Associations of T-Cell Receptor Repertoire Diversity with L-Asparaginase Allergy in Childhood Acute Lymphoblastic Leukemia

Abstract

Asparaginase is a critical component of therapy for childhood acute lymphoblastic leukemia (ALL), but it is commonly associated with allergy, which results in morbidity and poorer outcomes. The underlying basis of this allergy is undoubtedly immune-mediated, but the exact components of T-cell immunity have yet to be characterized. We performed longitudinal TCR sequencing of 180 bone marrow samples from 67 children with B-ALL treated as part of the Ma-Spore-ALL-2010 trial, and we evaluated the associations of TCR profile with asparaginase hypersensitivity, with functional validation of asparaginase activity in a separate cohort of 113 children. We found that a more diverse and dynamically changing TCR repertoire was associated with increased risk of clinical hypersensitivity and decreased L-asp activity. Allergic patients had a higher proportion of infrequent clonotypes, as well as a significantly lower degree of shared clonotypes amongst the cohort. Allergic patients also had significantly higher longitudinal variability of clonotypes across timepoints, where a higher dissimilarity between diagnosis and week 5 represented an 8.1-fold increased risk of an allergic event. After an allergy had occurred, there was shaping and convergence of the TCR repertoire towards a common antigen. Understanding the immunological basis of T-cell responses in allergy lays the groundwork for developing predictive biomarkers or strategies to mediate this common toxicity in childhood ALL.

Keywords: L-asparaginase; T-cell receptor repertoire; allergy; childhood acute lymphoblastic leukemia; hypersensitivity.

Figure 1

L-asparaginase allergy in childhood ALL, and the association of TCR-B diversity with occurrence of subsequent allergy: (A) Overall study schema. The phases of chemotherapy in the Ma-Spore 2010 protocol are shown, with timepoints of bone marrow sampling indicated by yellow arrows and occurrences of allergic episodes indicated by red crosses. (B) Number of patients with allergy in the cohort, and the distribution of allergy grades amongst the allergy episodes. The left pie chart depicts the proportion of patients with allergy (in red), and the right pie chart depicts the distribution of severity of 16 episodes of allergy. (C) Shannon’s entropy in non-allergy vs. allergy groups throughout all timepoints. Shannon’s entropy is plotted for each patient in the scatterplot at each timepoint and compared between pre-allergic and non-allergic patients; p-values determined by the Mann–Whitney test. (D) Inverse Simpson’s index in non-allergy vs. allergy groups throughout all timepoints. The inverse Simpson’s index is plotted for each patient in the scatterplot at each timepoint and compared between pre-allergic and non-allergic patients. Both Shannon’s entropy and the inverse Simpson’s index are higher in pre-allergic patients at all timepoints, indicating a higher TCR repertoire diversity in this group of patients who would later go on to develop allergy. p-Values determined by the Mann–Whitney test. (E) Shannon’s entropy in normal L-asp activity vs. L-asp inactivation in the functional validation cohort. Shannon’s entropy is plotted for each patient in the scatterplot and compared between patients with normal L-asp activity (blue) and patients with L-asp inactivation (red). L-asp inactivation is defined as an activity level of <100 U/L. p-Values determined by the Mann–Whitney test. For (B), (C), and (D), patients without allergy are indicated in blue, and patients who would go on to develop allergy are indicated in red. The median of each group is indicated by a bold horizontal line in the boxplot.

Figure 2

TCR-B diversity in allergic patients is characterized by a higher frequency of infrequently occurring clonotypes, which are less shared amongst patients and display more longitudinal variability: (A) Proportion of common (≥0.5%) and infrequent (<0.5%) clones in non-allergy vs. allergy groups. The proportions of common (≥0.5%) and infrequent clones (<0.5%) are shown in the bar chart. Non-allergy is shown in blue, and allergy is shown in red. Allergic patients have a significantly higher proportion of infrequent clones and a lower proportion of common clones. p-Values determined by the chi-squared test. (B) Proportion of rare clones (<0.005%) in non-allergy vs. allergy groups at diagnosis. The proportions of rare clones are plotted in the violin plots for the non-allergy vs. allergy groups. Non-allergy is shown in blue, and allergy is shown in red. The median in each group is noted as a bold horizontal black line. p-Values determined by the Mann–Whitney test. (C) Proportion of shared clonotypes in non-allergy vs. allergy groups. The definition of shared clonotypes is as described in the Materials and Methods section. The proportion of shared clonotypes is shown in the dot plot for non-allergy (blue) vs. pre-allergy (green). p-Values determined by the Mann–Whitney test. (D) Longitudinal similarity of clonotypes in non-allergy vs. allergy groups. Similarity of the TCR repertoires of the same patient between diagnosis and week 5 was measured by the Bhattacharyya similarity coefficient, and the definition is as described in the Materials and Methods section. The similarity coefficients of clones for each patient are plotted in the dot plot for non-allergy vs. allergy. Patients with allergy had significantly more variability between timepoints in their TCR repertoires compared to those without allergy. p-Values determined by the Mann–Whitney test. (E) Cumulative incidence of allergy in patients with similarity coefficients ≤0.05 vs. >0.05. Similarity of TCR repertoires of each patient between diagnosis and week 5 was measured by the Bhattacharyya similarity coefficient, and the definition is as described in the Materials and Methods section. We divided patients into two groups of similarity coefficients: ≤0.05 (i.e., less similar, shown in red) and >0.05 (i.e., more similar, shown in yellow). There was a significantly higher cumulative incidence of allergy when the similarity coefficient was <0.05, i.e., those who had more variability of clonotypes were at higher risk of allergy, and those with a more static repertoire were at less risk. p-Values determined by Cox proportional hazard regression. (F) The area under the curve (AUC) of the respective receiver operating characteristic (ROC) curves was calculated for the similarity coefficient of diagnosis vs. week 5.

Figure 3

Antigenic shaping of TCR repertoires after the occurrence of asparaginase allergy: (A) Interrelation between clones in non-allergy, pre-allergy, and post-allergy groups. The interrelation of clones to one another was determined using the normalized Damerau–Levenshtein distance, which measures the minimum number of substitutions, insertions, and/or deletions necessary to transform one sequence into another. Non-allergy is noted in blue, pre-allergy is noted in green, and post-allergy is noted in red. Clones were less interrelated in pre-allergic patients and became significantly more closely related after the occurrence of allergy, suggesting a convergent response. Overall p-values determined by the Kruskal–Wallis test, and p-values comparing specific pairs determined by the Mann–Whitney test. (B) Clonotypic relationships in a patient before and after grade 4 allergy. In a graphical representation of random sampling of 500 clones, each dot represents a clonotype. Clones that are similar/related are denoted by a gray line linking the two clones. Post-allergy, there is a striking convergence of clonotypes towards one another, indicating a convergent response towards a similar epitope. (C) Proportions of TRB-V and TRB-J segment use in pre- and post-allergy. The proportion of each segment’s usage is shown in the bar chart comparing the 4 patients, with sampling performed pre- and post-allergy. Asterisks indicate significant (p < 0.05) differential usage according to the Mann–Whitney test. TRB-V: pre-allergy is noted in blue, and post-allergy is noted in red. TRB-J: pre-allergy is noted in green, and post-allergy is noted in red. (D) Differentially expressed VJ segment usage. Hierarchical clustering was performed for the top 10 differentially utilized V and J gene segments. The left graph indicates TRB-J (high usage indicated in yellow and low usage indicated in dark blue), while the right graph indicates TRB-V (high usage indicated in red and low usage indicated in blue). The top horizontal panel indicates clustering of pre-allergy (dark blue) and post-allergy (light blue) samples.