Activity of decitabine as maintenance therapy in core binding factor acute myeloid leukemia

Abstract Background Posttherapy measurable residual disease (MRD) positivity in core binding factor acute myeloid leukemia (CBF‐AML) is associated with shorter relapse‐free survival (RFS). Elimination of MRD measured via quantitative reverse transcription polymerase chain reaction (qRTPCR) for disease specific transcripts can potentially lead to better outcomes in CBF‐AML. Methods We prospectively monitored the MRD using qRTPCR and flow cytometry on bone marrow samples in patients with newly diagnosed CBF‐AML who received decitabine (DAC) maintenance therapy after fludarabine/cytarabine/G‐CSF (FLAG)‐based induction/consolidation regimen. Negative qRTPCR (CMR) was defined as fusion transcript <0.01%. Results Thirty‐one patients with CBF‐AML including 14 with t(8;21) and 17 with inv(16) received parenteral DAC as maintenance therapy. Fifteen patients (48.3%) had completed FLAG‐based induction/consolidation but with positive MRD (0.35%, range = 0.01%–0.91%) (Group 1). Sixteen patients (51.7%) could not complete recommended consolidations with FLAG‐based regimen (due to older age or complications) and were switched to DAC maintenance (Group 2). In Group 2, eight patients (50%) had undetectable MRD (Group 2A) (all had qRTPCR ≤ 0.01%) and the other eight patients (50%) had residual fusion product by qRTPCR (0.1%, range = 0.02%–0.36%) (Group 2B) prior to starting DAC. Amongst the 23 patients who had a PCR ≥ 0.01% before maintenance therapy (Groups 1 and 2B), 12 patients (52%) attained a CMR as their best response (responders). The median pre‐DAC qRTPCR amongst responders were 0.03% compared to 0.14% in nonresponders (p = .002). The median estimated molecular RFS amongst responders were 93.9 months. At a median follow‐up of 59.3 months (13.2–106 months) from DAC initiation, 16 patients (51.6%) had to be initiated on a second line of therapy (40%, 25%, and 100% patients, respectively, in Groups1, 2A, and 2B). The median estimated time to new treatment between responders was 112.4 versus 5.8 months in nonresponders (hazard ratio = 0.16, 95% confidence interval = 0.04–0.54); however, there were no difference in overall survival between these groups (p = .37). Conclusion DAC is an effective maintenance therapy for CBF‐AML patients with persistent fusion transcript at a low level after FLAG‐based regimen. Attainment of CMR with DAC maintenance can lead to long‐term remission in patients who have persistent MRD positive after FLAG‐based regimen or are unable to receive the full course of consolidation therapy.

Conclusion: DAC is an effective maintenance therapy for CBF-AML patients with persistent fusion transcript at a low level after FLAG-based regimen. Attainment of CMR with DAC maintenance can lead to long-term remission in patients who have persistent MRD positive after FLAG-based regimen or are unable to receive the full course of consolidation therapy.

| INTRODUCTION
Core binding factor (CBF) acute myeloid leukemia (AML) is a subtype of AML, characterized by the presence of t(8;21)(q22;q22) or inv (16) (p13q22)/t(16;16) recurrent translocations, leading to the formation of unique RUNX1/RUNX1T1 (AML1/ETO) or CBFB/MYH112 fusion transcripts, respectively. These cytogenetic aberrations are associated with favorable response and sensitivity to high dose cytarabine based therapy. [1][2][3][4][5] Fludarabine/cytarabine/G-CSF (FLAG)-based regimen has been shown to improve the event-free survival in CBF-AML. 6,7 Despite the high remission rates of > 80% with chemotherapy, disease relapse remains a significant cause of treatment failure with 5-year overall survival (OS) in the range of 50%-60%. [8][9][10] The presence of disease defining recurrent fusion transcripts associated with CBF-AML enable the serial monitoring by real-time quantitative reverse transcription polymerase chain reaction (qRTPCR) for detection of measurable residual disease (MRD). [11][12][13] We and others have shown before that post induction monitoring of residual disease with qRTPCR can be useful to identify the patients with higher risk of relapse. 12,[14][15][16] Given the ability to detect early relapse with high sensitivity, disease monitoring using qRTPCR can be helpful in identifying appropriate candidates for further therapy and allogenic stem cell transplantation (allo-SCT) before frank hematological relapse, as conventionally CBF-AML patients in first remission are not considered candidates for allo-SCT. 17 Studies have shown that high level of PCR persistence after induction or consolidation predisposes the patients to relapse, and 3-4 log reductions from baseline are associated with better outcomes. 11,15,[18][19][20] Gene hypermethylation has been associated with increased risk of relapses in AML. 21,22 To counter this, hypomethylating agents (HMAs), such as decitabine (DAC) and azacitidine (AZA), have been studied as maintenance agents in AML. 23,24 Earlier data published by our group, from a smaller cohort of CBF-AML had shown that HMA maintenance controlled MRD and extended remission in patients who had residual qRTPCR after FLAG-based induction/consolidation or after ASCT. 25 The data needed validation in a larger cohort with longer follow-up and questions remain on the efficacy of HMA maintenance on maintaining MRD negativity in patients who attain negative MRD status after attenuated cycles of a FLAG-based regimen, and whether such maintenance can extend remission. Here, we further explored the role of DAC maintenance therapy with serial qRTPCR and flow cytometry MRD monitoring in CBF-AML patients who completed FLAG-based regimen with persistent positive qRTPCR or those with abbreviated induction/consolidation courses.

| METHODS
We obtained samples from bone marrow (and peripheral blood in patients with long-term follow-up) for serial qRTPCR and flow cytometry approximately every 3 months in patients with CBF-AML who received at least 1 cycle of DAC maintenance for persistent MRD (persistent fusion transcript) or because of inability to complete all planned consolidations of a FLAG-based regimen (due to age, infectious complications, persistent cytopenia etc.). The sensitivity for transcript qRTPCR detection was between 1 in 10 000 and 1 in 100 000.
The methods of the qRTPCR were similar to our previously published work and in line with a Europe against Cancer program. 25,26 An increase in qPCR from < 0.01% (CMR) was considered as molecular relapse. The planned number of DAC cycles were 12, but investigator discretion was allowed based on qRTPCR status and any other evidence of disease progression or toxicity. We collected the data of baseline hemoglobin, white blood cell counts, platelet counts, bone marrow blasts and cytogenetics (CBF defining and additional cytogenetic abnormalities [ACA]) and the myeloid panel gene mutation results (discussed later).  nostic laboratory at MDACC as described previously. 28 A minimum of 250X coverage with a detection sensitivity of~5% was used for variant calling.

| Statistical analysis
Patient and clinical characteristics were summarized using descriptive statistics. OS was calculated from the date of diagnosis to the date of death due to any cause and was censored at the last follow-up date.
As DAC maintenance was aimed at offsetting relapse or need for salvage therapy and in some patients, salvage was implemented before overt relapse, time to next treatment (TTNT) was calculated from the initiation of DAC maintenance to the first salvage regimen or death and censored at last follow-up. Salvage treatment included the first therapy received after the DAC maintenance in view of molecular relapse, progression, or hematological relapse and included allo-SCT done directly after DAC maintenance. Molecular relapse-free survival (mRFS) was calculated for responders in Groups 1 and 2B from attainment of CMR to loss of CMR, hematological relapse or death (whichever was earlier) and for patients in Group 2A from DAC maintenance initiation to similar endpoints mentioned above. mRFS was censored at last follow-up. Patient characteristics are summarized using frequency (%) for categorical variables and median (range) for continuous variables. Fisher's exact test was used to assess the association between categorical variables. Kaplan-Meier method was used to estimate the probabilities of TTNT and OS. Statistical analyses were performed using GraphPad Prism version 9, GraphPad Software. Patient characteristics and treatment are summarized in Table 1 and response to DAC maintenance and subsequent therapy in Table 2. and was thus continued DAC beyond the 12 cycles. He is also the only patient in this group who needed salvage therapy after having attained CMR as his best response.

| Response to DAC maintenance
Amongst the nine patients who did not need salvage therapy, two patients received less than 12 cycles of DAC due to therapy related cytopenia while remaining in persistent CMR and hence DAC was discontinued early.
Two of the six patients who needed salvage therapy proceeded directly to allo-SCT in view of molecular persistence (never attained CMR with DAC), two received salvage therapy followed by allo-SCT

| Group 2B
There were eight patients belonging to this group, all of whom had received less than the 7 designated cycles (median = 5, range = 1-6 cycles) of FLAG-based chemotherapy, and had persistent median qRTPCR positivity at 0.1% (range = 0.02%-0.36%). The median cycles of DAC received in this group were 4 (2-13 cycles). Two patients (25%) attained a CMR, and one patient had a 2-log reduction with DAC maintenance, but all three subsequently had a molecular relapse. All the patients in this group were initiated on a second line of therapy which was also the reason for early termination of DAC maintenance (only one patient continued DAC beyond 12 cycles). Five of them were able to proceed to an allo-SCT (four after salvage chemotherapy and one directly). The median PCR prior to initiation of the second line of therapy   Negative qRTPCR was reflected in negative flow MRD, but at any higher value the correlation was poor. Thus, PCR-based MRD assessment was more sensitive than flow-based assessment in our patients.

| Survival and TTNT
We assessed TTNT as an important endpoint in our study, as DAC maintenance, by virtue of clearing MRD, should preclude or delay need for salvage therapy. We  and with short follow-up. 25 Eleven of twelve patients who maintained remission with HMA had a reduction in qRTPCR either after the first or second cycle of HMA. In that cohort, patients were treated with either DAC or AZA for persistent low qRTPCR positivity after various induction/consolidation regimens, including allo-SCT and salvage therapy. Here, we report on a larger and a more homogenous group treated upfront with only FLAG-based regimens and DAC alone as the maintenance.
The importance of completing all scheduled cycles of induction/ consolidation chemotherapy for long-term survival in CBF-AML cannot be overemphasized. 4 In the MRC cohort, CBF-AML patients who completed all FLAG-and HDAC-based induction/consolidations had stellar outcomes 37 In our cohort also on comparing TTNT between Groups 1 and 2; there was a significant difference in the OS and there was a trend toward significance in TTNT between these two groups. Benefit of allo-SCT for patient with suboptimal qRTPCR response has been reported but may not be the most suitable option for all.
The risk of delaying allo-SCT lies in the possibility of rapid overt relapses while on DAC maintenance. For patients considered to be at high risk for such an event based on qPCR cut-offs, are better off being evaluated early for SCT. Randomized trials designed to compare the benefit of predefined duration of DAC maintenance versus allo-SCT for specific postconsolidation PCR cut-offs will help to make more informed decisions and help better identification of CBF-AML patients who should be transplanted earlier as they are poised to benefit less from DAC maintenance. The importance of completing all courses of chemotherapy in conjunction with PCR MRD transcript levels also need to be studied to understand the importance of each in long-term survival.

| CONCLUSION
CBF-AML patients who are unable to complete all planned consolidation therapy or have persistent disease specific transcripts detectable at low levels via qRTPCR benefit from DAC maintenance in terms of long mRFS and salvage treatment free remission or death. Larger studies will be required to designate the subset of patients who have the maximal chance of benefit from this approach.