Poster Presentation 37th Lorne Cancer Conference 2025

Establishing a precision medicine pipeline for paediatric acute myeloid leukaemia. (#128)

Patrick Connerty 1 , Angela Jinhan Xie 1 , Fatima El-Najjar 1 , Emma Henry 1 , Jie Mao 1 , Roxie Cadiz 1 , Gabor Tax 1 , Emmy M. Dolman 1 , Nisitha Jayatilleke 1 , Chelsea Mayoh 1 , Andrew S Moore 2 , John Colgan 1 3 , Glenn Marshall 1 3 , Richard Lock 1
  1. Children's Cancer Institute, Lowy Cancer Research Centre, School of Clinical Medicine, UNSW Medicine & Health, UNSW Centre for Childhood Cancer Research, UNSW Sydney, Sydney, NSW, Australia
  2. The University of Queensland Diamantina Institute & Children's Health, Brisbane, QLD, Aus
  3. Kids Cancer Centre, Sydney Children’s Hospital, Randwick, NSW, Aus

Background:

Precision medicine approaches have been effective in the treatment of adult acute myeloid leukaemia (AML), however, their role in paediatric AML treatment remains unknown. Our group is developing a precision medicine pipeline utilising a comprehensive panel of paediatric AML patient-derived xenografts (PDXs) combined with high-throughput drug screening (HTS) and molecular profiling to identify novel treatment approaches for paediatric AML.

 

Methods:

Primary paediatric AML cells from cryopreserved patient samples were transplanted into MISTRG mice to develop PDXs. Primary cells and PDXs were molecularly characterised for gene expression, structural variations (SVs) and single nucleotide variants (SNVs) by RNA-seq and DNA-seq. Variant allele frequency (VAF) was assessed between primary samples and PDXs. PDXs and primary AML samples were subjected to HTS with a 150 anti-cancer drug library and ‘hits’ were identified using standard methodology (Z-score, IC50 and AUC). MISTRG mice were inoculated with AML PDXs and treated with drugs identified from HTS screening and in vivo response was compared to ex vivo response.

 

Results:

To date, we have established 40 paediatric AML PDXs encompassing a range of cytogenetic subtypes and risk categories. DNA-seq analysis demonstrated >90% VAF concordance (<1.2-fold change) between primary samples and PDXs. Molecular characterisation of each PDX identified SVs, SNVs or mRNA expression predictive of specific drug sensitivity. A comparison of HTS between 3 primary samples and their respective PDXs revealed a high concordance in drug sensitivity (R2 0.697-0.892, P <0.001). HTS not only confirmed expected drug sensitivities (Z score <-2.0, IC50 <10µM) from the molecular curation but also identified drug sensitivities not predicted by the molecular data, specific to individual patients or more generally in paediatric AML. Finally, ex vivo drug sensitivity of PDXs was found to be concordant with in vivo drug response in MISTRG mice.

 

Conclusion:

These results demonstrate that AML PDXs derived in MISTRG mice provide a highly relevant representation of the primary disease at both the molecular and drug response levels.  A combination of molecular analysis and HTS can identify specific therapies for individual AML patients or discover novel drug sensitivities in paediatric AML that would otherwise be missed by molecular data alone.