Background and Aims

The molecular hallmark of ALL is characterized by recurrent, prognostic genetic alterations, many of which are cryptic by conventional cytogenetics. RNA-seq is a powerful next-generation sequencing technology with the ability to simultaneously identify cryptic gene rearrangements, sequence mutations and gene expression profiles in a single assay. We examined the feasibility and utility of incorporating RNA-seq into a prospective multicenter phase 3 clinical trial for children with newly-diagnosed ALL.

Methods

Patients enrolled on the DFCI ALL Consortium Protocol 16-001, who consented to optional studies and had available material for testing underwent RNA-seq.

Results

RNA-seq was performed in 173 patients with ALL (B-ALL: n=136; T-ALL: n=37). Median time for library preparation, sequencing, bioinformatic analysis and data interpretation was 17 [7-51] days. Median time from sample receipt to final report delivery was 35 [15-63] days. RNA-seq identified at least one alteration in 157 patients (91%) and fusion detection was 98% concordant with results obtained from conventional cytogenetic analyses for 46/47 (98%) gene fusions. An additional 49 gene fusions were identified by RNA-seq, many of which confer prognostic or therapeutic significance. Gene expression profiling and fusion detection enabled further molecular classification into the following B-ALL subgroups: ETV6-RUNX1/ETV6-RUNX1-like (n=31; 23%), high hyperdiploid (n=34; 25%), TCF3-PBX1 (n=7; 5%), KMT2A-rearranged (n=5; 4%), Philadelphia (Ph) chromosome-positive/Ph-like (n=19; 14%), DUX4-rearranged (n=11; 8%), PAX5-altered (n=7, 5%), ZNF384-rearranged (n=4, 3%), MEF2D-rearranged (n=1; 1%) and NUTM1-rearranged (n=1; 1%). RNA-seq identified 141 nonsynonymous mutations in 93 patients (54%); the most frequent were RAS-MAPK pathway mutations (36%). Among 68 patients with both low-density array and RNA-seq data for the Ph-like gene signature prediction, results were concordant in 63 patients (93%).

Conclusions

RNA-seq is feasible in the context of a multicenter prospective clinical trial. We identified several clinically-relevant genetic alterations not detected by conventional methods, supporting the integration of this technology in frontline trials for pediatric ALL.

Background and Aims

Philadelphia chromosome-like acute lymphoblastic leukemia (Ph-like ALL) is associated with high rates of chemoresistance and relapse. CRLF2 (cytokine receptor-like factor 2) rearrangements occur in 50% of Ph-like and 60% of Down Syndrome (DS)-ALL and induce constitutive kinase signaling. Current clinical trials are studying chemotherapy with the JAK inhibitor ruxolitinib in patients with CRLF2-rearranged Ph-like ALL, but results are not yet known. While chimeric antigen receptor T-cell (CART) immunotherapies have induced remarkable remissions in children with relapsed/refractory B-ALL, approximately 50% of CD19CART-treated patients will relapse again, many with CD19 antigen loss. New therapies are needed to prevent relapse and overcome immunotherapeutic resistance.

Methods

We previously developed thymic stromal lymphopoietin receptor (TSLPR; encoded by CRLF2) CAR T-cells and demonstrated potent preclinical activity in Ph-like ALL models (Qin Blood 2015), which led to a soon-to-open phase 1 clinical trial. In the current studies, we hypothesized that combinatorial targeting with bispecific TSLPRxCD19CART or TSLPRxCD22CART or with TSLPRCART+ruxolitinib will have superior activity against CRLF2-rearranged Ph-like and DS-ALL.

Results

TSLPRCART treatment of CRLF2-rearranged ALL cell line and patient-derived xenograft (PDX) models potently inhibited leukemia proliferation in vitro and in vivo and induced long-term ‘cure’ of xenograft mice. However, co-administration of TSLPRCART+ruxolitinib markedly diminished in vivo T cell proliferation, blunted cytokine production, and facilitated leukemia relapse, which could be abrogated by delaying ruxolitinib. Importantly, ruxolitinib co-treatment prevented severe TSLPRCART-induced cytokine release syndrome (CRS) and animal death. Interestingly, ruxolitinib withdrawal led to return of T-cell functionality with redetection of TSLPRCART in peripheral blood, IFN-γ production, and leukemia clearance upon CRLF2+ ALL rechallenge.

Conclusions

TSLPRCART has potent preclinical activity against both CRLF2-rearranged Ph-like and DS-ALL. Ruxolitinib co-administration impaired in vivo TSLPRCART functionality, but was protective against life-threatening CRS. Ongoing studies are defining optimal TSLPRCART+ruxolitinib sequence(s) and assessing in vivo efficacy of bispecific TSLPRCARTs in CRLF2-R Ph-like and DS-ALL PDX models for future clinical translation.

Background and Aims

With no opportunities for prevention but available highly cost-effective treatment opportunities, pediatric cancer survival fundamentally depends on accurate diagnosis. Comprehensive modern diagnostics are unavailable for many low- and middle-income countries (LMIC). We sought to develop a low-cost leukemia lineage and genomic classification approach using Oxford Nanopore Technology RNA sequencing.

Methods

We performed Nanopore RNA sequencing on 124 acute leukemia specimens across well-characterized lineage and genomic subtypes. Using only gene expression profiling, we trained a set of pairwise and one-vs-all partial least squares discriminant analysis classifiers among lineage and genomic subtypes. The output vectors from these were combined into a feature matrix for a support vector machine classifier. The composite model provides a set of predicted probabilities for each lineage and genomic subtype.

Results

Samples averaged 756,119 reads (465 Mbp) over an average of 11,755 unique genes per sample with a median read length of 840 bp. Overall, ninety seven percent (121/124) of cases were classified correctly into leukemia lineage (AML, B-ALL, T-ALL). The 117 cases with predicted probability above 80% were classified with 100% accuracy. Within the B-ALL subgroups of ETV6-RUNX1, hyperdiploid, TCF3-PBX1, KMT2A rearranged, and Ph+/Ph-like, 47/50 (94%) were correctly classified. Examining only the 40 cases of these core B-ALL subgroups with a predicted probability greater than 50%, all were classified correctly. The 18/23 (78%) of AML cases with high predicted probability for genomic subtypes were all correctly classified into core subgroups of KMT2A rearranged, core-binding factor, or “other”.

Conclusions

Our research lays the groundwork for an innovative diagnostic approach for acute leukemia in LMIC through unbiased RNA sequencing using Nanopore at material costs of $100-300 USD per sample and sequencer costs of $1000 USD. We expect genomic classification to improve with increasing samples for training the classifier. Next steps include expanding the genomic classification and developing the assay within the LMIC context.

Background and Aims

Molecular analyses in hematological malignancies may provide information about the etiological factors in leukemogenesis. Since molecular analyses are not yet available in Iraq; the country of repeated wars, we conceived of performing next-generation sequencing (NGS), to disclose the genomic landscape of Iraqi childhood acute lymphoblastic leukemia (ALL) and acute myelogenous leukemia (AML).

Methods

From June 2016 to February 2020, bone marrow paired samples (at diagnosis and remission) were collected from 66 Iraqi children aged < 16 years, including 55 ALL, and 11 AML. Flinders Technology Associates (FTA) cards were used to transfer the samples from Iraq to Japan. We performed whole-exome sequencing, targeted sequencing-based structural variation detection, and whole-genome sequencing.

Results

Among the 55 evaluated Iraqi children with ALL, there were (49, 89.1%) precursor B-ALL (B-ALL), and (6, 10.9%) precursor T-ALL (T-ALL) cases. TCF3-PBX1 was the most recurrent fusion gene (11, 22.4%). Moreover, RAS pathway mutations including NRAS, KRAS, FLT3, PTPN11, and BRAF, were abundant (22, 44.9%) in B-ALL, and they were significantly associated with high hyperdiploidy (10/12, 83.3%). Whereas no RAS mutation was identified in the 6 cases with T-ALL. Additionally, concerned with the AML subset, 5 out of 11 cases with AML had acute promyelocytic leukemia. Similarly, RAS mutations were frequent in patients with AML (4, 36.4%).

Conclusions

Apart from disclosing the unprecedented high frequency of TCF3-PBX1, NGS results confirmed our previous finding of frequent oncogenic RAS in Iraqi childhood ALL and AML compared to other regions. Our results suggest that the biology of childhood acute leukemia in Iraq is probably different, and the environment related to war and war-aftermath in Iraq may play a role along with the geographic or ethnic background.

Acknowledgments: We appreciated the 53rd-Annual SIOP acceptance for our abstract to be published in the official congress publications.

Background and Aims

Acute lymphoblastic leukemia (ALL) is one leading cause of cancer-related death in children, and cure rates for relapsed/refractory ALL remain suboptimal, highlighting the need for novel targeted therapies. We previously uncovered that ALL blasts are vulnerable to energy/ER-stress via AMP-activated protein kinase (AMPK) activation.

Methods

In order to identify genome-wide metabolic stress regulated genes, we used RNA-Seq and compared mRNA profiles in ALL cells treated with AICAR, an AMPK activator. RT-qPCR and ChIP assays were employed to validate the RNA-Seq data in Bp-ALL, T-ALL, and other mammalian cell models. Pharmacological inhibitors/activators, siRNA/shRNA, kinase assays, and genetic models were used to investigate AICAR-induced IEGs downregulation mechanism. We confirmed the in vivo relevance of our data, using an NSG mouse model.

Results

RNA-Seq identified several immediate-early genes (IEGs) as a subset of genes downregulated by AICAR. AICAR-induced IEGs downregulation was blocked by an adenosine uptake inhibitor but not with agents blocking its conversion to ZMP and AMPK activation, indicating AICAR was responsible for IEGs reprogramming. Further investigation established this mechanism was AMPK-independent. We demonstrated that AICAR directly inhibited PKD kinase activity and identified PKD as responsible for IEGs downregulation. PKD inhibition induced DUSPs downregulation, and decreased phosphorylation and nuclear export of HDAC4/5, which lowered histone H3 acetylation and decreased NFκB(p65) recruitment to IEGs promoters. Treatment with PKD inhibitors (CRT, NaPP1) led to growth inhibition and apoptosis in ALL cells and primary patient cells. Finally, we showed that treatment with CRT, regorafenib (multikinase inhibitor, including the NFκB pathway) and the combination CRT+regorafenib led to statistically significant survival of NSG mice engrafted with ALL-Luc cells.

Conclusions

We identified a previously unreported PKD-dependent adaptive survival mechanism in response to energy/metabolic stress in ALL through regulation of HDAC-IEGs and DUSPs expression. Further investigations to confirm a potential therapeutic role for targeting PKD in ALL are warranted.

Background and Aims

Acute lymphoblastic leukemia (ALL), as the most common cancer in children, is a biologically and clinically heterogeneous disease. Pediatric B-cell ALL (B-ALL) constitutes an aggressive neoplasia in which new biomarkers and therapies are required to obtain a better prognosis for the patients. Many long ncRNAs (lncRNAs) are involved in key biological processes. Dysregulation of lncRNAs has been described in numerous human diseases, including cancer, leading to oncogenic or tumor-suppressive activities. However, few lncRNAs have been directly associated with pediatric B-ALL. Here, we stablished lncRNA/mRNA expression profile associated with B-ALL in children.

Methods

We performed a comparative study of the lncRNAs and associated protein-coding genes profiles in 42 pediatric B-ALL samples and 4 bone marrows from healthy donors by microarray analysis. To validate our results, we analysed multiple external pediatric B-ALL datasets. We studied the possible gene expression regulation mechanisms of the lncRNA/mRNA candidates. Lastly, we compared the expression profiles and survival data of patients with B-ALL against different subtypes from these cohorts.

Results

We identified a subset of lncRNAs that were differentially expressed between different pediatric B-ALL subtypes and healthy bone marrow samples. The co-expression of TCL6/TCL1B and AL133346.1/CCN2 were consistently correlated across all datasets. B-ALL samples showed a higher expression of AL133346.1/CCN2 than T-ALL samples in external datasets. Gene Ontology Analysis showed overrepresented cancer-related pathways among the differentialy expressed lncRNA/mRNAs. Survival analysis demonstrated that patients with B-ALL with “high” expression levels of lncRNA TCL6 and mRNA CCN2 showed higher overall survival (OS) than those with “low” levels.

Conclusions

The classification of patients based on differentially lncRNA/mRNA, analysed in this study, could improve the diagnosis of pediatric B-ALL. “High” TCL6 and CCN2 levels could be associated with better clinical outcome of pediatric patients with B-ALL. These potential biomarkers could allow a better stratification and prognosis of children with B-ALL.