Early Hematopoietic Stem Cell Transplantation in the Treatment of High-Risk Infant Leukemia

INTRODUCTION: Acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML) are rare diseases in children less than one and two years old respectively. Infant ALL and AML demonstrates unique biological characteristics associated with aggressive clinical features, i.e. high frequency of KMT2A gene rearrangements, poor response to standard chemotherapy, and thus worse prognosis compared to older children. These patients (pts) are at high risk (HR). Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is a potentially curative therapy for HR infant leukemia especially in 1 st CR, but proportion of these patients unable to undergo allo-HSCT due to early events, mainly relapse or treatment-related toxicity during initial therapy. We investigated the role of early allo-HSCT from different types of donors. The primary end points were overall survival (OS), transplant-relate mortality (TRM), graft-versus-host disease (GVHD) free, relapse-free survival (GRFS). Secondary end points included neutrophil engraftment and acute and chronic GVHD. Methods: We included 50 infants with HR ALL (n=10, of with MLLr+ n=6) and AML (n=40, of with MLLr+ n=14) in CR 1. Median age at allo-HSCT 12,5 months old (range from 4 m.o. till 35 m.o.). Median time from diagnosis to allo-HSCT was 9 months (range from 3 till 21). Donor of SCT was: MSD-10% (n=5), MUD-46% (n=23), Haplo-44% (n=22). Majority of patients, 86% received MAC based on busulfan 12-16mg/kg (n=43, including IV busulfan-20pts), 14% (n=7) received RIC based on melphalan 140 mg/m 2. GVHD prophylaxis consisted of PTCy±CNI and m-TOR inhibitor in 62% (n=31), seroprophylaxis±CNI in 38% (n=19). Median follow-up was 64,3 months (range 1-220). Patient data were censored at the time of death or last follow-up. Probabilities of OS and GRFS were estimated using Kaplan-Meier curves. TRM was defined as any death that occurred in the absence of disease relapse; relapse was a competing risk for this event. Results: Engraftment rate was 82%. Cumulative incidences by day 100 of TRM in Haplo were 2% vs 4% in MSD+MUD (р=0,9). Ten-year rates OS in Haplo and MSD+MUD were 77,3% (±4,7%) vs 78,6% (±9%) respectively (p=0,73). Ten-year GRFS did not differ between two study cohorts being 50% vs 57% for Haplo and MSD+MUD (p=0,76). Clinically significant aGVHD (grades 2-4) after allo-HSCT was observed in 27% (n=6) Haplo and 32% (n=9) MSD+MUD (p=0,77). A total of 12% (n = 6) patients developed severe cGVHD. Conclusions: Early allogeneic HSCT in 1 st CR from different types of donors, including haploidentical is associated with favorable outcomes in infant acute leukemias. This approach should be evaluated in further controlled clinical studies. Disclosures Kulagin: X4 Pharmaceuticals, Alexion, Apellis, Biocad: Research Funding; Novartis, Generium, Sanofi, Roche, Johnson & Johnson, Pfizer: Speakers Bureau.

MLL-TFE3: a novel and aggressive KMT2A fusion identified in infant leukemia

Key Points A novel KMT2A-rearrangement, MLL-TFE3, was identified in an infant leukemia patient. MLL-TFE3 expression produces aggressive leukemia in a mouse model.

Infant Acute Lymphoblastic Leukemia with Atypical Presentation

Infant leukemia is a very rare disease and consists of 5% of all childhood leukemias. Infant leukemia usually presents with high leukocyte counts and numerous extra-hematological features including central nervous system and skin involvement. Herein, we report a 1-month-old girl presented with high phenylalanine level in national newborn screening program for phenylketonuria; and subsequently diagnosed to have acute lymphoblastic leukemia.

KMT2A-Rearranged Infant Acute Lymphoblastic Leukemia Cells Undergo ER-Stress-Induced Apoptosis Following Exposure to Proteasome Inhibitors

Infants diagnosed with KMT2A-rearranged (KMT2Ar) acute lymphoblastic leukemia (ALL) have a poor prognosis with an event free survival of 23-44%. To identify new treatment approaches we previously performed in vitro and in vivo assays to evaluate the activity of FDA approved compounds in 15 primary KMT2Ar infant leukemia samples. Three classes of agents were found to be active in these assays: proteasome inhibitors, anthracyclines, and histone deacetylase inhibitors (HDACi). KMT2Ar infant leukemia samples were exquisitely sensitive to the proteasome inhibitor bortezomib, requiring 10-100 fold less drug to achieve 50% toxicity when compared to non-KMT2Ar childhood ALL. Bortezomib is FDA approved for multiple myeloma and laboratory studies using this model system have previously demonstrated responses to be mediated through several mechanisms including NFKB inhibition, stabilization of cell cycle regulatory proteins, and perhaps most importantly the induction of an unfolded protein response (UPR) and endoplasmic reticulum (ER)-stress-induced apoptosis. To evaluate global protein dynamics in KMT2Ar ALL cells treated with bortezomib, we performed tandem mass tag (TMT) quantitative mass spectrometry on synchronized SEM cells exposed to either 50nM of bortezomib or DMSO at 0 hours (hr), 6hr, 12hr, 16hr, and 20hr. Applying pairwise comparison for 9232 unique proteins measured over the time course compared to untreated controls, we identified 1593 proteins with a log2 fold change >1.5 in bortezomib treated cells compared to 101 proteins in the DMSO control (FDR<0.01). Several proteins associated with ER-stress-induced apoptosis including ATF4, DDIT4, ATF3, TSC22D3 (GILZ), and PMAIP1 (NOXA) were upregulated more than 3-fold between 6 and 20hr, suggesting this pathway may play a role in bortezomib induced apoptosis of KMT2Ar cells (p<0.05 and log2 fold change of +/- 0.58). To validate this finding and further understand the role of the UPR and ER-stress-induced apoptosis, we evaluated seven key mediators of this pathway by western blot following bortezomib exposure on synchronized SEM cells over a 12 hour time course including ATF4, ATF6, CHOP, PERK, GADD34, CReP, and eIF2α as well as phosphorylated PERK (p-PERK) and eIF2α (p-eIF2α). This demonstrated a critical time point at 6hr where an increase in ATF4 (3.5 fold), CHOP (1.6 fold), and CReP (2.9 fold) protein levels was accompanied by a decrease in p-PERK (0.7 fold), and p-eIF2α (0.8 fold) whereas GADD34 levels remained constant. Although full-sized ATF6 (ATF6a) protein showed a considerable increase (1.9 fold), the levels of cleaved ATF6 (ATF6f) were only slightly increased (1.2 fold) consistent with ATF4-mediated upregulation of CHOP leading to increased protein synthesis along with ATP depletion, oxidative stress, and cell death. While GADD34 has been shown to be the main phosphatase that functions in a negative feedback loop to resolve cell stress, recent data suggests that stabilization of CReP mRNA by ER stress is able to reverse eIF2α phosphorylation at later stages of UPR leading to re-expression of key UPR proteins. Further, p-eIF2α-attenuated protein synthesis, and not ATF4 mRNA translation has been shown to promote cell survival. Our data support a model whereby the UPR and ER-stress in KMT2Ar ALL cells is induced upon exposure to bortezomib leading to increased levels of ATF4 and CHOP. Attenuation of p-eIF2α by CReP further contributes to cell death through the recovery of protein synthesis in a setting of limited protein folding capacity. These results support the use of proteasome inhibitors in KMT2Ar leukemia which is currently being formally evaluated in a Phase II clinical trial for newly diagnosed patients with infant ALL (NCT02553460). Disclosures Gruber: Bristol-Myers Squibb: Consultancy.

Late Effects in Survivors of Infant Acute Lymphoblastic Leukemia from the 3 Consecutive Japanese Nationwide Clinical Trials

Background Very few studies have reported the incidence of and risk factor for late effects of infant leukemia. In addition, the long-term impact of a preparative regimen for very young children undergoing allogeneic stem cell transplantation (SCT) is not evident. Method To examine the late effects in survivors of infant acute lymphoblastic leukemia (ALL) treated on the Japanese Infant Leukemia Study Group (JILSG) trials MLL96, MLL98 and the Japanese Pediatric Leukemia/Lymphoma Study Group (JPLSG) trial MLL03, we performed a cross-sectional survey using questionnaires by the attending pediatricians from 2015 to 2017. In all 3 studies, infants with KMT2A gene rearranged (MLL-R) ALL were allocated to receive SCT in their first remission. Choice of preparative regimen was left to the physician's choice in the MLL96/98 studies, while use of non-total body irradiation (TBI) regimen with busulfan (BU) was mandatory in the MLL03 study. Results Seventy out of 102 (69%) eligible subjects were evaluated: 33 were males and 37 were females. 62 patients had an MLL-R ALL. The median age of diagnosis was 5.5 months (range, 0-12m) and 35 (50%) were <6 months of age. At the time of analysis, the median age of the 70 patients was 13.5 years (range, 7-21y) and the median follow-up duration from diagnosis was 12.5 years (range, 7-20y). Fifty-seven (81%) of the 70 patients underwent SCT. Eight patients received the TBI-based regimen, while 35 received BU regimen and one received non-TBI, non-BU regimen. Thirteen patients had SCT twice and twelve patients received both TBI and BU regimen. Relapses occurred in seventeen patients (24%). Only two patients received cranial radiation therapy in 13 patients with central nervous system involved. At least one late effect was observed in 62 of the 68 patients (91%). Identified abnormalities included: underweight (body mass index [BMI]<18.5) (69%), hypogonadism (56%), teeth mal-development (52%), short stature (52%), bone abnormalities (28%), thyroid dysfunctions (27%), skin abnormalities (24%), ophthalmic dysfunctions (24%), pulmonary dysfunctions(22%), neurocognitive dysfunctions (22%), and cardiovascular dysfunctions (14%), gastrointestinal dysfunctions (5%), renal dysfunctions (4%), second malignancy (3%), and overweight (3%). The mean standard deviation (SD) heights at last follow up with TBI regimen were -4.2 and those with multiple SCT (BU+TBI) were -3.8. Both were significantly lower than those with single SCT by BU regimen (-1.4) (TBI vs BU, P=0.014; BU+TBI vs BU, P<0.001). The mean BMI at last follow up was lower in the TBI group (16.8) compared with that in the BU group (17.8) without statistical significance (TBI vs BU, P=0.342), whereas the mean BMI in the BU+TBI group (15.1) was significantly lower than that in the BU group (BU+TBI vs BU, P=0.020). Univariable analysis revealed that the risk of both thyroid dysfunctions (odds ratio [OR]=14.0, P=0.005) and the need of growth hormone (GH) treatment (OR=10.7, P=0.011) were significantly lower in the BU group than those in the TBI group. Hypothyroidism (OR=5.8, P=0.012), teeth mal-development (OR=2.8, P=0.066) and the need of GH treatment (OR=3.5, P=0.084) were more frequently observed in children who were diagnosed at less than 6 months of age. In addition, short stature (OR=23.3, P=0.003), thyroid dysfunctions (OR=4.1, P=0.023), cataract (OR=6.9, P=0.038) and alopecia (OR=6.8, P=0.038) were more frequently observed in relapsed patients. Overall, SCT significantly increased the incidence of short stature (OR=15.5, P=0.012) and underweight (OR=5.3, P=0.018). Conclusion In this study, survivors of infant leukemia had prolonged and various types of late effects. We observed serious growth failure in height and weight in patients treated with SCT and it was further pronounced in the TBI group and in the multiple SCT (BU+TBI) group. More than half of the patients had dental late effects and a third of all the patients had thyroid dysfunctions, which were more frequent in children diagnosed earlier. The number of patients over 20 years was limited in our study, more complications may increase in additional follow-up. Continuous long-term follow up and optimal intervention are crucial for survivors of infant leukemia who are transplanted. Disclosures Kada: Bayer Yakuhin: Other: personal fees for a member of independent data monitoring committee of clinical trials, outside of the submitted work..

The efficiency of murine MLL-ENL–driven leukemia initiation changes with age and peaks during neonatal development

MLL rearrangements are translocation mutations that cause both acute lymphoblastic leukemia and acute myeloid leukemia (AML). These translocations can occur as sole clonal driver mutations in infant leukemias, suggesting that fetal or neonatal hematopoietic progenitors may be exquisitely sensitive to transformation by MLL fusion proteins. To test this possibility, we used transgenic mice to induce one translocation product, MLL-ENL, during fetal, neonatal, juvenile and adult stages of life. When MLL-ENL was induced in fetal or neonatal mice, almost all died of AML. In contrast, when MLL-ENL was induced in adult mice, most survived for &gt;1 year despite sustained transgene expression. AML initiation was most efficient when MLL-ENL was induced in neonates, and even transient suppression of MLL-ENL in neonates could prevent AML in most mice. MLL-ENL target genes were induced more efficiently in neonatal progenitors than in adult progenitors, consistent with the distinct AML initiation efficiencies. Interestingly, transplantation stress mitigated the developmental barrier to leukemogenesis. Since fetal/neonatal progenitors were highly competent to initiate MLL-ENL–driven AML, we tested whether Lin28b, a fetal master regulator, could accelerate leukemogenesis. Surprisingly, Lin28b suppressed AML initiation rather than accelerating it. This may explain why MLL rearrangements often occur before birth in human infant leukemia patients, but transformation usually does not occur until after birth, when Lin28b levels decline. Our findings show that the efficiency of MLL-ENL–driven AML initiation changes through the course of pre- and postnatal development, and developmental programs can be manipulated to impede transformation.

How I treat infant leukemia

Leukemia in infants is rare but generates tremendous interest due to its aggressive clinical presentation in a uniquely vulnerable host, its poor response to current therapies, and its fascinating biology. Increasingly, these biological insights are pointing the way toward novel therapeutic approaches. Using representative clinical case presentations, we review the key clinical, pathologic, and epidemiologic features of infant leukemia, including the high frequency of KMT2A gene rearrangements. We describe the current approach to risk-stratified treatment of infant leukemia in the major international cooperative groups. We highlight recent discoveries that elucidate the molecular biology of infant leukemia and suggest novel targeted therapeutic strategies, including modulation of aberrant epigenetic programs, inhibition of signaling pathways, and immunotherapeutics. Finally, we underscore the need for increased global collaboration to translate these discoveries into improved outcomes.