Contributions
Abstract: EP335
Type: E-Poster Presentation
Session title: Acute lymphoblastic leukemia - Biology & Translational Research
Background
Despite intensified treatment regimens, the outcome for T-ALL patients remains inferior to pre-B-ALL patients, especially for patients that are refractory to treatment or that relapse within 4 years after the start of therapy. An important prognostic factor for cancer relapse, in general, is tumor heterogeneity at the time of diagnosis. Yet, it remains unclear to what extend intra-tumor heterogeneity appears in pediatric T-ALL patients, and especially whether this reflects the heterogeneity in T-cell developmental arrest or heterogeneity in acquired mutations. T-ALL is initiated by the activation of specific oncogenic transcription factors due to chromosomal rearrangements. Activation of specific oncogenes has been linked to four different subtypes, namely: ETP-ALL/Immature, TLX, NKX2-1/TLX1, or TALLMO. The gene expression signatures and immunophenotypes of these four T-ALL entities reflect differences in developmental arrests comparable to the various developmental stages of healthy thymocytes. In the current study, we aim to decipher the immunophenotypic make-up and intra-tumor heterogeneity in T-ALL patients at diagnosis. Accessing T-ALL intra-tumor heterogeneity will further increase our insights on T-ALL disease onset and development that may drive therapy resistance and relapse.
Aims
To investigate the level of immunophenotypic heterogeneity in T-ALL patients.
Methods
We have designed and optimized a 20-parameter multi-color flow cytometry panel to identify all known thymocyte developmental stages using thymus samples of healthy donors. Thereafter this optimized panel was used to screen 28 bio-banked diagnostic T-ALL patient samples from all four T-ALL subtypes. Bioinformatic analysis using t-stochastic neighbor embedding (t-SNE) was used to visualize immunophenotypic heterogeneity in each T-ALL sample. Samples were also matched with normal thymocyte subpopulations to assess the stage of developmental arrest. Lastly, we use the FlowSOM package to differentially compare levels of heterogeneity in matched diagnostic and relapse patient samples and diagnostic PDX models.
Results
We show immunophenotypic heterogeneity in 26 of 28 diagnostic T-ALL patient samples. Within each patient, immunophenotypic subclones reflect differential sites of normal developmental arrest. Next, we observed that immunophenotypic heterogeneity is preserved in diagnostic and relapsed PDX samples, where relapse PDX samples grow out from minor subpopulations that are present already at diagnosis.
Conclusion
T-ALL patients show a high degree of immunophenotypic heterogeneity at diagnosis suggesting differential sites of developmental arrest at disease initiation. Moreover, the persistence of immunophenotypic subclones after PDX generation shows the competition of subclones under selective pressure while still maintaining heterogeneity. Our future efforts will focus on correlating the immunophenotypic heterogeneity with underlying mutations to improve our understanding of T-ALL intra-tumor heterogeneity at different multi-omic levels. Deciphering the different layers of T-ALL heterogeneity will greatly impact the current targeted treatment strategies that focus on specific mutations but may not acknowledge underlying heterogeneity.
Keyword(s): Flow cytometry, Immunophenotype, T-ALL, Thymus
Abstract: EP335
Type: E-Poster Presentation
Session title: Acute lymphoblastic leukemia - Biology & Translational Research
Background
Despite intensified treatment regimens, the outcome for T-ALL patients remains inferior to pre-B-ALL patients, especially for patients that are refractory to treatment or that relapse within 4 years after the start of therapy. An important prognostic factor for cancer relapse, in general, is tumor heterogeneity at the time of diagnosis. Yet, it remains unclear to what extend intra-tumor heterogeneity appears in pediatric T-ALL patients, and especially whether this reflects the heterogeneity in T-cell developmental arrest or heterogeneity in acquired mutations. T-ALL is initiated by the activation of specific oncogenic transcription factors due to chromosomal rearrangements. Activation of specific oncogenes has been linked to four different subtypes, namely: ETP-ALL/Immature, TLX, NKX2-1/TLX1, or TALLMO. The gene expression signatures and immunophenotypes of these four T-ALL entities reflect differences in developmental arrests comparable to the various developmental stages of healthy thymocytes. In the current study, we aim to decipher the immunophenotypic make-up and intra-tumor heterogeneity in T-ALL patients at diagnosis. Accessing T-ALL intra-tumor heterogeneity will further increase our insights on T-ALL disease onset and development that may drive therapy resistance and relapse.
Aims
To investigate the level of immunophenotypic heterogeneity in T-ALL patients.
Methods
We have designed and optimized a 20-parameter multi-color flow cytometry panel to identify all known thymocyte developmental stages using thymus samples of healthy donors. Thereafter this optimized panel was used to screen 28 bio-banked diagnostic T-ALL patient samples from all four T-ALL subtypes. Bioinformatic analysis using t-stochastic neighbor embedding (t-SNE) was used to visualize immunophenotypic heterogeneity in each T-ALL sample. Samples were also matched with normal thymocyte subpopulations to assess the stage of developmental arrest. Lastly, we use the FlowSOM package to differentially compare levels of heterogeneity in matched diagnostic and relapse patient samples and diagnostic PDX models.
Results
We show immunophenotypic heterogeneity in 26 of 28 diagnostic T-ALL patient samples. Within each patient, immunophenotypic subclones reflect differential sites of normal developmental arrest. Next, we observed that immunophenotypic heterogeneity is preserved in diagnostic and relapsed PDX samples, where relapse PDX samples grow out from minor subpopulations that are present already at diagnosis.
Conclusion
T-ALL patients show a high degree of immunophenotypic heterogeneity at diagnosis suggesting differential sites of developmental arrest at disease initiation. Moreover, the persistence of immunophenotypic subclones after PDX generation shows the competition of subclones under selective pressure while still maintaining heterogeneity. Our future efforts will focus on correlating the immunophenotypic heterogeneity with underlying mutations to improve our understanding of T-ALL intra-tumor heterogeneity at different multi-omic levels. Deciphering the different layers of T-ALL heterogeneity will greatly impact the current targeted treatment strategies that focus on specific mutations but may not acknowledge underlying heterogeneity.
Keyword(s): Flow cytometry, Immunophenotype, T-ALL, Thymus