J Clin Oncol by American Society of Clinical Oncology INTRODUCTION

Transcription

1 JOURNAL OF CLINICAL ONCOLOGY O R I G I N A L R E P O R T Sensitive Detection of BCR-ABL1 in Patients With Chronic Myeloid Leukemia After Imatinib Resistance Is Predictive of Outcome During Subsequent Therapy Wendy T. Parker, Rebecca M. Lawrence, Musei Ho, Darryl L. Irwin, Hamish S. Scott, Timothy P. Hughes, and Susan Branford Wendy T. Parker, Rebecca M. Lawrence, Musei Ho, Hamish S. Scott, Timothy P. Hughes, and Susan Branford, Centre for Cancer Biology, SA Pathology; Hamish S. Scott, Timothy P. Hughes, and Susan Branford, University of Adelaide, Adelaide, South Australia; and Darryl L. Irwin, Sequenom, Herston, Queensland, Australia. Submitted February 2, 2011; accepted August 4, 2011; published online ahead of print at on October 11, Supported by Grant No from the National Health and Medical Research Council. The contents of this manuscript are solely the responsibility of the authors and do not reflect the views of the National Health and Medical Research Council. Authors disclosures of potential conflicts of interest and author contributions are found at the end of this article. Corresponding author: Wendy T. Parker, PhD, SA Pathology, IMVS, Frome Rd, PO Box 14 Rundle Mall, Adelaide, SA, 5000, Australia; wendy.parker@health.sa.gov.au by American Society of Clinical Oncology X/11/2999-1/$20.00 DOI: /JCO A B S T R A C T Purpose BCR-ABL1 mutation analysis is recommended to facilitate selection of appropriate therapy for patients with chronic myeloid leukemia after treatment with imatinib has failed, since some frequently occurring confer clinical resistance to nilotinib and/or dasatinib. However, could be present below the detection limit of conventional direct sequencing. We developed a sensitive, multiplexed mass spectrometry assay (detection limit, 0.05% to 0.5%) to determine the impact of low-level after imatinib treatment has failed. Patients and Methods Mutation status was assessed in 220 patients treated with nilotinib or dasatinib after they experienced resistance to imatinib. Results were detected by sequencing in 128 patients before commencing nilotinib or dasatinib therapy (switchover). In 64 patients, 132 additional low-level were detected by mass spectrometry alone (50 of 132 were to nilotinib and/or dasatinib). When patients received the inhibitor for which the mutation confers resistance, 84% of the low-level rapidly became dominant clones detectable by sequencing, including 11 of 12 T315I. Subsequent complete cytogenetic response rates were lower for patients with at switchover detected by sequencing (0%) or mass spectrometry alone (16%) compared with patients with other or no (41% and 49%, respectively; P.001). Failure-free survival among the 100 patients with chronic phase chronic myeloid leukemia when were detected at switchover by sequencing or mass spectrometry alone was 0% and 0% compared with 51% and 45% for patients with other or no (P.003). Conclusion Detection of low-level after imatinib resistance offers critical information to guide subsequent therapy selection. If an inappropriate kinase inhibitor is selected, there is a high risk of treatment failure with clonal expansion of the mutant. J Clin Oncol by American Society of Clinical Oncology INTRODUCTION Despite responses to imatinib in the majority of patients with chronic myeloid leukemia (CML), 1 a subset of patients shows primary or acquired resistance. Several mechanisms have been associated with acquired resistance, 2 but BCR-ABL1 kinase domain are the most common known mechanism. 3-5 Nilotinib and dasatinib are active against the majority of imatinib-, but some confer clinical resistance to nilotinib (Y253H, E255K, E255V, F359V, and F359C) 6 or dasatinib (V299L, T315A, F317L, F317I, F317V, and F317C), 7-10 or both (T315I). Other have decreased in vitro sensitivity to dasatinib, such as E255K and E255V. 11,12 However, these rarely emerge with dasatinib resistance, 7-10 and clinical responses in patients with chronic phase (CP) CML with these are similar to responses in patients with other considered sensitive to dasatinib. 10 Collectively, it is estimated that the considered clinically to nilotinib and/or dasatinib are detected in more than 40% of patients who acquire during imatinib therapy 13 and are associated with a poorresponsetonilotiniband/ordasatinibtherapy. 6, by American Society of Clinical Oncology 1

2 Parker et al Detecting these at imatinib failure facilitates selection of the most appropriate alternative therapy. Therefore, mutation analysis is always required before changing to other tyrosine kinase inhibitors (TKIs) or other therapy. 14 Conventional direct sequencing of the kinase domain is the most common mutation analysis technique, 15 but it has a detection limit between 10% and 20%. Resistant may be present below this level, for example, in patients in whom treatment with TKIs has ceased, which may re-establish predominance of the unmutated BCR- ABL1 clone Several methods have been described that detect with greater sensitivity However, highly sensitive detection of low-level mutants 24 was not always correlated with their subsequent clonal expansion in patients treated with TKIs. Here we describe a sensitive, multiplex mass spectrometry assay that uses the MassARRAY system (Sequenom, San Diego, CA) to detect the nilotinib and/or dasatinib clinically plus the most common imatinib-. We examined the clinical utility of this assay to detect low-level BCR-ABL1 present after imatinib resistance and to predict their subsequent expansion during nilotinib or dasatinib therapy. The vast majority of detectable by mass spectrometry alone before nilotinib or dasatinib therapy (switchover) rapidly expanded and were associated with inferior response. PATIENTS AND METHODS Patients and Sample Preparation The patient samples analyzed were a subset of those reported in previous studies. 10,32-37 The trials were run in accordance with the Declaration of Helsinki, and approvals were obtained from the relevant institutional review boards. Peripheral blood samples from 324 patients with CML for whom treatment with imatinib failed and who were subsequently treated with nilotinib or dasatinib were analyzed at our institution. Of these, 220 patients met our criteria for inclusion: a sample was available before nilotinib or dasatinib therapy commenced, and at least one sample was subsequently collected; of the patients in CP, those who were (defined in Data Supplement), 32,37 rather than intolerant to imatinib were included. Figure 1A indicates the time points when mutation analysis was performed. One hundred patients were in CP, 64 were in accelerated phase (AP), and 56 were in blast crisis (BC). Overall, 89 patients were treated with nilotinib and 131 with dasatinib (Data Supplement). The median follow-up for therapy with nilotinib or dasatinib for patients in CP was 18 months (range, 2 to 33 months), in AP was 12 months (range, 1 to 36 months), and in BC was 3 months (range, 1 to 27 months). Peripheral blood samples collected at diagnosis from 13 patients with CP CML and 10 healthy donors were also analyzed. RNA was extracted from blood leukocytes and was reverse transcribed. 38 Mutation Detection Sequencing of the complete BCR-ABL1 kinase domain was performed as described 39 by using two cdna preparations per sample. Details on mass spectrometry BCR-ABL1 mutation analysis are provided in the Data Supplement. Allele-specific oligonucleotide polymerase chain reaction (ASO-PCR) for the T315I mutation was performed by using the method of Gruber et al. 23,30 Statistical Analysis The 2 and Fisher s exact tests were used to assess differences in treatment response according to switchover mutation status. RESULTS Establishment of the BCR-ABL1 Mass Spectrometry Assay The high-throughput mass spectrometry assay enabled simultaneous detection of 31 in up to 87 samples in a single experiment. The included in the assay were those considered clinically to nilotinib and/or dasatinib, 6,10,13 plus detected at a frequency of 1% or more in patients treated with imatinib with analyzed at our institution. 13 The single-allele base extension reaction (SABER) 40 protocol was used for the assay design (Data Supplement). This method restricts primer extension to the mutant sequence, conferring improved detection of in an abundance of unmutated sequences. However, the ability to estimate the relative amount of mutant is lost. The detection limit was determined by using dilutions of plasmids containing BCR-ABL1 and ranged from 0.05% for T315I to 0.5% for E255V and M244V. The detection limit for the majority of was approximately 0.2% (Data Supplement). The variation in detection limits is likely due to individual extension primer sequences and to primer interactions inherent in a multiplex reaction. ASO-PCR was used in a subset of patients to validate the mass spectrometry detection limit of the T315I mutation (Data Supplement). Mutation Analysis at Switchover and During Nilotinib or Dasatinib Therapy by Direct Sequencing were detected at switchover in 128 (59%) of 220 patients to imatinib (n 169 ; Fig 1B and Data Supplement). Multiple were detected in 31 patients (14%; maximum of four per patient). Of the detected by sequencing, 89% were included in the mass spectrometry assay design. Among the 169, 55 were to nilotinib and/or dasatinib and therefore would influence therapeutic decisions. 13 These 55 were detected in 50 patients (23%), and 26 of 50 patients received the inhibitor for which the confer resistance (eg, Y253H in a patient treated with nilotinib). These 26 patients had various (T315I, n 12; E255K, n 4; F317L, n 4; F359V, n 4; Y253H, n 3; E255V, n 1; and two patients had two switchover ). Mutation analysis was also performed by sequencing during nilotinib or dasatinib therapy (Data Supplement). During therapy with nilotinib or dasatinib, 115 new were detected in 86 (39%) of 220 patients at a median of 6 months (range, 1 to 36 months) after therapy started (these were not detected by sequencing at switchover). Of these 115, 91 (79%) conferred resistance to the inhibitor received (detected in 75 patients; 36 were T315I). Mutation Analysis at Switchover by Mass Spectrometry Mass spectrometry was used only at switchover to determine whether were present below the detection limit of sequencing. By mass spectrometry, 281 were detected in 131 patients. Sixty patients (27%) had multiple (maximum of 10 per patient). Of the 281 detected by mass spectrometry, 132 were not detected by sequencing (detected in 64 patients; Fig 1B and Data Supplement). detected by mass spectrometry alone are by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

3 Sensitive Mutation Analysis After Imatinib Predicts Response A Mutation analyses performed Sequencing 1. At switchover 2. During nilotinib or dasatinib therapy as mandated by the study protocol 3. End of study Mass spectrometry 1. At switchover B Sequencing at switchover 220 Patients Mass spectrometry at switchover No No 128 patients patients 131 patients patients Nilotiniband/or dasatinib- Nilotiniband dasatinib-sensitive Nilotiniband/or dasatinib- Nilotiniband dasatinib-sensitive 50 patients 55 * 78 patients patients patients 89 to therapy received sensitive to therapy received to therapy received sensitive to therapy received 26 patients patients patients patients 31 CCyR: 0% CCyR: 29% CCyR: 41% CCyR: 46% CCyR: 7% CCyR: 42% CCyR: 47% CCyR: 49% Fig 1. Flow diagram of the BCR-ABL1 mutation analyses performed. (A) Time points at which the mutation analysis was performed by each technique. (B) The switchover samples of the 220 patients were analyzed by the two independent mutation detection techniques and the patients were grouped according to the type of identified by sequencing (left) or by mass spectrometry (right). to nilotinib and/or dasatinib at switchover are those reported to confer clinical resistance, irrespective of the actual therapy received by individual patients. These are further subcategorized as or sensitive according to the actual therapy received. The number of detected and rates of achieving a complete cytogenic response (CCyR) are shown for each group. (*) Among these 50 patients, an additional 21 sensitive to nilotinib and dasatinib were detected. ( ) Among these 26 patients, an additional two to nilotinib or dasatinib that remain sensitive to the therapy received were detected. ( ) Among these 71 patients, an additional 87 sensitive to nilotinib and dasatinib were detected. ( ) Among these 45 patients, an additional 21 to nilotinib or dasatinib that remain sensitive to the therapy received were detected. referred to as low-level. Figure 2A compares the frequencies of the 20 most common 13 detected by sequencing or mass spectrometry alone. In general, the composition of the detected by the two techniques was similar, except for E459K. Of the 64 patients with low-level, 53 also had other switchover detectable by sequencing, and 11 had low-level only (one low-level mutation each; Data Supplement). Fifty (38%) of the 132 low-level were to nilotinib and/or dasatinib and therefore would influence therapeutic decisions. 13 They were detected in 35 patients in all disease phases: CP, 13 (13%) of 100; AP, 12 (19%) of 64; and BC, 10 (18%) of 56. One or more switchover that would influence therapeutic decisions after failure of imatinib therapy 13 were detected in 71 (32%) of 220 patients by mass spectrometry compared with 50 (23%) of 220 patients by sequencing (P.03; Fig 2B). Among the 50 low-level to nilotinib and/or dasatinib, 25 were detected in patients who received the inhibitor for which the mutation confers resistance (eg, Y253H in a patient treated with nilotinib). These 25 (T315I, n 12; Y253H, n 4; F317L, n 4; E255K, n 2; E255V, n 2; and F359V, n 1) were detected in 22 patients: CP, 7 (7%) of 100; AP, 8 (13%) of 64; and BC, 7 (13%) of 56). Low-Level at Switchover That Confer Resistance to the Inhibitor Received Rapidly Expanded During Therapy With Nilotinib or Dasatinib Mutation status was monitored by sequencing during therapy with nilotinib or dasatinib, and the detection of new was compared with the low-level detected at switchover by mass spectrometry. Table 1 provides details on the 22 patients with by American Society of Clinical Oncology 3

4 Parker et al A T315I n = 12 n = 12 B F359V F317L Y253H E255V E255K F359C M351T E459K G250E M244V H396R E355G L248V n = 2 n = 2 n = 3 n = 3 n = 4 n = 7 n = 5 n = 7 n = 7 n = 6 n = 6 n = 6 n = 7 n = 11 n = 8 n = 10 n = 8 n = 10 n = 10 n = 9 n = 12 n = 11 n = 11 n = 14 n = 16 n = 16 Frequency of Patients (%) All that influence therapeutic decision 6 13 Resistant to both nilotinib and dasatinib 4 Sequencing Mass spectrometry 4 Dasatinib (not T315I) Nilotinib (not T315I) F486S F359I E279K Q252H Y253F D276G n = 1 n = 1 n = 2 n = 3 n = 4 n = 3 n = 4 n = 4 n = 4 n = 5 n = 4 n = 6 Low-level by mass spectrometry alone by sequencing Type of Mutation Frequency of (%) Fig 2. Type of detected by sequencing and mass spectrometry at switchover in our patient cohort. (A) Frequency of the 20 most common to imatinib detected at switchover, arranged from the top according to the most frequently detected by mass spectrometry alone. Top, that would influence therapeutic decisions after imatinib. Bottom, that do not confer clinical resistance to nilotinib or dasatinib. The 20 represented 87% of all detected by sequencing at switchover. Frequency is expressed as a percentage of low-level detectable by mass spectrometry alone (gold bars; n 132) and detectable by sequencing (blue bars; n 169). Low-level were not detectable by sequencing. (B) Frequency of patients in whom one or more of their detected at switchover would influence therapeutic decisions after treatment with imatinib failed. For this analysis, the detected by mass spectrometry included those that were detected by mass spectrometry alone and those that were also detected by sequencing. Patients in the group to both nilotinib and dasatinib included those with T315I plus those with multiple to both nilotinib and dasatinib (eg, Y253H and F317L). By sequencing, one patient (0.5%) had multiple to both nilotinib and dasatinib (not T315I). By using mass spectrometry, five such patients (2%) were identified. low-level that confer resistance to the inhibitor received. In the 12 patients with a low-level T315I mutation, T315I rapidly expanded and became detectable by sequencing in 11 (92%) of 12 patients at a median of 3 months (range, 1 to 12 months). The exception was a patient in AP for whom the only sample received was when the patient ceased dasatinib therapy at 9 months (patient 135; Table 1). Figure 3 shows follow-up BCR-ABL1 levels and mutation analysis for three representative patients in CP. Of the remaining 13 of 25 low-level that confer resistance to the inhibitor received, 10 (77%) expanded and became detectable by sequencing during therapy with nilotinib or dasatinib (median, 3.5 months; range, 1 to 24 months). The exceptions were F317L in two patients treated with dasatinib (patients 155 and 161; Table 1) and E255K in one patient treated with nilotinib (patient 23). However, in these three patients, T315I was also detectable by sequencing at switchover (n 1), by mass spectrometry at switchover (n 1), or became detectable by sequencing after 3 months (n 1). This suggests a hierarchy of emergent according to mutant sensitivity to the drug. 11,12 Taken together, 21 (84%) of 25 low-level detected at switchover by mass spectrometry alone that confer resistance to the inhibitor received became detectable by sequencing during therapy with nilotinib or dasatinib. In three of four patients in whom the low-level mutation did not emerge, a different, more mutation was detected. The remaining 25 of 50 low-level to nilotinib or dasatinib detected at switchover in 19 patients retained sensitivity to the inhibitor received (eg, Y253H in a patient treated with dasatinib). Three (12%) of these 25 expanded and became detectable by sequencing in two patients during therapy with nilotinib or dasatinib (Y253H and E255K in a patient treated with dasatinib and F317L in a patient treated with nilotinib), but 22 (88%) were not detected (Table 2) by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

5 Sensitive Mutation Analysis After Imatinib Predicts Response Table 1. Characteristics of Patients With Low-Level Resistant to Nilotinib and/or Dasatinib Detected at Switchover by Mass Spectrometry Alone That Confer Resistance to the Inhibitor Received Month Low-Level Month of Complete Patient Disease Phase Sequencing (relative % mutant) Mass Spectrometry According to Inhibitor Received Inhibitor Received Resistant Mutation Became Detectable by Sequencing Follow-Up (months) Sequencing of Last Sample Received (relative % mutant) Cytogenetic Response, if Achieved Disease Status at Last Sample Received 15 CP D276G (50) D276G, T315I Nilotinib 1 3 T315I (100) Progression 23 CP M244V (10), E255V (50), T315I (10) M244V, E255V, T315I, E255K, Q252H, E355G Nilotinib 15 E255V (80), T315I (20) No complete hematologic response 44 CP G250E (30), M351T (25) G250E, M351T, E255V, T315I, Nilotinib 1.5, E255V, T315I; 10, F359V 10 E255V (40), T315I (40), F359V No cytogenetic response F359V (40) 67 CP M244V (50) M244V, T315I, Y253H, E255K, E355G, F359V Dasatinib 2 24 E255K (20), T315I (50) 12 Complete cytogenetic response 75 CP No mutation T315I Dasatinib 6 6 T315I (20) Minimal cytogenetic response 91 CP E355G (80) E355G, T315I, F359V Dasatinib 2 17 T315I (95) No cytogenetic response 98 CP No mutation T315I Dasatinib 2 4 T315I (90) Progression 108 AP G250E (20), E255K (70) G250E, E255K, E255V, Nilotinib 6 6 E255V (90) No cytogenetic response H396R, E459K 111 AP E255K (100) E255K, Y253H Nilotinib 9 9 Y253H (20), E255K (100) Progression 114 AP No mutation Y253H Nilotinib 2 4 Y253H (80) Progression 126 AP A397P (100) T315I Dasatinib 3 18 T315I (100) No cytogenetic response 134 AP Y253H (50), A397P (30) Y253H, F317L, M351T Dasatinib F317L (100) 3 Complete cytogenetic response 135 AP E255K (20), F359V (50) E255K, F359V, T315I, M244V, E255V, M351T, F486S Dasatinib No intervening samples received 9 M244V (90) Minor cytogenetic response 155 AP No mutation F317L Dasatinib 32 T315I (95) detected at 3 months No cytogenetic response 161 AP S417Y (60) T315I, F317L, E355G Dasatinib 9 9 T315I (100) No cytogenetic response 183 BC No mutation T315I Nilotinib 1 1 T315I (100) No cytogenetic response 192 BC F486S (30) F486S, Y253H, F317L, E459K Nilotinib 2 2 Y253H (50), F486S (30) Progression 218 BC L248V (60), F486S (10) L248V, F486S, T315I, G250E, M351T, E355G, F359V, Dasatinib 12 (no prior samples) 12 T315I (100) Minimal cytogenetic response F359C, H396R, E459K 219 BC No mutation T315I Dasatinib 1 1 T315I (50) No cytogenetic response 242 BC E355G (80) E355G, E255K Nilotinib 2 5 T315I (90), E355G (100) 2 Progression 244 BC F317L (20) F317L, Y253H Nilotinib 3 3 Y253H (50) Progression 248 BC M244V (30), G250E (35), Y253F (25) M244V, G250E, Y253F, F317L, L248V, E255V, F359C, E459K Dasatinib 4 4 V299L (50), F317L (100) No cytogenetic response Abbreviations: AP, accelerated phase; BC, blast crisis; CP, chronic phase. Standard font, detected by sequencing; bold and underlined, low-level detected by mass spectrometry alone that confer resistance to the inhibitor received; italics, low-level detected by mass spectrometry alone that retain sensitivity to the inhibitor received. Median follow-up for the 22 patients was 7.5 months (range, 1 to 32 months). Bold and underlined, low-level detected by mass spectrometry alone at switchover that confer resistance to the inhibitor received and that expanded during therapy with nilotinib or dasatinib. For CP patients, progression was defined as loss of major cytogenetic response or complete hematologic response or transformation to AP or BC; for AP patients, progression was defined as transformation to BC or as defined by the investigator; for BC patients, progression was defined as loss of major hematologic response or minor hematologic response in responding patients over 2 weeks or no reduction in switchover blast levels over 4 weeks; minimal cytogenetic response, 66% to 95% Philadelphia-positive cells; minor cytogenetic response, 36% to 65% Philadelphia-positive cells. Mutation not included in the design of the mass spectrometry assay. The low-level E255K mutation detected at switchover by mass spectrometry alone was subsequently detected by sequencing at 2 and 3 months. Sequencing of the last sample received (at 5 months) revealed E355G and the more T315I mutation. Of the 132 low-level, 82 conferred clinical resistance to imatinib but not to nilotinib or dasatinib (detected in 44 patients). Twelve (15%) of these 82 expanded during therapy with nilotinib (F359I, n 2; G250E, D276G, H396R, and E459K, n 1 each) or therapy with dasatinib (M244V, n 3; L248V, G250E, and E355A, n 1 each), but 70 (85%) were not detected by sequencing during therapy with nilotinib or dasatinib. Overall, of 132 low-level, 107 were classified as sensitive to the inhibitor received (eg, M244V or Y253H in a patient treated with dasatinib). Fifteen (14%) of these 107 low-level expanded and became detectable by sequencing during therapy with nilotinib or dasatinib compared with 21 (84%) of 25 of the low-level that confer resistance to the inhibitor received (P.001). Switchover That Confer Resistance to the Inhibitor Received Were Associated With Low Complete Cytogenetic Response Rates Patients with at switchover that confer resistance to the inhibitor received had low rates of complete cytogenetic response (CCyR) and/or its equivalent of 1% BCR-ABL1 (international scale) 41 to treatment with nilotinib or dasatinib compared with that of patients who did not have these. Figure 4A shows the CCyR rates according to mutation status and disease phase at switchover. Patients were classified according to their switchover mutation status on the basis of the following hierarchical rank: (1) detectable by sequencing that confer resistance to the inhibitor received, (2) low-level that confer resistance to the inhibitor received, (3) detected by sequencing and/or mass spectrometry that by American Society of Clinical Oncology 5

6 Parker et al A Switchover 1 month 12 months BCR-ABL1 IS (%) Undetectable Time Since Switchover (months) Sequencing MassARRAY Intensity * Wild type * * B 5,650 5,700 5,750 5,800 5,850 5,900 5,950 6,000 Mass (Da) Switchover 3 months 12 months BCR-ABL1 IS (%) T315I first detected by direct sequencing at switchover Time Since Switchover (months) Sequencing MassARRAY Intensity * * T315I * 5,650 5,700 5,750 5,800 5,850 5,900 5,950 6,000 Mass (Da) C BCR-ABL1 IS (%) T315I first detected by direct sequencing at 2 months Time Since Switchover (months) Sequencing MassARRAY Intensity Switchover 2 months 11 months * * T315I * 5,650 5,700 5,750 5,800 5,850 5,900 5,950 6,000 Mass (Da) Fig 3. Follow-up BCR-ABL1 levels and mutation analysis for three representative patients in chronic phase. The asterisk in the sequencing chromatograms indicates the T315I mutation site, 944C T. In the mass spectrometry mass spectra, the dotted lines represent the expected molecular weights of the peaks in the T315I assay. The black arrow at 5,659 Da indicates the peak if the T315I mutation is not detected. The gold arrow at 5,986 Da indicates the peak if T315I is detected. A peak from another assay in the multiplex is indicated by the black arrow head at 5,834 Da. (A) Patient 81 had no detectable by sequencing or mass spectrometry at switchover and maintained a major molecular response during dasatinib therapy. (B) Patient 23 had a T315I mutation detectable at switchover by sequencing and mass spectrometry. This patient experienced failure of nilotinib therapy and had no significant reduction in BCR-ABL1. (C) Patient 91 had T315I detected by mass spectrometry alone at switchover. The mutation rapidly expanded and was detectable by sequencing at 2 months. The patient experienced failure of dasatinib therapy and, similar to patient 23, had no significant reduction in BCR-ABL by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

7 Sensitive Mutation Analysis After Imatinib Predicts Response Table 2. Low-Level BCR-ABL1 Resistant to Nilotinib and/or Dasatinib That Were Detectable at Switchover by Mass Spectrometry Alone and Their Subsequent Detection by Sequencing According to the Inhibitor Received BCR-ABL1 Mutation T315I (n 12) No. Emerged/ No. Detected % Nilotinib Resistant (n 28) No. Emerged/ No. Detected % Dasatinib Resistant (n 10) No. Emerged/ No. Detected % Emerged on nilotinib 3/ /9 89 1/6 17 Emerged on dasatinib 8/9 89 2/ /4 50 Not including T315I. Low-Level at Switchover That Confer Resistance to the Inhibitor Received Were Associated With Treatment Failure in Patients in CP The association between switchover and failure-free survival during therapy with nilotinib or dasatinib was determined for the patients in CP according to the provisional definitions of the European LeukemiaNet. 14 At 18 months, the frequency of failure-free survival was 0% for patients with that confer resistance to the inhibitor received, detected at switchover by sequencing or by mass spectrometry alone, compared with 51% for patients with other and 45% for those with no (P.003, Fisher s exact test). Figure 4B depicts the duration of failure-free survival for these groups. remain sensitive to the inhibitor received (other ), and (4) no mutation detected by either method. The CCyR rates for the four groups were 0%, 16%, 41%, and 49%, respectively (P.001). Among the patients with low-level that confer resistance to the inhibitor received, three achieved CCyR (CP, AP, BC, n 1 each), including one of the 12 patients with low-level T315I. Interestingly, this patient (patient 67; Table 1) maintained a CCyR with dasatinib treatment after T315I became detectable by sequencing, as did another patient with a low-level mutation to dasatinib (patient 134; F317L). In eight of the 95 patients who had only other, their at switchover were those rare that were not included in the mass spectrometry assay design. Five (63%) achieved CCyR. DISCUSSION We demonstrated that BCR-ABL1 present below the detection limit of conventional direct sequencing before treatment with second-generation TKIs after patients have exhibited imatinib resistance have biologic and clinical significance. Sensitive detection of low-level that confer clinical resistance to nilotinib and/or dasatinib 6,10 is highly predictive of their rapid clonal expansion under the selective pressure of nilotinib or dasatinib therapy. that would influence therapeutic decisions after failure of imatinib therapy 13 were detected in significantly more patients by mass spectrometry than by sequencing (32% v 23%; P.03). Patients with low-level at switchover that confer resistance to the A Patients (%) Resistant by sequencing (0% for all phases) Low-level Other No Chronic phase Accelerated phase Blast crisis 2 Disease Phase at Switchover No. at risk CCyR Total B Probability of Failure-Free Survival (%) Resistant by sequencing (n = 9) Low-level (n = 6) Other (n = 47) No (n = 38) Time Since Switchover (months) Fig 4. Rates of complete cytogenetic response (CCyR) and failure-free survival according to the switchover mutation status. Within the disease phases, each patient was grouped into one of four independent groups on the basis of their switchover mutation status and the following hierarchical rank: (1) one or more detectable by sequencing that confers resistance to the inhibitor received termed by sequencing in the figure; (2) one or more low-level detected by mass spectrometry alone that confers resistance to the inhibitor received termed low-level ; (3) other detectable by sequencing and/or mass spectrometry; and (4) no mutation detected by either method. Three patients (chronic phase, n 1; accelerated phase, n 2; patients 23, 108, and 111; Table 1) had both a mutation detectable by sequencing and a low-level mutation, and these patients were included in group 1. (A) The rates of achieving a CCyR for the 220 patients according to their disease phase and mutation status at switchover. (B) Failure-free survival by 18 months of nilotinib or dasatinib therapy for the 100 chronic-phase patients according to their mutation status at switchover (as defined above). The CCyR rates and failure-free survival shown in this figure cannot be directly compared with those in Figure 1B and Data Supplement because of the independent hierarchy used to categorize patients for this analysis by American Society of Clinical Oncology 7

8 Parker et al therapy received had low CCyR rates compared with patients without these. Furthermore, in patients in CP, low-level were associated with poor failure-free survival with therapy using nilotinib or dasatinib (0%), which was the same as that for patients with detectable by sequencing at switchover. Importantly, when a patient received the secondgeneration TKI that retains sensitivity to a low-level mutation present at switchover, the mutation rarely expanded. The significance of low-level for TKI resistance may vary according to clinical context and the detection limit of the mutation assay. In patients in all disease phases after imatinib resistance, we found that detection of low-level by mass spectrometry (average detection limit, 0.2%) was highly predictive of their expansion on alteration of the selective environment. This is in contrast to the detection of low-level in imatinib-naive patients (detection limit, 0.001%) reported in a study by Willis et al, 24 which did not invariably expand. In patients with Philadelphia-positive acute lymphoblastic leukemia, however, the situation is different; detection of before imatinib therapy was highly predictive of their subsequent clonal expansion. 42 In support of our findings, Ernst et al 43 reported that low-level F317L detected by ASO-PCR expanded during dasatinib therapy in four patients with CML (after imatinib therapy failed). Although T315I is not the only clinically relevant mutation for patients for whom imatinib therapy has failed, outcomes for patients with this mutation are poor, and there is currently no approved TKI that retains sensitivity to T315I. Median overall survival and progression-free survival for patients with T315I are significantly lower than for patients without T315I. 44,45 The current recommendation for patients with this mutation is allogeneic stem-cell transplantation 14 or experimental therapy. By using mass spectrometry in our cohort of patients, we detected T315I in 11 patients at switchover, up to 12 months earlier than by sequencing. Only one of these patients achieved and maintained a CCyR, which suggests that patients with low-level T315I after imatinib resistance would benefit from transplantation or experimental therapy such as ponatinib 46 rather than nilotinib or dasatinib. However, T315I can emerge in a proportion of patients during therapy in whom the mutation could not be detected by mass spectrometry at switchover (25 in our cohort). The responses for the other patients with pre-existing low-level to nilotinib and/or dasatinib were also generally poor if the patient received the inhibitor for which the mutation confers resistance. However, in addition to the patient we referred to earlier with low-level T315I who maintained CCyR, a patient with low-level F317L at switchover also achieved and maintained CCyR on dasatinib. In both patients, several other were detected before and during dasatinib therapy. Whether other modified the sensitivity of T315I and F317L to dasatinib is unknown. 47 Interestingly, a recent case report described a patient with CML with a dominant F317L mutation who achieved and maintained a major molecular response with dasatinib therapy. 48 Among the detected by sequencing at switchover in our cohort, 11% could not be detected by the current design of the mass spectrometry assay. These are rare, have each been detected in less than 1% of all imatinib-treated patients tested at our institution since 2001, 13 and are not considered clinically to nilotinib or dasatinib. 6,10 Indeed, the patients in our study with these rare at switchover achieved a CCyR rate comparable with that of patients with other considered sensitive to nilotinib and dasatinib. This suggests there is little advantage in adding these to the multiplex assay. However, the mass spectrometry assay is amenable to addition of when necessary, which may be important for new inhibitors under development. 49 Alternatively, a simplified assay could be developed to incorporate only the that are clinically relevant for patients treated with second- or thirdgeneration TKIs. One such multiplex ASO-PCR assay has been described 30 that detects some dasatinib- (T315I, V299L, and F317L); however, the assay could not distinguish between the three, which could limit its applicability for therapeutic decisions. The singleplex ASO-PCR assay that specifically detects the T315I mutation, 31 although highly sensitive, does not provide a comprehensive assessment of that could have an impact on therapy response. Some of the low-level detected at switchover in patients who received the inhibitor that remains sensitive to the mutation expanded and became detectable by sequencing (14%). These presumably expanded because of differential sensitivity to the TKI compared with the unmutated BCR-ABL1 clone, or as part of compound mutants, either as passengers or by altering drug sensitivity. Compound could alter the sensitivity to TKIs compared with single. 8,47 In summary, sensitive detection of after imatinib resistance by mass spectrometry was highly predictive of their subsequent expansion and treatment failure if the patient was treated with the inappropriate inhibitor. This demonstrates the clinical utility of sensitive mutation detection to identify patients after imatinib failure who harbor biologically significant that are below the level of detection with direct sequencing and who may benefit from alternative TKI therapy or stem-cell transplantation. The current recommendation is that mutation analysis be performed before changing TKIs. 14 Although the mass spectrometry assay described herein is experimental, the findings suggest that sensitive analysis of a limited range of clinically relevant could be incorporated into clinical practice to aid selection of appropriate therapy after treatment with imatinib has failed. AUTHORS DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a U are those for which no compensation was received; those relationships marked with a C were compensated. For a detailed description of the disclosure categories, or for more information about ASCO s conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors. Employment or Leadership Position: Darryl L. Irwin, Sequenom (C) Consultant or Advisory Role: Timothy P. Hughes, Ariad, Bristol-Myers Squibb, Novartis; Susan Branford, Novartis Stock Ownership: None Honoraria: Timothy P. Hughes, Ariad, Bristol-Myers Squibb, Novartis; Susan Branford, Bristol-Myers Squibb, Novartis Research Funding: Timothy P. Hughes, Bristol-Myers Squibb, Novartis; Susan Branford, Ariad, Bristol-Myers Squibb, Novartis Expert Testimony: None Other Remuneration: None by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

9 Sensitive Mutation Analysis After Imatinib Predicts Response AUTHOR CONTRIBUTIONS Conception and design: Wendy T. Parker, Rebecca M. Lawrence, Darryl L. Irwin, Hamish S. Scott, Timothy P. Hughes, Susan Branford Provision of study materials or patients: Darryl L. Irwin Collection and assembly of data: Wendy T. Parker, Rebecca M. Lawrence, Musei Ho, Darryl L. Irwin Data analysis and interpretation: Wendy T. Parker, Darryl L. Irwin, Hamish S. Scott, Timothy P. Hughes, Susan Branford Manuscript writing: All authors Final approval of manuscript: All authors REFERENCES 1. Deininger M, O Brien SG, Guilhot F, et al: International Randomized Study of Interferon Vs STI571 (IRIS) 8-Year Follow up: Sustained survival and low risk for progression or events in patients with newly diagnosed chronic myeloid leukemia in chronic phase (CML-CP) treated with imatinib. 51st ASH Annual Meeting and Exposition, New Orleans, LA, December 5-8, 2009 (abstr 1126) 2. Apperley JF: Part I: Mechanisms of resistance to imatinib in chronic myeloid leukaemia. Lancet Oncol 8: , Gorre ME, Mohammed M, Ellwood K, et al: Clinical resistance to STI-571 cancer therapy caused by BCR-ABL gene mutation or amplification. Science 293: , von Bubnoff N, Schneller F, Peschel C, et al: BCR-ABL gene in relation to clinical resistance of Philadelphia-chromosome-positive leukaemia to STI571: A prospective study. Lancet 359: , Hochhaus A, Kreil S, Corbin AS, et al: Molecular and chromosomal mechanisms of resistance to imatinib (STI571) therapy. Leukemia 16: , Hughes T, Saglio G, Branford S, et al: Impact of baseline BCR-ABL on response to nilotinib in patients with chronic myeloid leukemia in chronic phase. J Clin Oncol 27: , Soverini S, Colarossi S, Gnani A, et al: Resistance to dasatinib in Philadelphia-positive leukemia patients and the presence or the selection of at residues 315 and 317 in the BCR-ABL kinase domain. Haematologica 92: , Shah NP, Skaggs BJ, Branford S, et al: Sequential ABL kinase inhibitor therapy selects for compound drug- BCR-ABL with altered oncogenic potency. J Clin Invest 117: , Soverini S, Gnani A, Colarossi S, et al: Philadelphiapositive patients who already harbor imatinib- Bcr-Abl kinase domain have a higher likelihood of developing additional associated with resistance to second- or third-line tyrosine kinase inhibitors. Blood 114: , Müller MC, Cortes JE, Kim DW, et al: Dasatinib treatment of chronic-phase chronic myeloid leukemia: Analysis of responses according to preexisting BCR- ABL. Blood 114: , O Hare T, Walters DK, Stoffregen EP, et al: In vitro activity of Bcr-Abl inhibitors AMN107 and BMS against clinically relevant imatinib- Abl kinase domain mutants. Cancer Res 65: , Redaelli S, Piazza R, Rostagno R, et al: Activity of bosutinib, dasatinib, and nilotinib against 18 imatinib- BCR/ABL mutants. J Clin Oncol 27: , Branford S, Melo JV, Hughes TP: Selecting optimal second-line tyrosine kinase inhibitor therapy for chronic myeloid leukemia patients after imatinib failure: Does the BCR-ABL mutation status really matter? Blood 114: , Baccarani M, Cortes J, Pane F, et al: Chronic myeloid leukemia: An update of concepts and management recommendations of European LeukemiaNet. J Clin Oncol 27: , Jones D, Kamel-Reid S, Bahler D, et al: Laboratory practice guidelines for detecting and reporting BCR-ABL drug resistance in chronic myelogenous leukemia and acute lymphoblastic leukemia: A report of the Association for Molecular Pathology. J Mol Diagn 11:4-11, Müller MC, Lahaye T, Hochhaus A: Resistance to tumor specific therapy with imatinib by clonal selection of mutated cells [in German]. Dtsch Med Wochenschr 127: , Branford S, Hughes TP: Mutational analysis in chronic myeloid leukemia: When and what to do? Curr Opin Hematol 18: , Hanfstein B, Müller MC, Kreil S, et al: Dynamics of mutant BCR-ABL-positive clones after cessation of tyrosine kinase inhibitor therapy. Haematologica 96: , Roche-Lestienne C, Soenen-Cornu V, Grardel- Duflos N, et al: Several types of of the Abl gene can be found in chronic myeloid leukemia patients to STI571, and they can pre-exist to the onset of treatment. Blood 100: , Kreuzer KA, Le Coutre P, Landt O, et al: Preexistence and evolution of imatinib mesylate clones in chronic myelogenous leukemia detected by a PNA-based PCR clamping technique. Ann Hematol 82: , Deininger MW, McGreevey L, Willis S, et al: Detection of ABL kinase domain with denaturing high-performance liquid chromatography. Leukemia 18: , Sorel N, Chazelas F, Brizard A, et al: Doublegradient-denaturing-gradient gel electrophoresis for mutation screening of the BCR-ABL tyrosine kinase domain in chronic myeloid leukemia patients. Clin Chem 51: , Gruber FX, Lamark T, Anonli A, et al: Selecting and deselecting imatinib- clones: Observations made by longitudinal, quantitative monitoring of mutated BCR-ABL. Leukemia 19: , Willis SG, Lange T, Demehri S, et al: Highsensitivity detection of BCR-ABL kinase domain in imatinib-naive patients: Correlation with clonal cytogenetic evolution but not response to therapy. Blood 106: , Khorashad JS, Anand M, Marin D, et al: The presence of a BCR-ABL mutant allele in CML does not always explain clinical resistance to imatinib. Leukemia 20: , Vivante A, Amariglio N, Koren-Michowitz M, et al: High-throughput, sensitive and quantitative assay for the detection of BCR-ABL kinase domain. Leukemia 21: , Ernst T, Erben P, Müller MC, et al: Dynamics of BCR-ABL mutated clones prior to hematologic or cytogenetic resistance to imatinib. Haematologica 93: , Poláková KM, Lopotová T, Klamová H, et al: High-resolution melt curve analysis: Initial screening for in BCR-ABL kinase domain. Leuk Res 32: , Pelz-Ackermann O, Cross M, Pfeifer H, et al: Highly sensitive and quantitative detection of BCR- ABL kinase domain by ligation PCR. Leukemia 22: , Gruber FX, Ernst T, Kiselev Y, et al: Detection of drug- clones in chronic myelogenous leukemia patients during dasatinib and nilotinib treatment. Clin Chem 56: , Nicolini FE, Chomel JC, Roy L, et al: The durable clearance of the T315I BCR-ABL mutated clone in chronic phase chronic myelogenous leukemia patients on omacetaxine allows tyrosine kinase inhibitor rechallenge. Clin Lymphoma Myeloma Leuk 10: , Kantarjian HM, Giles F, Gattermann N, et al: Nilotinib (formerly AMN107), a highly selective BCR-ABL tyrosine kinase inhibitor, is effective in patients with Philadelphia chromosome-positive chronic myelogenous leukemia in chronic phase following imatinib resistance and intolerance. Blood 110: , le Coutre P, Ottmann OG, Giles F, et al: Nilotinib (formerly AMN107), a highly selective BCR- ABL tyrosine kinase inhibitor, is active in patients with imatinib- or -intolerant acceleratedphase chronic myelogenous leukemia. Blood 111: , Giles FJ, Larson RA, Kantarjian HM, et al: Nilotinib in patients with Philadelphia chromosomepositive chronic myelogenous leukemia in blast crisis (CML-BC) who are or intolerant to imatinib. J Clin Oncol 26:376s, 2008 (suppl; abstr 7017) 35. Cortes J, Rousselot P, Kim DW, et al: Dasatinib induces complete hematologic and cytogenetic responses in patients with imatinib- or -intolerant chronic myeloid leukemia in blast crisis. Blood 109: , Apperley JF, Cortes JE, Kim DW, et al: Dasatinib in the treatment of chronic myeloid leukemia in accelerated phase after imatinib failure: The START a trial. J Clin Oncol 27: , Hochhaus A, Kantarjian HM, Baccarani M, et al: Dasatinib induces notable hematologic and cytogenetic responses in chronic-phase chronic myeloid leukemia after failure of imatinib therapy. Blood 109: , Branford S, Hughes T: Diagnosis and monitoring of chronic myeloid leukemia by qualitative and quantitative RT-PCR, in Illand H, Hertzgerg M, Marlton P (eds): Myeloid Leukemia: Methods and Protocols (Methods in Molecular Medicine). Totawa, New Jersey, Humana Press, 2006, pp Branford S, Rudzki Z, Walsh S, et al: Detection of BCR-ABL in patients with CML treated with imatinib is virtually always accompanied by clinical resistance, and in the ATP phosphate-binding loop (P-loop) are associated with a poor prognosis. Blood 102: , Ding C, Chiu RW, Lau TK, et al: MS analysis of single-nucleotide differences in circulating nucleic acids: Application to noninvasive prenatal diagnosis. Proc Natl Acad Sci U S A 101: , Hughes TP, Hochhaus A, Branford S, et al: Long-term prognostic significance of early molecular response to imatinib in newly diagnosed chronic by American Society of Clinical Oncology 9

10 Parker et al myeloid leukemia: An analysis from the International Randomized Study of Interferon and STI571 (IRIS). Blood 116: , Pfeifer H, Wassmann B, Pavlova A, et al: Kinase domain of BCR-ABL frequently precede imatinib-based therapy and give rise to relapse in patients with de novo Philadelphiapositive acute lymphoblastic leukemia (Ph ALL). Blood 110: , Ernst T, Gruber FX, Pelz-Ackermann O, et al: A co-operative evaluation of different methods of detecting BCR-ABL kinase domain in patients with chronic myeloid leukemia on second-line dasatinib or nilotinib therapy after failure of imatinib. Haematologica 94: , Nicolini FE, Mauro MJ, Martinelli G, et al: Epidemiologic study on survival of chronic myeloid leukemia and Ph( ) acute lymphoblastic leukemia patients with BCR-ABL T315I mutation. Blood 114: , Nicolini FE, Morisset S, Hochhaus A, et al: The presence of the BCR-ABL T315I mutation in chronic phase chronic myelogenous leukemia to tyrosine kinase inhibitors profoundly compromises overall survival and progression free survival: Preliminary results of a matched pair analysis. 52nd ASH Annual Meeting and Exposition, Orlando, FL, 2010 (abstr 3410) 46. O Hare T, Shakespeare WC, Zhu X, et al: AP24534, a pan-bcr-abl inhibitor for chronic myeloid leukemia, potently inhibits the T315I mutant and overcomes mutation-based resistance. Cancer Cell 16: , von Bubnoff N, Manley PW, Mestan J, et al: Bcr-Abl resistance screening predicts a limited spectrum of point to be associated with clinical resistance to the Abl kinase inhibitor nilotinib (AMN107). Blood 108: , Faber E, Mojzikova R, Plachy R, et al: Major molecular response achieved with dasatinib in a CML patient with F317L BCR-ABL kinase domain mutation. Leuk Res 34:e91 e93, Zhang J, Adrián FJ, Jahnke W, et al: Targeting Bcr-Abl by combining allosteric with ATP-bindingsite inhibitors. Nature 463: , 2010 Acknowledgment We thank Franz Gruber, MD, for his assistance in setting up the allele-specific oligonucleotide polymerase chain reaction (ASO-PCR) assay and the staff of the Leukemia Unit, Genetics and Molecular Pathology, SA Pathology, for their excellent technical support by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

11 Sensitive detection of BCR-ABL1 in patients with chronic myeloid leukemia after imatinib resistance is predictive of outcome during subsequent therapy Wendy T Parker, Rebecca M Lawrence, Musei Ho, Darryl L Irwin, Hamish S Scott, Timothy P Hughes, Susan Branford Data Supplement Methods Further details on patient samples The patients analyzed in this study are a subset of the previously reported studies: 1 nilotinib trial (CAMN107A 2101 Phase 2) 1-3 and 4 dasatinib trials (CA (START-A) 4, CA (START- B) 5, CA (START-C) 6,7, CA (START-L) 5 ). Imatinib resistance for the CP patients was defined as failure to achieve complete hematologic response (CHR) after 3 months, cytogenetic response after 6 months, major cytogenetic response (MCyR) after 12 months, or loss of a hematologic or cytogenetic response at any time during treatment with imatinib. 1,7 Patients had to have been treated with an imatinib dose of at least 600 mg daily for 3 months (nilotinib-treated patients 1 ), or greater than 600 mg daily (dasatinib-treated patients 7 ) or 600 mg daily for those with a highly imatinib mutation (dasatinib-treated patients 7 ). Peripheral blood samples of patients treated with nilotinib were scheduled to be collected prior to nilotinib therapy (switchover), at 1, 2 and 3 months and every 3 months thereafter during follow-up. Samples of CP patients treated with dasatinib had the same collection time points, whereas samples of AP and BC patients treated with dasatinib were collected at switchover, study end, and 3 monthly throughout therapy if a complete cytogenetic response (CCyR) was achieved. Additional samples were received during therapy for some of these patients who did not achieve a CCyR. Mutation analysis was scheduled to be performed by sequencing for all patients at switchover, and in the event of progression or loss of response, and at end of study. Mutation analysis was also performed by sequencing in the event of a greater than 2-fold rise in BCR-ABLl transcript level. For nilotinib-treated patients with detectable by sequencing at switchover, mutation analysis was performed on every sample as 1