Clinical OMICs Presents. Conquering CML. The Breakthrough Paradigm Change of Treatment-free Remission

Transcription

1 Clinical OMICs Presents Conquering CML The Breakthrough Paradigm Change of Treatment-free Remission Sponsored by Produced by

2 Conquering CML The Breakthrough Paradigm Change of Treatment-free Remission 4 A MESSAGE FROM MOLECULARMD Fritz Eibel, Senior Vice President of Marketing, MolecularMD 5 FEATURE ENESTfreedom Trial Drives Clinical Adoption of Treatment-free Remission as Therapeutic Goal in Chronic Myelogenous Leukemia 11 INFOGRAPHIC Treatment-free Remission: The ability for eligible patients who achieved a sustained 4.5 to maintain deep molecular response after discontinuing Tasigna 12 Q&A with Chad Galderisi, Executive Vice President, Chief Medical Officer and Co-founder of MolecularMD Sponsored by 15 NEWS Dx BCR-ABL Test Supports Tasigna Label Update 17 POSTERS Photo credits: cover & above: DepositPhotos/lighthouse and AdobeStock/ravital; p. 5: AdobeStock/Picasa; p. 6: DepositPhotos/ Dmyrto_Z; p.7: Wikipedia Creative Commons/ J3D3; p. 8: AdobeStock/Peathegee, Inc.; p.11: Infographic: Nora Wertz; p. 15: DepositPhotos/auntspray. 3 A Clinical OMICs ebook

3 Welcome to the MolecularMD ebook Knocking on the Front Door of Precision Medicine Great advancements have been made in the Precision Medicine movement, which are a result of gaining a deeper biological understanding of disease through the use of powerful life science tools. Achieving insights like understanding a therapeutic s mode of action, or identifying a cancer s vulnerability are proving to be transformative to healthcare. Biomarkers have and continue to play a key role in the successes for targeted drugs and immunotherapies. In fact, PhRMA states that 73% of the cancer medicines in the pipeline are likely to be personalized medicines, and that over the next five years there will be a 69% increase in the development of these types of therapies. Biomarkers have become the new norm and they are integral to a drug development program. Today, the use of a biomarker for patient selection improves the probability three-fold for approval of a drug candidate. At MolecularMD, it is our belief that diagnostics aid in clinical decision making, and our role is to assist clients in de-risking their development programs. With more than ten years of experience, MolecularMD has built out the capabilities, and knowledge, to serve as a trusted advisor to our clients in the biopharma industry. One of MolecularMD s demonstrated successes has been in the area of Chronic Myeloid Leukemia (CML) and a novel treatment approach which is now being used by progressive clinicians. In this publication you will find many interesting details about the Dx BCR-ABL Test being deployed as a Companion Diagnostic. MolecularMD s Dx Test is used in conjunction with Tasigna to identify, initiate, and monitor Treatment-free Remission, which enables patients to safely go off the drug. This is personalized medicine in its truest sense, providing patients with the right drug at the right time. In the spirit of advancing a dialogue, I would like to invite you to provide feedback on the articles in this ebook. Our success in Precision Medicine relies on a collaborative exchange between the research, clinical, pharma and regulatory communities. Fritz Eibel Senior VP, Marketing feibel@molecularmd.com 4 A Clinical OMICs ebook

4 ENESTfreedom Trial Drives Clinical Adoption of Treatment-free Remission as Therapeutic Goal in Chronic Myelogenous Leukemia The international team of researchers that conducted the ENESTfreedom clinical trial concluded that frontline therapy with the second-generation tyrosine kinase inhibitor (TKI) nilotinib could be safely stopped in patients with chronic-phase (CP) chronic myeloid leukemia (CML) who had achieved a deep molecular response % IS. 1 Discontinuation of TKI therapy led to treatment-free remission (TFR) in more than half of patients. The safety of attempting discontinuation of treatment was validated by the fact that nearly all patients who experienced relapse were able to once again achieve their previous level of molecular response when TKI therapy was re-started. Furthermore, no patients in the ENESTfreedom study died of CML or had disease progression. TKI treatment for CML-CP is no longer necessarily a lifelong therapy. Testing will determine if the patient is in the 49% who achieve TFR. TFR in ENESTfreedom was defined as a sustained major molecular response (<M) for at least 48 weeks, and those results have now been updated to 96 weeks. 2! A Clinical OMICs ebook

5 Regular testing identifies patients who fail to maintain M. Upon restarting treatment with Tasigna, 98% of patients regain M. ENESTfreedom was the first prospective trial to focus on TFR following frontline nilotinib treatment. TKI treatment for Ph+ CML is no longer necessarily a lifelong therapy. It is a treatment that can be discontinued safely in a proportion of patients, says Dr. Andreas Hochhaus, Professor of Internal Medicine, Hematology and Oncology, Jena University Hospital, Jena, Germany, and lead author of the ENESTfreedom study. Over the past few years, as TFR study data have begun to suggest that the duration of a deep molecular response (D, 4.0, BCR-ABL< or = 0.01% IS) might be even more relevant than the duration of TKI therapy as a predictor of TFR success, the ability of second-generation TKIs such as nilotinib to induce more dramatic decreases in the leukemic load in a greater proportion of patients treated has become an increasingly important factor. 3 The thinking was that second-generation TKIs, might allow us to recruit more patients to a deep molecular response, and more quickly, leading to earlier achievement of TFR and potentially more patients ultimately achieving TFR, says Professor Tim Hughes of the University of Adelaide and South Australian Health and Medical Research Institute (SAHI), Australia, who has been involved in TFR trials in CML for more than a decade. The ENESTfreedom results supported this supposition. How large is the cohort of Ph+ CML patients that could potentially be eligible to discontinue TKI treatment? Based on the initial ENESTfreedom results, 51.6% of patients with at least 3 years of frontline nilotinib therapy and sustained deep molecular response (D) were able to achieve TFR. The recently updated ENESTfreedom data, which extended TFR data collection from 48 to 96 weeks, showed that 48.9% of patients entering TFR maintained that status without the need for re-initiation of TKI treatment at 96 weeks follow-up. Translating Trial Results to Clinical Practice Results of the ENESTfreedom trial and other recent studies such as Euro-Ski, as well as numerous review articles on the outlook for patients with Ph+ CML, have raised awareness of the potential for TFR in CML. 4-7 The growing body of literature 6 A Clinical OMICs ebook

6 is leading to increased acceptance of and urgency for TFR as a new therapeutic goal in Ph+ CML The ENESTfreedom study used predefined patient eligibility criteria for safe discontinuation of nilotinib. We have minimum recommendations for discontinuation, but not optimum recommendations, says Dr. Hochhaus. Determining optimum eligibility criteria will require comparative studies that evaluate patient outcomes depending on variation of at least three key parameters: duration of TKI treatment; duration of deep molecular response; and depth of molecular response. drug cessation is conducted under the appropriate conditions, it is very safe to attempt. Appropriate conditions include not only meeting the eligibility criteria, but also monthly monitoring patients in TFR using robust and reliable RT-qPCR. An additional clinical study ENESTop was performed for second line nilotonib treatment. 12 Setting Sights on TFR The potential advantages of TFR in Ph+ CML are many and encompass a combination of treatment-, cost-, and patient-related factors. 13 Overall, treatment-related benefits of discontinuing a TKI include the reduced risk of an adverse drug reaction We have minimum recommendations and of future drug interactions. for discontinuation, but not optimum Cost factors include eliminating recommendations. patient copays and insurer Dr. Andreas Hochhaus, costs for ongoing TKI treatment senior author of the ENESTfreedom study as well as the expense to both patients and the healthcare The need for more data should not stand in the way of starting to change clinical practice, says Dr. Hughes. As long as system for treating the effects of noncompliance that may result from poor tolerability of a TKI. Chronic myeloid leukemia. From a patient perspective, although TKIs are generally well tolerated, side effects may compromise an individual s health, ability to function, and health-related quality of life (QoL). Clinical trials of TKIs that have incorporated a patient-reported outcome measure developed specifically for use by patients with Ph+ CML have shown that while most patients experience few TKI-related symptoms or negative effects on QoL, 30-40% reported symptoms including fatigue and muscle soreness or cramping A Clinical OMICs ebook

7 completed at least 2 years of frontline treatment with nilotinib on entry into the study. All patients had to have achieved a molecular response (4.5, BCR-ABL < % IS) to TKI treatment to participate in the study. Treatment-free Remission: For young women with CML, in particular, the desire to start a family may drive an urgency to discontinue treatment. These symptoms may occur at moderate to severe levels, can impact QoL and compliance, and may continue indefinitely. Importantly, symptoms that may seem tolerable to a patient when they accompany a short-term, or at least time-limited, therapy can present a greater burden with lifelong treatment. For young women with CML, in particular, the desire to start a family may drive an urgency to discontinue treatment. Freedom from TKI Treatment in Ph+ CML ENESTfreedom was a phase 2 clinical trial that enrolled patients 18 years of age or older who had Philadelphia chromosome (BCR-ABL1)-positive Ph+ CML and had On study entry, patients began a 1-year consolidation phase during which they continued nilotinib treatment, with ongoing molecular response assessment every 12 weeks. Patients with sustained D (4.0) were eligible to stop nilotinib treatment and enter the TFR phase of the study. Patients in whom a D was not sustained during the TFR phase, nilotinib treatment was re-initiated. The primary endpoint of the trial was the proportion of patients who remained below M at 48 weeks post-tki discontinuation. Secondary endpoints included the proportion of patients off treatment with a sustained D at 48 weeks, treatment-free survival (TFS) over time, the need to re-initiate nilo- 8 A Clinical OMICs ebook

8 [88] patients; 85 [87] regained M and 76 [81] achieved 4.5 by the end of the study. QoL scores were high at the end of the consolidation phase and showed minimal change on cessation or re-initiation of treatment with nilotinib. Adverse events decreased from 83.2% during the consolidation phase (among patients who went on to the TFR phase) to 65.8% during the 48-week TFR period. The ENESTfreedom trial data supports the feasibility and saftey of TFR with precise and accurate monitoring of patients Molecular Response () levels. tinib, and disease progression. The trial monitored for adverse events and for patient-reported outcomes throughout the study, which included overall QoL and presence and severity of problems related to mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Of the 203 patients who completed the consolidation phase, 190 (88.4%) entered the TFR phase of the study. At the 48-week point of TFR, 98 of the 190 (51.6%) patients remained below M without requiring treatment re-initiation [93/190 (48.9%) at 96 weeks]. Re-initiation of nilotinib due to loss of D during TFR was necessary in 86 The results of the ENESTfreedom trial clearly demonstrated that TKI treatment can be safely discontinued in appropriate patients that meet the eligibility criteria and with ongoing monitoring of molecular response using RT-qPCR. This trial adds to the growing body of evidence supporting the feasibility and safety of TFR in patients with Ph+ CML. Greater awareness and acceptance of TFR as a therapeutic goal will help bring this breakthrough into mainstream practice. 9 A Clinical OMICs ebook

9 References 1. Hochhaus A, Masszi T, Giles FJ, et al. Treatment-free remission following frontline nilotinib in patients with chronic myeloid leukemia in chronic phase: Results from the ENEST freedom study. Leukemia 2017;31: Ross DM, Masszi T, Casares MTG. Durable treatment-free remission in patients with chronic myeloid leukemia in chronic phase following frontline nilotinib: 96-week update of the ENESTfreedom study. J Cancer Res Clin Oncol 2018; in press. 3. Hochhaus A, Saglio G, Hughes TP, et al. Longterm benefits and risks of frontline nilotinib vs imatinib for chronic myeloid leukemia in chronic phase: 5-year update of the randomized ENESTnd trial. Leukemia 2016;30: Hughes TP, Hochhaus A, Kim D-W, et al. Treatment-free remission (TFR) among patients with chronic myeloid leukemia in chronic phase (CML-CP) not initially eligible for treatment discontinuation due to unstable deep molecular response (D): Enestfreedom and Enestop. Blood 2017;130(Suppl 1): Mahon F-X, Richter J, Guilhot J, et al. Cessation of tyrosine kinase inhibitors treatment in chronic myeloid leukemia patients with deep molecular response: Results of the Euro-Ski trial. Blood 2016;128: Saussele S, Richter J, Guilhot J, et al. Duration of Deep Molecular Response Has Most Impact on the Success of Cessation of Tyrosine Kinase Inhibitor Treatment in Chronic Myeloid Leukemia Results from the EURO-SKI Trial. Oral Presentation at American Society of Hematology annual meeting, December 2017, Atlanta, GA. 7. Takahashi N, Tauchi T, Kitamura K, et al. Deeper molecular response is a predictive factor for treatment-free remission after imatinib discontinuation in patients with chronic phase chronic myeloid leukemia: the JALSG-STIM213 study. In J Hematol 2017;doi.org/ /s x. 8. Hughes TP, Ross DM. Moving treatment-free remission into mainstream clinical practice in CML. Blood 2016;128(1): Saussele S, Richter J, Hochhaus A, Mahon F-X. The concept of treatment-free remission in chronic meyloid leukemia. Leukemia 2016;30: Elsayed AG, Srivastava R, Jamil MO. Treatment-free remission: A new therapeutic goal in chronic myelogenous leukemia. Curr Oncol Rep 2017;19(12): Narra RK, Flynn KE, Atallah E. Chronic myeloid leukemia the promise of tyrosine kinase inhibitor discontinuation. Curr Hematol Malig Rep 2017;12(5): Annals of Internal Medicine Feb 20. doi: /M [Epub ahead of print] Treatment-Free Remission After Second-Line Nilotinib Treatment in Patients With Chronic Myeloid Leukemia in Chronic Phase: Results From a Single-Group, Phase 2, Open-Label Study. 13. Khoury HG, Williams LA, Atallah E, Hehlmann R. Chronic Myeloid Leukemia: What Every Practitioner Needs to Know in Am Soc Clin Oncol Educ Book 2017;37: A Clinical OMICs ebook

10 Dx = The FDA authorized test that supports treatment-free remission by accurately and precisely monitoring Molecular Response in eligible patients. M=Major Molecular Response M is defined as a molecular response of 3.0; BCR-ABL is less than or equal to 0.100% IS. C=Complete Molecular Response C is defined as a molecular response of 4.5 and is necessary for TFR. TFR= Treatment-free Remission The ability of a patient to maintain Deep Molecular Response after Tasigna 88 % ~ 49 % 98 % A vast majority of patients could be eligible for TFR* Almost half of patients maintain TFR With frequent testing almost all patients regain M With a complete molecular Frequent BCR-ABL testing 98% of patients regained M response, 88% of patients adequately monitors patients and upon re-initiation of Tasigna. are eligible for TFR. confirms appropriate molecular response status. While many tests claim to detect 4.5, the Dx BCR-ABL Test is one of the few that accurately and precisely quantifies to 4.5 Accuracy Precision Dx BCR-ABL Test was the clinical trial assay developed under IDE and validated for the Tasigna TFR clinical studies, with over 10,000 clinical data points to date Dx Test Qualifications First FDA authorized Test for TFR Kit Manufactured Under ISO 13485:2003 Rigorous One-step RT-qPCR test with calibrators and controls to 4.5 Why is testing important? Why select the Dx Test? What does TFR look like? How often should you test? Dx Test Result to Assess TFR The Tasigna label update requires an increased testing regimen: 1 st YEAR 2 nd YEAR 3 rd YEAR+ Year 1: every month Year 2: every 6 weeks Year 3 and thereafter: every 12 weeks For TFR, patients are tested more frequently with the Dx BCR-ABL Test. 3-fold increase in Year 1 2-fold increase in Year 2* * See Novartis Tasigna Package Insert

11 Q&A Chad Galderisi CHAD GALDERISI is Executive Vice President, Chief Medical Officer and a Scientific Co-founder of MolecularMD, which created the Dx BCR-ABL companion diagnostic for the Novartis drug Tasigna (nilotinib) to measure whether chronic myeloid leukemia (CML) patients with sustained deep molecular response could safely discontinue therapy and remain in treatment free remission (TFR). Here, Dr. Galderisi describes the creation of the diagnostic and how MolecularMD was awarded the contract from Novartis based on the superiority of its BCR-ABL test. CO: How did MolecularMD first start working with Novartis on nilotinib? Galderisi: In April of 2012 Novartis needed a robust, sensitive, and precise assay capable of reliably detecting at least 4.5 and also distinguishing 4 from 4.5 which is required for patients attempting TFR. The clinical trial assay was designed, developed, and validated in less than three months in order to meet the extremely ambitious Novartis timeline to start testing all patients for the consolidation phase of their TFR clinical trials. 12 A Clinical OMICs ebook

12 Dx s sensitivity, reproducibility, and limit of detection surpassed all other assays on the market and ultimately secured MolecularMD the contract for both multi-year trials. CO: What is it that makes the Dx BCR-ABL test stand apart? CG: The Reverse Transcription (RT) and PCR are done in the same well and the test uses in vitro transcribed RNA calibrators, which go through the RT step, thus mimicking patient samples. The LOD/LOQ validation was extensive incorporating 60 replicates at each clinical decision point with multiple operators, lots, and instruments. The assay enzyme system selected was extremely robust and the RNA input for the test was also optimized to ensure that each patient sample tested obtained at least a 4.5 sensitivity. We measured accuracy with droplet digital PCR to verify that our copy numbers were accurate using a completely different methodology that was almost as sensitive as Dx. The test also incorporates a 4.5 and M control on every run in order to ensure that the assay is Dx s performing appropriately within sensitivity, acceptable specifications. There reproducibility, and are several labs and kits that limit of detection claim 4.5 sensitivity, but they surpassed all other do not utilize controls that confirm the consistent measurement market... assays on the of 4.5 and they do not have FDA authorization for TFR. The Dx BCR- ABL Test stands alone as the companion diagnostic for Tasigna and the only diagnostic assay with a claim for Treatment-free Remission. 13 A Clinical OMICs ebook

13 CO: Anything else you d like to mention? CG: I think we have assembled a really great team. The Core Team for Dx is dedicated and persistent. There have been a lot of setbacks and challenges along the journey but the team persevered through it all. Overall, this is a great opportunity for CML patients to safely and reliably go off therapy without having to worry about having an appropriate test to monitor potential relapse. Patients now know when they have been treated with Tasigna for at least three years, and have sustained a level of at least 4.5, then they have an approximately 50 percent chance of successfully staying off therapy. This breakthrough is really all about patient outcomes and having this test is great for them and their hopefully successful cessation of treatment. 14 A Clinical OMICs ebook

14 Dx BCR-ABL Test Supports Tasigna Label Update In late December, in what marked a significant breakthrough for patients with Philadelphia chromosome positive Ph+ Chronic Myeloid Leukemia (CML), the U.S. Food and Drug Administration (FDA) updated the label for the Novartis cancer drug Tasigna (nilotinib) that included information for physicians about how to discontinue use of the drug in certain patients with Ph+ CML. The label update contains information on treatment-free remission (TFR) data that makes Tasigna the first and to date, only second generation tyrosine kinase inhibitor (TKI) whose patients may attempt to discontinue taking the drug based on their response to treatment. The TFR label update was supported by two separate clinical studies, ENESTfreedom and ENESTop. Minimal residual disease (D) and molecular response () data for the trials were generated with the Dx BCR-ABL Test from MolecularMD. TFR is defined as the ability of patients to sustain major molecular response (M) or deep molecular response (D) after discontinuing use of the drug. The two clinical studies evaluated patients over the age of 18 with Ph+ CML in the chronic phase and their potential to maintain D. Of these patients, 49% who discontinued Tasigna remained in TFR for two years. Of the patients who lost TFR, 98% regained D when they began taking the drug again. In the FDA s announcement of the label change, Richard Pazdur, M.D., Director of 15 A Clinical OMICs ebook

15 the FDA s Oncology Center of Excellence and acting director of the Office of Hematology and Oncology Products in the FDA s based on the validation data demonstrated by its use in the ENESTfreedom and ENESTop trials. This is a very significant achievement, said Chad Galderisi, Executive Vice President, Chief Medical Officer and Co-founder of MolecularMD. There have been some investigator studies of TFR prior to ENESTfreedom and ENESTop, but these were the first studies that allowed the drug label to change in the U.S. for TFR. The implications of TFR are far-reaching, as Ph+ CML strikes people of all ages and, until now, those with the disease faced a lifetime of treatment. For some women with Ph+ CML, the potential to go off Tasigna opens up the possibility of starting a family. Yet for others, who may be responding to the treatment, are nonetheless afflicted with side effects which impact quality of life. For patients who achieve deep molecular response, it is important to understand they can safely discontinue treatment, Galderisi added. Center for Drug Evaluation and Research noted that while TFR marked an advancement in CML therapy, any discontinuation of treatment still means patients must be regularly monitored for disease recurrence. This is exactly the role served by MolecularMD s Dx BCR-ABL Test which was concurrently authorized by the FDA as the companion diagnostic to monitor TFR in patients who discontinue use of Tasigna. It is currently the only test authorized by FDA for monitoring D for TFR patients, Patients who discontinue use of Tasigna are monitored using the Dx BCR-ABL Test monthly in the first year, every six weeks in year two, and every 12 weeks for year three and beyond. According to Dan Snyder, CEO of MolecularMD, this marks a major advancement in CML treatment practice. The FDA authorized Dx BCR-ABL Test ensures that physicians have the information needed to identify patients that meet the stringent eligibility criteria to attempt TFR and provides the robust sensitivity and accuracy necessary for monitoring minimal residual disease with confidence, Snyder concluded. 16 A Clinical OMICs ebook

16 Going Beyond M to the Analysis of Deep Molecular Response Kristy A. Drafahl, Julie A. Toplin, David L. Smith, Chad D. Galderisi, MolecularMD Corp., Portland, OR Analytical Validation-Traceability to the International Scale The NIBSC World Health Organization Laboratory provides a method for the calibration of a manufacturer s BCR-ABL secondary standards to the WHO-NIBSC BCR-ABL primary standards. This method is described in the 1st WHO International Genetic Reference Panel for quantitation of BCR-ABL translocation by RQ-PCR, Instruction for Use. This method will allow assignment of International Scale (IS) values to WHO BCR-ABL Reference Panel Secondary Standards in addition to assigning a conversion factor for each assay. Briefly, on five different days, a different WHO primary panel of four freeze-dried materials was reconstituted and extracted using the Maxwell CSC RNA Blood Purification kit on the Maxwell CSC instrument. The rehydrated/extracted WHO Primary samples (200 ng RNA input/bcr-abl or ABL well for Dx and 200 ng RNA input/bcr-abl and 4 ng RNA input/abl well for ddpcr) were tested over ten days using three lots of reagents for Dx or two lots of reagents for ddpcr. The BCR-ABL/ABL % values were calculated from the averaged or merged BCR-ABL and ABL copy numbers (normalized for RNA input) for each WHO primary sample tested. The resulting BCR-ABL/ABL % value was converted to log10. This is the measured result. The mean of the WHO-assigned log10 IS value and the measured result log10 value were calculated for each test of the WHO primary panel members. The difference between the WHO-assigned log10 IS value and the measured result log10 value was calculated for each WHO primary panel level tested. The mean difference between the WHO-assigned log10 IS value and the measured result log10 value was calculated for all levels of primary standards. A regression analysis of the mean and the differences was performed to determine if the differences between the assigned value and the measured result were uniform across all levels of WHO primary standards. The conversion factor was defined as the anti-log of the mean difference. No significant trend (p>0.05) should be observed in the regression analysis of the mean and the differences of the WHO-assigned log10 IS value and the measured result log10 value. Dx Over the course of the study, five WHO International Genetic Reference Panels were tested by four operators across four instruments using three separate lots of Dx kits over the course of twenty days. One outlier (Z=3.91 based on the entire dataset) in the dataset was detected based on the Grubbs test and was excluded from the analysis as the root cause was identified; with an N>100, the Grubbs value for outliers was Z The results for the study are shown in Table 1. The conversion factor (anti-log of the mean difference) was The bias plot comparing the mean vs the difference between measured and assigned values is shown in Figure 3. The p-value for the regression was , showing no significant bias across the levels. ddpcr Over the course of the study, five WHO International Genetic Reference Panels were tested by four operators on one instrument using two separate lots of ddpcr kits over the course of ten days. The required number of samples (n=80) met the acceptance criteria. One outlier (Z=6.09) in the dataset was detected based on the Grubbs test but was included in the analysis as no root cause for the failure could be identified. The results for the study are shown in Table 2. The conversion factor (anti-log of the mean difference) was The bias plot comparing the mean vs the difference between measured and assigned values is shown in Figure 4. The p-value for the regression was , showing no significant bias across the levels. Introduction Quantification of BCR-ABL transcript levels in blood samples is routinely used to assess CML patient response to targeted kinase inhibitor (TKI) therapy. With the approval of front and second line drugs for CML treatment, BCR-ABL measurements such as major molecular response (3) and deep molecular response (4.5) play an increasingly critical role in patient management decisions. Moreover, timing and depth of these responses have been shown to be predictive of long term outcomes. The MolecularMD Dx BCR-ABL Test is intended to be used in the monitoring of BCR-ABL mrna transcript levels in patients diagnosed with CML. Serial monitoring for BCR-ABL mrna transcript levels is used as an aid in identifying CML patients being treated with nilotinib as candidates for initiating and monitoring of treatment-free remission. The Dx BCR-ABL Test is a reproducible and highly sensitive IS-standardized solution to measure D in CML as a BCR/ABL quantitation assay that can reliably measure BCR-ABL/ABL down to % (4.5) on the WHO International Scale. In this study, we show the validation and comparison of two highly sensitive methods, droplet digital PCR (ddpcr) and RT-qPCR (Dx ), for the analysis of 4.5 using RNA extracted from EDTA and PAXgene Blood RNA tubes. Transcript Target Level N Nominal % IS % IS Std Dev % IS Nominal Std Dev Detection Rate Tube Type Bias TE EDTA e13a /38 (100%) EDTA e13a /38 (100%) EDTA e13a /39 (100%) EDTA e13a /38 (100%) EDTA e14a /39 (100%) EDTA e14a /38 (100%) EDTA e14a /39 (100%) EDTA e14a /38 (97%) 4.5e e13a /36 (100%) PAXgene e13a /35 (100%) PAXgene e13a /36 (92%) PAXgene e13a /36 (86%) PAXgene e14a /36 (100%) PAXgene e14a /36 (97%) PAXgene e14a /36 (89%) PAXgene e14a /34 (85%) PAXgene Slope: Intercept: R 2 : Going Beyond M to the Analysis of Deep Molecular Response Going Beyond M to the Analysis of Deep Molecular Response Going Beyond M to the Analysis of Deep Molecular Response Kristy A. Drafahl, Julie A. Toplin, David L. Smith, Chad D. Galderisi, MolecularMD Corp., Portland, OR Kristy A. Drafahl, Julie A. Toplin, David L. Smith, Chad D. Galderisi, MolecularMD Corp., Portland, OR Going Beyond M to the Analysis of Deep Molecular Response Going Beyond M to the Analysis of Deep Molecular Response Going Beyond M to the Analysis of Deep Molecular Response Kristy A. Drafahl, Julie A. Toplin, David L. Smith, Chad D. Galderisi, MolecularMD Corp., Portland, OR Kristy A. Drafahl, Julie A. Toplin, David L. Smith, Chad D. Galderisi, MolecularMD Corp., Portland, OR Analytical Validation-Traceability to the International Scale The NIBSC World Health Organization Laboratory provides a method for the calibration of a manufacturer s BCR-ABL secondary standards to the WHO-NIBSC BCR-ABL primary standards. This method is described in the 1st WHO International Genetic Reference Panel for quantitation of BCR-ABL translocation by RQ -ABL translocation by RQ - -PCR, Instruction for Use. This method will allow assignment of International Scale (IS) values to WHO BCR-ABL Reference Panel Secondary Standards in addition to assigning a conversion factor for each assay. Briefly, on five different days, a different WHO primary panel of four freeze-dried materials was reconstituted and extracted using the Maxwell CSC RNA Blood Purification kit on the Maxwell CSC instrument. The rehydrated/extracted WHO Primary samples (200 ng RNA input/bcr-abl or ABL well for Dx and 200 ng RNA input/bcr-abl and 4 ng RNA input/abl well for ddpcr) were tested over ten days using three lots of reagents for Dx or two lots of reagents for ddpcr. The BCR-ABL/ABL % values were calculated from the averaged or merged BCR-ABL and ABL copy numbers (normalized for RNA input) for each WHO primary sample tested. The resulting BCR-ABL/ABL % value was converted to log10. This is the measured result. The mean of the WHO-assigned log10 IS value and the measured result log10 value were calculated for each test of the WHO primary panel members. The difference between the WHO-assigned log10 IS value and the measured result log10 value was calculated for each WHO primary panel level tested. The mean difference between the WHO-assigned log10 IS value and the measured result log10 value was calculated for all levels of primary standards. A regression analysis of the mean and the differences was performed to determine if the differences between the assigned value and the measured result were uniform across all levels of WHO primary standards. The conversion factor was defined as the anti-log of the mean difference. No significant trend (p>0.05) should be observed in the regression analysis of the mean and the differences of the WHO-assigned log10 IS value and the measured result log10 value. Dx Over the course of the study, five WHO International Genetic Reference Panels were tested by four operators across four instruments using three separate lots of Dx kits over the course of twenty days. One outlier (Z=3.91 based on the entire dataset) in the dataset was detected based on the Grubbs test and was excluded from the analysis as the root cause was identified; with an N>100, the Grubbs value for outliers was Z The results for the study are shown in Table 1. The conversion factor (anti-log of the mean difference) was The bias plot comparing the mean vs the difference between measured and assigned values is shown in Figure 3. The p-value for the regression was , showing no significant bias across the levels. ddpcr Over the course of the study, five WHO International Genetic Reference Panels were tested by four operators on one instrument using two separate lots of ddpcr kits over the course of ten days. The required number of samples (n=80) met the acceptance criteria. One outlier (Z=6.09) in the dataset was detected based on the Grubbs test but was included in the analysis as no root cause for the failure could be identified. The results for the study are shown in Table 2. The conversion factor (anti-log of the mean difference) was The bias plot comparing the mean vs the difference between measured and assigned values is shown in Figure 4. The p-value for the regression was , showing no significant bias across the levels. Introduction Quantification of BCR-ABL transcript levels in blood samples is routinely used to assess CML patient response to targeted kinase ABL transcript levels in blood samples is routinely used to assess CML patient response to targeted kinase inhibitor (TKI) therapy. With the approval of front and second line drugs for CML treatment, BCR-ABL measurements such as major molecular response (3) and deep molecular re ABL measurements such as major molecular response (3) and deep molecular response (4.5) play an increasingly critical role in patient management decisions. Moreover, timing and depth of these responses have been shown to be predictive critical role in patient management decisions. Moreover, timing and depth of these responses have been shown to be predictive of long term outcomes. The MolecularMD Dx BCR-ABL Test is intended to be used in the monitoring of BCR-ABL mrna transcript levels in patients diagnosed with CML. Ser ABL mrna transcript levels in patients diagnosed with CML. Serial monitoring for BCR-ABL mrna transcript levels is used as an aid in identifying CML patients being treated with nilotinib as candidates for initiating and monitoring levels is used as an aid in identifying CML patients being treated with nilotinib as candidates for initiating and monitoring of treatment-free remission. The Dx BCR-ABL Test is a reproducible and highly sensitive IS-standardized solution to measure D in CML as a BCR/ABL quantitation assay that can reli standardized solution to measure D in CML as a BCR/ABL quantitation assay that can reliably measure BCR-ABL/ABL down to % (4.5) on the WHO International Scale. In this study, we show the validation and comparison of two highly sensitive methods, droplet digital PCR (ddpcr) and RT In this study, we show the validation and comparison of two highly sensitive methods, droplet digital PCR (ddpcr) and RT-qPCR (Dx ), for the analysis of 4.5 using RNA extracted from EDTA and PAXgene Blood RNA tubes. (Dx ), for the analysis of 4.5 using RNA extracted from EDTA and PAXgene Blood RNA tubes. Transcript Target Level N Nominal % IS % IS Std Dev % IS Nominal Std Dev Detection Rate Tube Type Bias TE EDTA e13a /38 (100%) EDTA e13a /38 (100%) EDTA e13a /39 (100%) EDTA e13a /38 (100%) EDTA e14a /39 (100%) EDTA e14a /38 (100%) EDTA e14a /39 (100%) EDTA e14a /38 (97%) 4.5e e13a /36 (100%) PAXgene e13a /35 (100%) PAXgene e13a /36 (92%) PAXgene e13a /36 (86%) PAXgene e14a /36 (100%) PAXgene e14a /36 (97%) PAXgene e14a /36 (89%) PAXgene e14a /34 (85%) PAXgene Slope: Intercept: R 2 : Analytical Validation-Limit of Blank Dx Based on reported data from multiple clinical studies, low level expression of the BCR-ABL transcript is expected in healthy individuals and therefore cannot be used to establish the limit of blank (Biernaux et al). RNA extracted with TRIzol reagent from thirty known BCR-ABL negative cell lines was used to establish the limit of blank. RNA samples extracted from thirty cell lines using TRIzol reagent were tested in duplicate using 200ng of RNA input on each of three lots for a total of 180 replicates. This input decreased the ABL1 copies in the cell lines to patient levels (100, ,000 copies of ABL1). The number of samples with detectable BCR-ABL was tabulated and the percent of BCR- ABL positive samples was calculated. The limit of blank was calculated using the nonparametric option per CLSI EP17-A2. The 95 th percentile of the distribution of results and rank position corresponding to the 95 th percentile were calculated for each lot (Rank Position= *0.95=57.5). The LoB is interpolated from the 57 th and 58 th values for each lot, all of which were zero (undetected). The LoB is therefore set to zero in %IS. ddpcr TRIzol-extracted RNA from 20 known BCR-ABL negative cell lines was used to establish LoB. RNA samples were evaluated with two (2) lots of ddpcr Test reagents (total n=40) using 200 ng RNA input for BCR-ABL and 4 ng RNA input for ABL. This input decreased the ABL1 copies in the cell lines to patient levels (100, ,000 copies of ABL1). The number of samples with detectable BCR-ABL was tabulated and the percent of BCR-ABL positive samples was calculated. The limit of blank was calculated using the nonparametric option per CLSI EP17-A2. The 95 th percentile of the distribution of results and rank position corresponding to the 95 th percentile were calculated for each lot (Rank Position= *0.95=19.475). The LoB is interpolated from the 19 th and 20 th values for each lot, all of which were zero (undetected). The LoB is therefore set to zero in %IS. The 1D plots from one lot are shown in Figures 7a ( BCR-ABL) and 7b (ABL). Analytical Validation-Precision and Reproducibility Dx PAXgene (precision study) Two series of samples were created by separately diluting blood drawn into PAXgene Blood RNA tubes from pooled e13a2 and pooled e14a2 positive CML patient samples with high BCR-ABL/ABL ratios into blood from non-diseased subjects drawn into PAXgene Blood RNA tubes. Five levels were created with targeted BCR-ABL/ABL % IS ratios from 10% IS to % IS. One testing site was used with three operators; each operator used exclusively one 7500 Fast DX instrument. Two sample panels were extracted with three sets of reagents and tested using three lots of reagents over five non-consecutive days for each lot for a total of 15 non-consecutive days over the course of approximately one month. A total of 90 measurements were made for each panel member for each BCR-ABL transcript type. The total number of tests run across transcript types, RNA inputs, and panel members is 900 tests. The difference in replicates per level (N) is due to either failed plates or invalid samples. The results of this study are presented in Table 7. The total precision expressed in SD of the value for each sample across lots, instruments, operators, and days must be less than All levels passed the acceptance criteria for the study. EDTA (reproducibility study) Two series of samples were created by separately diluting blood from EDTA blood collection tubes from pooled e13a2 and pooled e14a2 positive CML patient sample with high BCR-ABL/ABL ratios into blood from non-diseased subject blood drawn into EDTA blood collection tubes. Five levels were created with targeted BCR-ABL/ABL % IS ratios from 10% IS to % IS. Three testing sites with two 7500 Dx instruments were used with two operators at each site. Each day at each site, each operator ran two plates, one on each instrument. Each plate had one replicate of the e13a2 five member panel and one replicate of the e14a2 five member panel. For each of three lots, five runs were done. A total of 15 measurements were made per instrument, per operator. For each site, a total of 60 replicates were run (15 measurements per instrument per site x 2 instruments x 2 operators=60) across each BCR-ABL transcript type. The total number of tests run across transcript types and panel members was 1800 tests. The difference in replicates per level (N) is due to either failed plates or invalid samples. The results of this study are presented in Table 8. The total variability expressed in SD of the value for each sample across sites, lots, instruments, operators, and days must be less than The total variability expressed in standard deviation passed the acceptance criteria (SD < 0.25) for both BCR-ABL transcripts across all five levels tested. ddpcr-intermediate Precision Two (2) sample panels containing either PAXgene or EDTA blood-derived RNA and representing both BCR-ABL transcripts (i.e., e13a2 and e14a2) were formulated. The panels included five (5) different % IS targeted concentrations of each transcript and were tested on three (3) non-consecutive days using two (2) different lots of BCR-ABL ddpcr Test kits. All three (3) operators tested both lots of reagents on a single day. A total of 18 measurements were made per level for a total of 90 measurements for each BCR-ABL transcript and blood tube type. The total precision of the BCR-ABL ddpcr Test for all variables, including lot, operator, and day was less than 0.25 SD log10 for all samples tested that were above the LoQ of the assay. The results of this study are presented in Table 9. were above the LoQ of the assay. Methods Both assays are designed to detect BCR-ABL (e13a2/b2a2 or e14a2/b3a2) and ABL transcript levels. Total RNA extracted from blood is used as the template in one-step thermal cycling protocols with BCR-ABL and ABL target-specific primers in separate wells. The Bio-Rad QX200 and ABI 7500 Fast Dx systems are used for the ddpcr and Dx assays, respectively. International Scale (IS) conversion factors based on the 1st WHO International Genetic Reference Panel were established for both assays. Limit of blank (LoB), limit of detection (LoD), precision, linear range, and specificity studies were conducted using diluted CML patient samples or cell lines. A method comparison of the two assays was performed using 100 CML patient samples for both blood collection tube types. PAXgene samples are extracted from 4 PAXgene blood RNA tubes using the Qiagen Blood RNA Extraction Kit with a modified protocol which allows isolation of two tubes on one column to increase the concentration of RNA. The EDTA samples are extracted on the Promega Maxwell CSC instrument using the Promega Blood RNA kit using a modified protocol that allows isolation of RNA from 5mL of blood rather than 2.5mL in order to increase the RNA concentration. Dx The MolecularMD Dx BCR-ABL Test is a quantitative real-time polymerase chain reaction test that provides sensitive and specific quantitation of BCR-ABL (BCR-ABL1, hereafter BCR-ABL) transcripts (e13a2/b2a2 or e14a2/b3a2) and ABL (ABL1, hereafter ABL) transcript levels in RNA extracted from peripheral blood samples collected from CML patients. Total RNA is extracted from peripheral blood and serves as the template for RT-qPCR. The test is performed using a one-step RT-qPCR protocol wherein the reverse transcription and quantitative real-time PCR reactions are performed in the same well. BCR-ABL and ABL amplicons are generated and detected in real-time using TaqMan MGB probes. Quantitation is achieved using RNA calibration standards and linear regression analysis provided by the 510(k) cleared ABI 7500 Fast Dx PCR Instrument Software. BCR- ABL transcript levels are measured in relation to the ABL transcript as an endogenous reference. The BCR-ABL/ABL ratio is calculated and converted to the International Scale by the Dx BCR-ABL Test Software. Calibrators are analyzed for both BCR-ABL and ABL in every run simultaneously with patient samples. RNA controls provided at M and 4.5 allow for on-plate verification of test accuracy and reproducibility over this critical transcript range. The integrated conversion factor provides test results on the International Scale harmonized to the WHO. RNA input was validated at 1µg for EDTA samples and 2µg for PAXgene samples. Samples and controls are run in duplicate for both BCR- ABL and ABL. A schematic of the Dx workflow is shown in Figure 1. ddpcr The BCR-ABL ddpcr Assay is designed to run on the BioRad QX200 system and is used as both a reference method for validation of the Dx BCR-ABL Test as well as an in-process QC assay used for manufacture of the Dx RNA calibrators. Samples and controls are run in duplicate wells with results calculated from merged value of the duplicate wells. BCR-ABL RNA input was validated at 1µg per well for EDTA samples and 2µg per well for PAXgene samples. In order to achieve sufficient negative droplet counts, the RNA input for each ABL well was validated at 20ng for EDTA samples and 40ng for PAXgene samples. The test is performed using a one-step, end-point PCR protocol wherein the reverse transcription and end-point PCR reactions are performed in the same well. BCR-ABL and ABL amplicons are generated and detected using TaqMan MGB probes. Absolute quantitation is achieved by collecting positive and negative fluorescence data from the sample droplets and data fitting to a Poisson distribution. BCR-ABL transcript levels are measured in relation to the ABL transcript as an endogenous reference. The BCR-ABL/ABL ratio is calculated and converted to the International Scale. RNA controls provided at 1, M, and 4.5 allow on-plate verification of test accuracy and reproducibility over this critical transcript range. The integrated conversion factor provides test results on the International Scale. A schematic of the ddpcr workflow is shown in Figure 2. Figure 2: ddpcr workflow Poster H51 Poster H51 Sample Type Molecular Response BCR-AB/ABL % IS Tube Type Transcript Level SD CV EDTA e13a2 Level Level Level Level Level Level Level Level Level Level e14a2 Level Level Level Level Level Level Level Level Level PAXgene e13a2 Level Level Level Level Level Level Level Level Level * Level e14a2 Level Level Level Level Level Level Level Level Slope: Intercept: R 2 : Analytical Validation-Linear Range Dx Two 10-level series of PAXgene and EDTA samples were created with targeted BCR-ABL/ABL % IS ratios from 15% IS to % IS. A high sample was created by pooling blood drawn into PAXgene Blood RNA tubes or EDTA tubes collected from e14a2 positive CML patient samples with high BCR-ABL/ABL ratios and diluting this pool into blood from nondiseased subjects (NDS). For the e13a2 branch of the study, an artificial sample was created using the BV173 cell line (e13a2 BCR-ABL positive cell line) spiked into non-diseased blood. The mean and standard deviation for the value for each level as well as the mean and %CV for the BCR-ABL/ABL %IS are presented in Table 3. Regression analysis for each tube type and transcript is shown in Figures 4a-d. The assay was found to be linear from 0.99 to 4.6 for EDTA and from 0.93 to 4.5 for PAXgene samples. The upper linear range for Dx was truncated at 1.0 for both tube types. Table 3: Dx linear range ddpcr Two 10-level series of dilutions were made by diluting a pool of a high CML patient sample collected in either PAXgene blood RNA tubes or EDTA tubes and diluted into nondiseased donor blood collected in the same blood tube type. Five (5) replicates were tested from each of the two (2) RNA panels. Only the e14a2 transcript was evaluated in this study. The mean and standard deviation for the value for each level as well as the mean and %CV for the BCR-ABL/ABL %IS is presented in Table 4. Regression analysis for each tube type and transcript is shown in Figures 5a-b. The assay was found to be linear from 0.8 to 5.0 for EDTA samples and from 0.9 to 4.9 for PAXgene samples. Analytical Validation-Accuracy Blood from CML patients ranging from 10% BCR-ABL/ABL IS (1) to 0.001% BCR-ABL/ABL IS (5) was drawn into four PAXgene Blood RNA collection tubes and two EDTA blood collection tubes. All specimens were acquired under 21 CFR 50.25(c) informed consent and applicable GCP regulations. Approximately 100 samples between BCR-ABL/ABL IS 1-10% and % were analyzed with the Dx and ddpcr assays. Deming regressions for EDTA samples and PAXgene samples are presented in Figures 6a and 6b respectively. Summary statistics for the regression analyses and predicted differences at each clinical decision point is shown in Table 5. The results of the study demonstrate that no clinically significant bias was seen in the Molecular Response value between the two assays, regardless of the blood tube type. Slope: Intercept: R 2 : Slope: Intercept: R 2 : Slope: Intercept: R 2 : Sample Type Molecular Response BCR-AB/ABL % IS Tube Type Transcript Level SD CV EDTA e14a2 Level Level Level Level Level Level Level Level Level Level PAXgene e14a2 Level Level Level Level Level Level Level Level Level Level Slope: Intercept: R 2 : Tube Type Sample Size Deming Slope Deming Y-Intercept Pearson Coefficient Predicted Difference 3 Predicted Difference 4 Predicted Difference 4.5 EDTA [0.94, 0.99] [0.0027, 0.17] [-0.053, ] [-0.10, ] [-0.13, ] PAXgene [0.95, 1.0] [-0.11, 0.056] [-0.11, ] [-0.15, ] [-0.17, ] Tube Type Transcript Level % IS SD % IS Nominal Observed Bias SD Total Error N EDTA e13a2 Level EDTA e13a2 Level EDTA e13a2 Level EDTA e13a2 Level EDTA e13a2 Level EDTA e14a2 Level EDTA e14a2 Level EDTA e14a2 Level EDTA e14a2 Level EDTA e14a2 Level PAXgene e13a2 Level PAXgene e13a2 Level PAXgene e13a2 Level PAXgene e13a2 Level PAXgene e13a2 Level PAXgene e14a2 Level PAXgene e14a2 Level PAXgene e14a2 Level PAXgene e14a2 Level PAXgene e14a2 Level Tube Type Transcript Kit Lot Pooled SD (SDL) L (total sample replicates) J (total patient samples) Cp LoD (SqRt) LoD (%IS) LoD () EDTA e13a e14a PAXgene e13a e14a difference (Md) Standard Deviation (Sd) N 119 Grubbs' critical value for N 3.38 Number of samples over critical value 1 (Z=3.91) Conversion Factor 1.10 p= difference (Md) Standard Deviation (Sd) N 80 Grubbs' critical value for N 3.31 Number of samples over critical value 1 Conversion Factor 0.86 p= Transcript Level N Value Between Operator Between Day Between Lot Total SD %CV SD %CV SD %CV SD %CV e13a e13a e13a e13a e13a e14a e14a e14a e14a e14a BCR-ABL Transcript Level Between Site Between Operator Between Day Between Instruments Between Lot Total N SD %CV SD %CV SD %CV SD %CV SD %CV %CV SD e13a2 Level e13a2 Level e13a2 Level e13a2 Level e13a2 Level e14a2 Level e14a2 Level e14a2 Level e14a2 Level e14a2 Level Transcript Level n EDTA PAXgene Avg Between Op Between Day Between Lot Total Avg Between Op Between Day Between Lot Total SD %CV SD %CV SD %CV SD %CV SD %CV SD %CV SD %CV SD %CV e13a e13a e13a e13a e13a e14a e14a e14a e14a e14a Figure 4a: Dx linear regression for EDTA e13a2 Figure 4b: Dx linear regression for EDTA e14a2 Figure 4c: Dx linear regression for PAXgene e13a2 Figure 4d: Dx linear regression for PAXgene e14a2 Table 4: ddpcr Linear Range Figure 5a: ddpcr Linear Regression EDTA e14a2 Figure 5b: ddpcr Linear Regression PAXgene e14a2 Figure 7a: ddpcr LoB (BCR-ABL) Figure 7b: ddpcr LoB (ABL) Figure 6a: Deming Regression for EDTA samples Figure 6b: Deming Regression for PAXgene samples Table 5: Summary statistics for Accuracy study Table 7: Dx PAXgene precision study Table 8: Dx EDTA Reproducibility study Table 9: Intermediate Precision study Figure 8a: Pooled SD Figure 8b: Parametric LoD Table 10: Dx LoD Table 11: Dx LoQ Table 12: ddpcr LoD and LoQ Figure 9a: BCR-ABL ddpcr Figure 9b: ABL ddpcr Dx ddpcr LoD EDTA tubes: 5.5 PAXgene tubes: 5.4 EDTA tubes: 5.0 PAXgene tubes: 4.5 LoQ EDTA tubes: 4.8 PAXgene tubes: 4.6 EDTA tubes: 5.1 PAXgene tubes: 4.5 Linear Range EDTA: PAXgene: EDTA: PAXgene: Assay Range (based on on LoQ and linearity) EDTA: PAXgene: EDTA: PAXgene: Table 13: Dx and ddpcr assay specifications Figure 3: Bias plot for Dx (mean vs Difference) Table 1: Summary statistics for Dx WHO study Figure 4: Bias plot for ddpcr (mean vs Difference) Table 2: Summary statistics for ddpcr WHO study Figure 1: Dx workflow *Nominal value was 4.5 and used to set the linear range Analytical Validation-Limit of Detection and Limit of Quantitation Dx-LoD Three CML patient RNA samples for each BCR-ABL transcript were diluted into non-diseased subject blood to create five member panels. The RNA panels were tested using three lots of the Dx BCR-ABL Test kit (n=20 per lot for a total of n=60 replicates per level and per sample). The mean BCR-ABL/ABL % IS values and values were calculated for each level for all blood tube types and BCR-ABL transcripts tested. The level nearest the expected LoD was used for analysis. The data was analyzed to determine if the BCR-ABL/ABL % IS values were normally distributed (p-value > 0.05). The data was square root transformed since the non-transformed data was not normally distributed. Per CLSI EP17-A2, since the transformed data was normally distributed, the parametric approach to calculating the LoD was used. The pooled Sd was calculated using the formula shown in Figure 8a. The LoD was calculated using the formula shown in Figure 8b where L is the total number of sample replicates and J is the total number of patient samples. The value represents the 95th percentile from the normal distribution for b=0.05. Note: the abstract for this poster described a probit approach to determining the LoD. Since the time of the abstract submission, the LoD was recalculated using the parametric approach rather than probit. The LoD was determined to be 5.5 for EDTA and 5.4 for PAXgene samples (Table 10). Dx-LoQ The standard deviation of the BCR-ABL/ABL % IS value and the value was calculated for each level for all extraction methods and BCR-ABL transcripts tested. The nominal BCR-ABL/ABL % IS and values were calculated based on value assigned by the BCR-ABL ddpcr Test. The log10 bias between the nominal values and the Dx BCR-ABL Test for each level was calculated by subtracting the average measured values for each level from the values. The total error (TE) was calculated using the following equation: Total Error = Bias + 2 x SD. For the limit of quantification of the Dx BCR-ABL Test, the Total Error must be 0.5 log10. For BCR-ABL/ABL % IS log10 values less than or equal to the LoQ, this criteria ensures that there is a greater than 95% probability that the measured value will be within one half log of the true value. The LoQ for EDTA was 4.8 and for PAXgene was 4.6 (Table 11). ddpcr-lod and LoQ Two separate series of samples was created by diluting RNA extracted from blood drawn into either PAXgene Blood RNA tubes or EDTA tubes from e13a2 and e14a2 positive CML patient samples with high BCR-ABL/ABL ratios into RNA extracted from blood from non-diseased subjects drawn into the same tube type. Four levels were created for each transcript type with targeted BCR-ABL/ABL % IS ratios at 4.0, 4.5, 4.8, and 5.0 and the samples were all tested in 20 replicates using two lots (total n=40). The average BCR-ABL/ABL % IS values were calculated for all replicates for each level for both extraction methods and BCR-ABL transcripts. The BCR-ABL/ABL % IS and nominal values were calculated based on the dilution scheme from the 4.5 clinical sample. The percent hit rate was calculated based on the number of replicates with detectable BCR-ABL ratios out of the number of total number of tested replicates (n=40). The LoD was calculated as the BCR-ABL/ABL % IS nominal value where at least 95% of replicates were detected by direct point estimate. For EDTA, the LoD was 5.0 and for PAXgene, the LoD was 4.5 (Table 12). Example 1D ddpcr plots are shown in Figures 9a and 9b. The total error (TE) was calculated using the following equation: Total Error = Bias + 2 x SD. For the limit of quantification of the Dx BCR-ABL Test, the Total Error (TE) = bias +2SD 0.5 log10. For BCR-ABL/ABL % IS log10 values less than or equal to the LoQ, this criteria ensures that there is a greater than 95% probability that the measured value will be within one half log of the true value. The TE for PAXgene-derived e14a2 transcripts was greater than 0.5 at 4.6. There were no sample levels at exactly 4.5 so a linear interpolation of samples with measurements bracketing 4.5 (i.e., Level 1 at 4.0 and Level 2 at 4.6) was performed using JMP statistical software s interpolate function. The resulting TE was 0.5 and therefore the acceptance criterion (TE 0.5 at 4.5) was met for this condition. The LoQ for EDTA samples was 5.1 and for PAXgene samples was 4.5. Conclusions Both assays met acceptance criteria for all validation studies. The sequence of the BCR-ABL and ABL PCR amplicons are 100% homologous to the reference sequences. In precision studies, the SD of the value was below 0.25 in the established linear range for all samples across lot, operator and day for both assays. As expected, the level of variability was proportional to levels. No clinically significant bias at the clinical decision points was observed when the molecular response values were compared between the two assays for both tube types. A summary of the assay specifications is shown in Table 13. For Dx, the assay software truncates the assay range at 4.5 for both tube types. In this study, BCR-ABL quantification assays based on RT-qPCR (Dx ) and ddpcr methods were validated for detecting 4.5 using RNA extracted from EDTA or PAXgene Blood RNA tubes. These assays provide important diagnostic tools for monitoring deep therapeutic responses in CML patients treated with TKI therapies. Introduction Discontinuation of tyrosine kinase inhibitor therapy for treatment-free remission (TFR) in patients with chronic myelogenous leukemia in chronic phase (CML-CP) and a sustained deep molecular response is an emerging treatment goal 1-6 ENESTfreedom is the first study to investigate TFR in patients who achieved a sustained deep molecular response with frontline nilotinib 2 In the primary analysis of ENESTfreedom, the rate of TFR at 48 weeks was 51.6%, similar to rates reported in studies of TFR following imatinib despite a shorter duration of prior therapy (median of 3.6 years in ENESTfreedom vs 5-7 years in prior studies) 2-6 Methods ENESTfreedom (NCT ) is an ongoing, single-arm, phase 2 study (Figure 1) Figure 1. ENESTfreedom Study Design Adults with CML-CP b2a2 and/or b3a2 transcripts 2 years frontline nilotinib 4.5 at screening (central laboratory) Nilotinib consolidation phase (52 weeks) Nilotinib continuation phase (52 weeks) TFR-2 phase Nilotinib treatment reinitiation phase TFR phase (up to 264 weeks after last patient enters TFR phase) RQ-PCR every 4 weeks for the first 48 weeks, every 6 weeks for the second 48 weeks, and then every 12 weeks Sustained D a No sustained D a Sustained D a Enroll Loss of M Loss of M RQ-PCR every 12 weeks D, deep molecular response; IS, International Scale; M, major molecular response (BCR-ABL1 IS 0.1%); 4.5, BCR-ABL1 IS %; RQ-PCR, real-time quantitative polymerase chain reaction (standardized to the IS). a Sustained D defined as the following (in the last 4 quarterly PCR assessments): 4.5 in the last assessment, no assessment worse than 4 (BCR-ABL1 IS 0.01%), and 2 assessments between 4 and 4.5. Here we present updated results from ENESTfreedom based on a cutoff date of 31 October 2016, at which time all patients who entered the TFR phase had completed 96 weeks of TFR, entered the reinitiation phase, or discontinued from the study Building on results from the primary analysis reported previously, 2 the current report provides longer-term efficacy and safety updates Updated efficacy analyses include the rate of patients remaining in TFR (with M) at 96 weeks and evaluation of potential predictors of TFR success at 48 weeks Updated safety analyses include the characterization of safety during TFR over time for the subset of patients who remained in the TFR phase for > 48 weeks Results Of 190 patients who entered the TFR phase, 93 (48.9%) remained in TFR at the data cutoff; of the 88 patients who reinitiated nilotinib due to loss of M, 69 (78.4%) remained in the reinitiation phase at the data cutoff (Figure 2) Figure 2. Patient Flow and Disposition Still in TFR phase at data cutoff n = 93 Entered TFR phase n = 190 Discontinued TFR phase n = 97 Entered reinitiation phase due to loss of M n = 88 Still in reinitiation phase at data cutoff n = 69 Discontinued study n = 9 Patient decision: 3 Loss of M: 3 Physician decision: 1 Death: 1 Loss to follow-up: 1 Discontinued study n = 19 AEs: 7 Physician decision: 5 Patient decision: 3 Death: 3 Lack of efficacy: 1 AE, adverse event. At diagnosis, 32.6%, 26.3%, and 14.7% of patients had low, intermediate, and high Sokal risk scores, respectively; Sokal risk scores at diagnosis were not available for 26.3% of patients (Table 1) At the data cutoff date, the median duration of follow-up in the TFR phase was 75.9 weeks (range, weeks) Table 1. Patient Characteristics in the TFR Population TFR Population (n = 190) Median age at study entry (range), years 55 (21-86) Sokal risk score at diagnosis, n (%) Low 62 (32.6) Intermediate 50 (26.3) High 28 (14.7) Missing 50 (26.3) Median time from first 4.5 until TFR entry (range), months 30.4 ( ) Median total nilotinib duration prior to TFR (range), months 43.5 ( ) Treatment-Free Remission Ninety-three of 190 patients (48.9%;, 41.6%-56.3%) remained off treatment and in M at week 96 of the TFR phase Eighty-eight of 190 patients also had 4.5 Five of the 98 patients who were in TFR at 48 weeks were no longer in TFR at 96 weeks Three patients lost M after 48 weeks (at 54, 78, and 92 weeks) Two patients discontinued from the study without loss of M (due to patient decision and loss to follow-up) The estimated rate of treatment-free survival (TFS) at 96 weeks was 50.9% (, 43.6%-57.8%; Figure 3) Figure 3. Kaplan-Meier Estimated TFS a,b Treatment-Free Survival, % Time Since TFR Start, weeks No. at risk:events Patients 190 Events 94 Censored 96 Censored observation 190:0 99:89 120:70 95:91 75:93 8:93 0:94 a TFS was defined as the time from the start of TFR until the earliest of any of the following: loss of M, reinitiation of nilotinib for any reason, progression to accelerated phase/blast crisis, or death due to any cause. b By the data cutoff date, 1 patient lost M at week 120, at which time only 8 patients were considered to be at risk, resulting in the artificial drop seen at the end of the curve. Response to Retreatment Of the 88 patients who reinitiated nilotinib due to loss of M: Eighty-seven (98.9%) regained M (Figure 4) The remaining 1 patient discontinued from the study (due to patient decision) without regaining M 7.1 weeks after reinitiating nilotinib Eighty-one (92.0%) regained 4.5 by the data cutoff date (Figure 5) Of the patients who regained M but not 4.5, 1 remained on study, and 5 discontinued from the study by the data cutoff date 5 to 25 weeks after reinitiation of nilotinib (2 patients due to AEs, 1 patient due to lack of efficacy, and 2 patients due to individual decisions) Figure 4. Cumulative Incidence of M Regained After Nilotinib Reinitiation Cumulative n/n Cumulative % Time Since Start of Retreatment, weeks Patients Who Regained M, % / / / / / / Figure 5. Cumulative Incidence of 4.5 Regained After Nilotinib Reinitiation Time Since Start of Retreatment, weeks 0/ / / / / / / / / / Cumulative n/n Cumulative % Patients Who Regained 4.5, % Patients with low Sokal risk scores at diagnosis and patients with 4.5 in all assessments during the consolidation phase remained in TFR at higher rates than other patients (Table 2) Table 2. TFR Rates According to Sokal Risk and 4.5 Stability TFR Rate at 48 Weeks, n/n (%; ) TFR Population (n = 190) Sokal risk score at diagnosis Low 39/62 (62.9; ) Intermediate 25/50 (50.0; ) High 9/28 (32.1; ) BCR-ABL1 IS level in the consolidation phase % in all assessments 90/170 (52.9; ) > % in 1 assessment 8/20 (40.0; ) Safety Three new deaths were reported since the 48-week analysis; an overall total of 8 deaths were reported in the study by the data cutoff (Table 3) Table 3. Deaths Deaths, n (%) Consolidation Phase (N = 215) TFR Phase (n = 190) Reinitiation Phase (n = 88) Post- treatment Follow-Up a Total 2 (0.9) 1 (0.5) 3 (3.4) 2 Cardiac arrest 1 (0.5) Suicide 1 (0.5) Acute myocardial infarction (1.1) 0 Respiratory failure (1.1) b 0 Other cancers a,b Unknown cause 0 1 (0.5) 1 (1.1) 0 a Deaths were reported > 30 days after patients discontinued from the study. b New deaths reported since the 48-week analysis. Safety analyses included the subgroup of patients remaining in TFR for > 48 weeks (n = 100) Fewer patients in this subgroup had AEs during the second 48 weeks of TFR (62.0%) vs the first 48 weeks of TFR (76.0%) or the 1-year consolidation phase (85.0%) The frequency of cardiovascular events reported in each study period are shown in Table 4 The incidence of AEs in the musculoskeletal-pain grouping increased during the TFR phase vs the consolidation phase, but decreased over time in the TFR phase (Table 5) Table 4. Cardiovascular Events (all grades) a,b Patients, n (%) Consolidation Phase (n = 100) TFR Phase First 48 Weeks (n = 100) Second 48 Weeks (n = 100) Cardiovascular events 3 (3.0) 2 (2.0) 1 (1.0) Ischemic cerebrovascular events 1 (1.0) 1 (1.0) 0 Ischemic heart disease 1 (1.0) 0 1 (1.0) Peripheral arterial occlusive disease 1 (1.0) 1 (1.0) 0 a Among patients who remained in TFR for > 48 weeks (n = 100). b Each listed AE group includes a predefined set of individual AEs. Reported frequencies include all patients with 1 new or worsening AE in the group reported during the indicated study period. Table 5. Musculoskeletal Pain and Other Clinically Notable AE Groups (all grades) a,b Patients, n (%) Consolidation Phase (n = 100) TFR Phase First 48 Weeks (n = 100) Second 48 Weeks (n = 100) Musculoskeletal pain 17 (17.0) 34 (34.0) 9 (9.0) Fluid retention 3 (3.0) 4 (4.0) 4 (4.0) Edema and other fluid retentions 2 (2.0) 3 (3.0) 4 (4.0) Severe 1 (1.0) 1 (1.0) 0 Hepatotoxicity 2 (2.0) 2 (2.0) 0 Cardiac failure 0 1 (1.0) 0 Rash 5 (5.0) 1 (1.0) 1 (1.0) Myelosuppression (thrombocytopenia) 1 (1.0) 0 0 Pancreatitis 1 (1.0) 0 0 Significant bleeding (1.0) Gastrointestinal hemorrhage (1.0) a Among patients who remained in TFR for > 48 weeks (n = 100). b Each listed AE group includes a predefined set of individual AEs. Reported frequencies include all patients with 1 new or worsening AE in the group reported during the indicated study period. Conclusions 48.9% of patients who had a sustained deep molecular response with frontline nilotinib therapy remained in remission at 96 weeks after stopping treatment Ninety-six-week results from ENESTfreedom were consistent with the previously reported results at the time of the primary analysis Five patients discontinued from the TFR phase between 48 and 96 weeks; only 3 of these patients discontinued TFR due to loss of M Of patients who reinitiated nilotinib due to loss of M, 98.9% regained M, and 92.0% regained 4.5 The frequency of AEs, including musculoskeletal-pain AEs, decreased during the second 48 weeks of TFR In this study, low Sokal risk at diagnosis and continuous 4.5 in the consolidation phase appeared to be associated with higher TFR rates; however, these results must be interpreted with caution due to small patient numbers in some subgroups These results support use of TFR for patients in sustained deep molecular response with frontline nilotinib References 1. Hughes TP, Ross DM. Blood. 2016;128: Hochhaus A, et al. Leukemia Mar 17. [Epub ahead of print]. 3. Etienne G, et al. J Clin Oncol. 2017;35: Ross DM, et al. Blood. 2013;122: Richter J, et al. Haematologica. 2016;101 [abstract S145]. 6. Rousselot P, et al. J Clin Oncol. 2014;32: Acknowledgments The authors thank all study participants and their families, the study investigators, and staff at participating study sites. The Dx BCR-ABL Test was developed, validated, and run by MolecularMD (Portland, OR). Financial support for medical editorial assistance was provided by Novartis Pharmaceuticals. The authors thank Karen Kaluza, PhD, of ArticulateScience LLC for medical editorial assistance with this poster presentation. Presented at the EHA Annual Meeting; Madrid, Spain; June 22-25, Durable Treatment-Free Remission Following Frontline Nilotinib in Patients With Chronic Myeloid Leukemia in Chronic Phase: ENESTfreedom 96-Week Update David M. Ross, 1 Tamas Masszi, 2 María Teresa Gómez Casares, 3 Andrzej Hellmann, 4 Jesper Stentoft, 5 Eibhlin Conneally, 6 Valentin Garcia Gutierrez, 7 Norbert Gattermann, 8 Philipp D. le Coutre, 9 Bruno Martino, 10 Susanne Saussele, 11 Francis J. Giles, 12 Jerald P. Radich, 13 Giuseppe Saglio, 14 Prashanth Gopalakrishna, 15 Weiping Deng, 16 Nancy Krunic, 17 Veronique Bedoucha, 15 Andreas Hochhaus 18 1 SA Pathology, Adelaide, Australia; 2 Semmelweis University, Budapest, Hungary; 3 Hospital Universitario de Gran Canaria Dr. Negrín, Las Palmas de Gran Canaria, Spain; 4 Medical University of Gdańsk, Gdańsk, Poland; 5 Aarhus University Hospital, Aarhus, Denmark; 6 St James s Hospital, Dublin, Ireland; 7 Hospital Universitario Ramón y Cajal, Madrid, Spain; 8 Universitätsklinikum Düsseldorf, Düsseldorf, Germany; 9 Charité - Universitätsmedizin Berlin, Berlin, Germany; 10 Azienda Ospedaliera Bianchi Melacrino Morelli, Reggio Calabria, Italy; 11 III. Med. Klinik, Medizinische Fakultät Mannheim der Universität Heidelberg, Mannheim, Germany; 12 Northwestern Medicine Developmental Therapeutics Institute, Robert H. Lurie Comprehensive Cancer Center of Northwestern University, Chicago, IL, USA; 13 Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, WA, USA; 14 University of Turin, Orbassano, Italy; 15 Novartis Pharma AG, Basel, Switzerland; 16 Novartis Pharmaceuticals Corporation, East Hanover, NJ, USA; 17 Novartis Institutes for BioMedical Research, Cambridge, MA, USA; 18 Abteilung Hämatologie/Onkologie, Universitätsklinikum Jena, Jena, Germany P601 Text Q30144 to: 8NOVA (86682) USA only North, Central and South Americas; Caribbean; China UK, Europe & Russia Sweden, Europe Scan this QR code Copies of this poster obtained through QR (Quick Response) code are for personal use only and may not be reproduced without written permission of the authors. Visit the web at: 17 A Clinical OMICs ebook Conquering CML The Breakthrough Paradigm Change of Treatment-free Remission POSTERS Going Beyond M to the Analysis of Deep Molecular Response Durable Treatment-Free Remission Following Frontline Nilotinib in Patients With Chronic Myeloid Leukemia in Chronic Phase: ENESTfreedom 96-Week Update Durable Treatment-Free Remission After Stopping Second-Line Nilotinib in Patients With Chronic Myeloid Leukemia in Chronic Phase: ENESTop 96-Week Update Click for Full Screen N 80 Grubbs' critical value for N 3.31 Number of samples over critical value 1 Conversion Factor 0.86 p= data cutoff n = 69 AE, adverse event.