD34- cells (median 100, range 11.73 to 100% BCR-ABLmutant/BCR-ABLtotal). However, ligation-PCR detected further low level mutations within the 16 patient panel, which where present at a higher frequency in the CD34+ cells (n=8; Y252H, E255K/V, T315I, F317Lx2, F359Vx2) than in the CD34- (n=2; Y252H, T315I) or the total white cells (n=3; Y252H, T315I, F359V). Furthermore, the proportion of mutated BCR-ABL within the patients with low level mutations was higher in the CD34+ cells (median 4.57%, range 0.64 to 7.82%) then in the CD34- (median 1.18%, range 0.75 to 1.61%) or total white cells (median 1.24%, range 0.7–6.85%). Within the 10 patients with high level mutations, the low level mutations were exclusively detected in CD34+ cells (p=0.014). Conclusions:Whereas high level mutations are present at the same level in total white cells, CD34+ and CD34- cells, we confirm our hypothesis that low level mutations are predominantly detectable in the early progenitor fraction. This is consistent with a spontaneous background of potentially resistant mutations in the stem/progenitor population which have the potential to develop a resistant leukemic phenotype on ineffective TKI treatment. Disclosures:Al-Ali:BMS: Honoraria, Membership on an entity's Board of Directors or advisory committees. Niederwieser:BMS: Honoraria, Membership on an entity's Board of Directors or advisory committees; Novartis: Honoraria, Research Funding. Lange:BMS: Honoraria; Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding.