Figure 4.
Serial lymphocyte phenotyping prior to and following CD22 CAR T-cell infusion in patients with vs without carHLH. (A) Bone marrow T-cell percentage (% of lymphocytes) prior to CAR infusion (pre-CAR) in those who would and would not go on to develop carHLH. Graph shows scatter plot of all values with lines at median and IQR. (B) Bone marrow NK cell percentage (% of lymphocytes) pre-CAR in those who would and would not go on to develop carHLH. Graph shows scatter plot of all values with lines at median and IQR. (C) Ratio of T:NK cells was calculated for subjects using pre-CAR infusion baseline bone marrow flow. A higher T:NK ratio was observed in subjects who would subsequently go on to develop carHLH. Graphs show scatter plot of all values with lines at median and IQR. (D) Pre-CAR disease burden in the bone marrow, reflecting bone marrow blasts as percent of marrow mononuclear cells by flow cytometry. There were no substantial differences seen between the 2 groups. (E) Bone marrow CAR T-cell (percent of CD3+ T cells) percentage at restaging in those with and without carHLH. Graph shows scatter plot of all values with lines at median and IQR. (F) Bone marrow NK cell percentage (percent of lymphocytes) at restaging in those with and without carHLH. Graph shows scatter plot of all values with lines at median and IQR. (G) Bone marrow T:NK cell ratio at restaging in those with and without carHLH. Graph shows scatter plot of all values with lines at median and IQR. (H) Bone marrow CAR T:NK cell ratio at restaging in those with and without carHLH. Graph shows scatter plot of all values with lines at median and IQR. (I) In carHLH patients, there is a relative predominance of CD22 CAR T cells in patients with carHLH at day 14 time point, which persists through day 28 time point. Line graphs are drawn through median value for each time point, with thin vertical lines denoting IQR. (J) In subjects with carHLH, NK cells comprise a smaller percentage of the cellular milieu at pre-CAR infusion baseline on the day of CAR infusion (day 0) as well as at the day 14 and day 28 time points. Patients with carHLH are characterized by an absence of NK cell reexpansion at the latter time point, compared with a marked reexpansion toward baseline levels in those without carHLH (supplemental Figure 4L). Line graphs are drawn through median value for each time point, with thin vertical lines denoting IQR. (K) A ratio of T:NK cells was calculated at all time points. Higher peripheral blood T:NK ratio was seen in carHLH patients at pre-CAR infusion baseline, day 14 (coinciding with median day of carHLH onset), and day 28 time points. A trend toward higher T:NK ratio was observed at the day 0 time point (P = .08). Line graphs are drawn through median value for each time point, with thin vertical lines denoting IQR. (L) In subjects with carHLH, monocytes comprise a smaller percentage of the cellular milieu from day 7 time point onwards, with most substantial differences corresponding to median carHLH onset. Line graphs are drawn through the median value for each time point, with thin vertical lines denoting IQR. (M-O) Heatmap representation of row-normalized mean frequencies of immune cell subsets between carHLH and non-carHLH groups at preinfusion (M), peak CAR expansion (N), and at day 28 (O) time points. Frequencies were analyzed using a Mann-Whitney U test. *P < .05, **P < .01, ***P < .001, ****P < .0001. (P) Paired flow cytometric analysis of perforin expression in peripheral blood at preinfusion and day 14 post-CAR T-cell infusion in patients with carHLH (n = 6) and those without carHLH (n = 5). No differences were observed in expression levels between the cohorts at either time point. (Q) Paired flow cytometric analysis of granzyme B expression in peripheral blood at preinfusion and day 14 post-CAR T-cell infusion in patients with carHLH (n = 6) and those without carHLH (n = 5). No differences were observed in expression levels between the cohorts at either time point. (R) Among those with CRS, peak CAR expansion as percentage of total lymphocyte population was higher in those who underwent CD4/CD8 T-cell selection (TCS) compared with those who underwent CD3/CD28 T-cell enrichment (TCE) during CAR manufacturing process. Color indicates carHLH status for a given datapoint. Graphs show scatter plot of all values with lines at median and IQR.

Serial lymphocyte phenotyping prior to and following CD22 CAR T-cell infusion in patients with vs without carHLH. (A) Bone marrow T-cell percentage (% of lymphocytes) prior to CAR infusion (pre-CAR) in those who would and would not go on to develop carHLH. Graph shows scatter plot of all values with lines at median and IQR. (B) Bone marrow NK cell percentage (% of lymphocytes) pre-CAR in those who would and would not go on to develop carHLH. Graph shows scatter plot of all values with lines at median and IQR. (C) Ratio of T:NK cells was calculated for subjects using pre-CAR infusion baseline bone marrow flow. A higher T:NK ratio was observed in subjects who would subsequently go on to develop carHLH. Graphs show scatter plot of all values with lines at median and IQR. (D) Pre-CAR disease burden in the bone marrow, reflecting bone marrow blasts as percent of marrow mononuclear cells by flow cytometry. There were no substantial differences seen between the 2 groups. (E) Bone marrow CAR T-cell (percent of CD3+ T cells) percentage at restaging in those with and without carHLH. Graph shows scatter plot of all values with lines at median and IQR. (F) Bone marrow NK cell percentage (percent of lymphocytes) at restaging in those with and without carHLH. Graph shows scatter plot of all values with lines at median and IQR. (G) Bone marrow T:NK cell ratio at restaging in those with and without carHLH. Graph shows scatter plot of all values with lines at median and IQR. (H) Bone marrow CAR T:NK cell ratio at restaging in those with and without carHLH. Graph shows scatter plot of all values with lines at median and IQR. (I) In carHLH patients, there is a relative predominance of CD22 CAR T cells in patients with carHLH at day 14 time point, which persists through day 28 time point. Line graphs are drawn through median value for each time point, with thin vertical lines denoting IQR. (J) In subjects with carHLH, NK cells comprise a smaller percentage of the cellular milieu at pre-CAR infusion baseline on the day of CAR infusion (day 0) as well as at the day 14 and day 28 time points. Patients with carHLH are characterized by an absence of NK cell reexpansion at the latter time point, compared with a marked reexpansion toward baseline levels in those without carHLH (supplemental Figure 4L). Line graphs are drawn through median value for each time point, with thin vertical lines denoting IQR. (K) A ratio of T:NK cells was calculated at all time points. Higher peripheral blood T:NK ratio was seen in carHLH patients at pre-CAR infusion baseline, day 14 (coinciding with median day of carHLH onset), and day 28 time points. A trend toward higher T:NK ratio was observed at the day 0 time point (P = .08). Line graphs are drawn through median value for each time point, with thin vertical lines denoting IQR. (L) In subjects with carHLH, monocytes comprise a smaller percentage of the cellular milieu from day 7 time point onwards, with most substantial differences corresponding to median carHLH onset. Line graphs are drawn through the median value for each time point, with thin vertical lines denoting IQR. (M-O) Heatmap representation of row-normalized mean frequencies of immune cell subsets between carHLH and non-carHLH groups at preinfusion (M), peak CAR expansion (N), and at day 28 (O) time points. Frequencies were analyzed using a Mann-Whitney U test. *P < .05, **P < .01, ***P < .001, ****P < .0001. (P) Paired flow cytometric analysis of perforin expression in peripheral blood at preinfusion and day 14 post-CAR T-cell infusion in patients with carHLH (n = 6) and those without carHLH (n = 5). No differences were observed in expression levels between the cohorts at either time point. (Q) Paired flow cytometric analysis of granzyme B expression in peripheral blood at preinfusion and day 14 post-CAR T-cell infusion in patients with carHLH (n = 6) and those without carHLH (n = 5). No differences were observed in expression levels between the cohorts at either time point. (R) Among those with CRS, peak CAR expansion as percentage of total lymphocyte population was higher in those who underwent CD4/CD8 T-cell selection (TCS) compared with those who underwent CD3/CD28 T-cell enrichment (TCE) during CAR manufacturing process. Color indicates carHLH status for a given datapoint. Graphs show scatter plot of all values with lines at median and IQR.

Close Modal
This Feature Is Available To Subscribers Only

or Create an Account

Close Modal
Close Modal