Back

Proteogenomic analysis identifies the persistence of targetable lesions and proteomes through disease evolution in pediatric acute lymphoblastic leukemia

Lorentzian, A. C.; Guo, M.; Ergin, E. K.; Akella, N. M.; Rolf, N.; Lim, C. J.; Reid, G. S.; Maxwell, C. A.; Lange, P. F.

2022-03-08 oncology
10.1101/2022.03.04.22271927 medRxiv
Show abstract

Childhood acute lymphoblastic leukemia (ALL) genomes show that relapses often arise from subclonal outgrowths. However, the impact of clonal evolution on the actionable proteome and response to targeted therapy is not known. Here, we present a comprehensive retrospective analysis of paired ALL diagnosis and relapsed specimen. Targeted next generation sequencing and proteome analysis indicated persistence of actionable genome variants and stable proteomes through disease progression. Paired viably-frozen biopsies showed high correlation of drug response to variant-targeted therapies but in vitro selectivity was low. Proteome analysis prioritized PARP1 as a new pan-ALL target candidate needed for survival following cellular stress; diagnostic and relapsed ALL samples demonstrated robust sensitivity to treatment with two PARP1/2 inhibitors. Together, these findings support initiating prospective precision oncology approaches at ALL diagnosis and emphasize the need to incorporate proteome analysis to prospectively determine tumor sensitivities, which are likely to be retained at disease relapse. STATEMENT OF SIGNIFICANCEWe discover that disease progression and evolution in pediatric acute lymphoblastic leukemia is defined by the persistence of targetable genomic variants and stable proteomes, which reveal pan-ALL target candidates. Thus, personalized treatment options in childhood ALL may be improved with the incorporation of prospective proteogenomic approaches initiated at disease diagnosis.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.