P365 Triple checkpoint blockade targeting PD -1, TIM-3 ... · PD-1, TIM-3, and LAG-3, eventually...
1
Triple checkpoint blockade targeting PD-1, TIM-3, and LAG-3 improves T cell reinvigoration and antitumor efficacy over single and double combinations Johanna K. Kaufmann, Geeta Sharma, Srimoyee Ghosh, Sujatha Kumar, Kevin G. Coleman, Sridhar Ramaswamy, David Jenkins TESARO, Inc., Waltham, MA, USA 33 rd Annual Meeting of the Society for Immunotherapy of Cancer, Washington, DC, November 7–11, 2018 • Triple blockade of PD-1, TIM-3, and LAG-3 resulted in highly effective reversal of T cell exhaustion in a murine model. • Triple blockade improved tumor growth control over single or double combinations. • Targeting immune checkpoints expressed on multiple cell types engaged additional mechanisms of tumor immune control. • Triple combination treatment with checkpoint inhibitors was able to reinvigorate human TILs most effectively. • Overall, these findings support the concept of double and triple combinations of antibodies that block PD-1, TIM-3, and LAG-3. 1. Koyama S et al. Nat Commun. 2016;7:10501. 2. Shayan G et al. Oncoimmunology. 2017;6:e1261779. 3. Taube JM et al. Clin Cancer Res. 2015;21:3969-3976. 4. Thommen DS et al. Cancer Immunol Res. 2015;3:1344-1355. 5. Ghosh S et al. Cancer Res. 2018;78 suppl 13:5690. 6. NCT02715284, NCT02817633, NCT03250832, NCT03307785, NCT03574779, NCT03602859, NCT03651206, NCT03680508. Writing and editorial support, funded by TESARO, Inc. (Waltham, MA, USA) and coordinated by Hemant Vyas, PhD, of TESARO, Inc., was provided by Ashfield Healthcare Communications (Middletown, CT, USA). In vitro exhaustion assay Splenocytes from MBP-Tracker mice with CD4 + T cell receptor transgenic T cells 72 h peptide stimulation with APL-MBP Exhausted T cells 72 h restimulation of CD4 + T cells with APL-MBP peptide presented on irradiated APCs with immune checkpoint inhibitors Measure T cell activity by IFN-γ release Splenocytes were isolated from mice engineered to express a CD4 + T cell receptor recognizing an epitope of myelin basic protein (MBP). Cells were stimulated with either MBP peptide to generate effector T cells or with a superagonistic altered peptide ligand (APL) of MBP to generate exhausted T cells. Immune cells were isolated by density gradient centrifugation and rested for 4 days before restimulation with APL-MBP-loaded and irradiated antigen-presenting cells (APCs). IFN-γ release was quantified in supernatants by ELISA (Figure 3). Ex vivo TIL reinvigoration Primary resections from treatment-naïve ovarian cancer patients were dissociated and CD45 + tumor infiltrating lymphocytes (TILs) were enriched by magnetic bead isolation. TILs were then restimulated with S. aureus enterotoxin B (SEB) and cytokine release into the supernatant was quantified by cytometric bead array (CBA, Figure 6). Ovarian cancer tumor resection at clinical site Dissociation into single cell suspension 48 h restimulation with 100 μg/ml SEB with immune checkpoint inhibitors TIL enrichment Measure T cell activity by IFN-γ and IL-2 release Syngenic in vivo model for efficacy and pharmacodynamics EMT6 murine mammary carcinoma cells were implanted subcutaneously (sc.) into flanks of Balb/c mice and animals were randomized across treatment groups on day 4 post-inoculation, when tumor volumes ranged between 50 and 80 mm 3 (n = 10). Animals were dosed biweekly (BIW) with single or multiple antibodies at 200 μg per antibody for each mouse. Tumor growth and peripheral and local immune cell composition on day 20 were monitored as indicated (Figures 4 and 5). -4 0 4 9 13 18 20 immunophenotyping of spleens and TILs antibody treatment, BIW EMT6 sc. in Balb/c tumor volume measurement METHODS CONCLUSIONS REFERENCES ACKNOWLEDGEMENTS Figure 2. Co-expression of PD-1, TIM-3, and LAG-3 on human tumor- infiltrating lymphocytes (TILs) isolated from multiple tumor tissues Fresh tumor tissue samples across 15 different tumor types were collected and dissociated to single cell suspensions before immune profiling by flow cytometry. For each tumor type, the frequency of PD-1, TIM-3, and LAG-3 co-expressing cells was determined for 4 TIL subsets, defined by classical lineage markers. Here, we explore the functional effects of triple blockade of PD-1, TIM-3, and LAG-3 on T cell reinvigoration, tumor immune contexture, and antitumor activity in preclinical models. RESULTS BACKGROUND OBJECTIVES Figure 1. Therapeutic concept T cell Anti-PD(L)-1 PD-1 TIM-3 α-TIM-3 Tumor cells PDL-1 α-PD-1 TIM-3 Boost anti- tumor activity T cell invigoration Amplify immune function Overcome secondary resistance LAG-3 α-LAG-3 LAG-3 α-PD-1 α-LAG-3 α-TIM-3 T cells upregulate multiple checkpoint receptors in response to chronic activation, including PD-1, TIM-3, and LAG-3, eventually resulting in a dysfunctional or exhausted phenotype. Triple blockade targeting all three of these immune checkpoint receptors may therefore be more effective at T cell reinvigoration. In addition, increases in TIM-3 and LAG-3 expression have been described in the context of acquired resistance to PD-1/PD-L1 axis blockade, which may be overcome by combination checkpoint blockade. 1-4 Furthermore, TIM-3 and LAG-3 have been described to modulate activity of other cell types, including NK and dendritic cells, suggesting potential additional mechanisms of action during triple immune checkpoint blockade. Exhausted T cell Exhausted T cell CD8 + T cells CD4 + T cells Figure 3. Triple combination treatment was able to fully reverse the exhausted phenotype in an antigen-specific CD4 + T cell model (a) In vitro exhausted CD4 + T cells were restimulated for 72 h with a superagonistic peptide in the presence of single, double, and triple antibody combinations at 0.32 µM total antibody concentration. For single and double combinations, the remaining thirds of antibody were topped off with isotype control. Means and standard deviations of normalized IFN-γ secretions are shown, n = 4. The dotted line indicates IFN-γ secretion by non-exhausted T cells. * p < 0.05, ** p < 0.01, **** p < 0.0001 by one-way ANOVA with Tukey’s multiple comparisons test. (b) A dose-titration of triple antibody combination and isotype control was performed in the same assay as in (a). Means and standard deviations of normalized IFN-γ secretions are shown, n = 4. The dotted line indicates IFN-γ secretion by non-exhausted T cells * p < 0.05, *** p < 0.001, **** p < 0.0001 over isotype control at the same concentration by two-way ANOVA with Sidak’s multiple comparisons test. One of two representative experiments is shown. a b Figure 4. Maximum TGI was observed in the triple combination group in an EMT-6 syngenic mouse breast cancer model Triple combination treatment targeting human PD-1, TIM-3, and LAG-3 using TSR-042, TSR-022, and TSR-03, respectively, was also most effective in an MDA-MB-436 triple-negative breast cancer model in humanized NOG-EXL mice. 5 Isotype control α-PD-1 α-TIM-3 α-LAG-3 α-PD-1 + α-TIM-3 α-PD-1 + α-LAG-3 α-PD-1 + α-LAG-3 + α-TIM-3 TGI 24% TGI 6% TGI 11% Figure 5. Triple combination treatment modulated multiple aspects of the tumor microenvironment EMT6 tumors were harvested on day 20 from all remaining animals. Tumors were dissociated into single-cell suspension and infiltrating immune cells were characterized by flow cytometry. Frequencies of selected cell populations of interest are shown as means and standard deviations, along with values for individual animals (n = 7-9). ** p < 0.01, **** p < 0.0001 by one-way ANOVA with Dunnett’s multiple comparisons test vs isotype control. T cells Myeloid cells Figure 6. Triple combination treatment is able to reinvigorate TILs from human ovarian cancer patients TILs were isolated from primary ovarian cancer resections from treatment-naïve patients. IFN-γ and IL-2 secretion was quantified after restimulation with 100 μg/ml SEB for 48 h in the presence of immune checkpoint inhibitory antibodies targeting PD-1 (TSR-042), TIM-3 (TSR- 022), and LAG-3 (TSR-033). All three proprietary antibodies (TESARO, Inc.) are being evaluated in clinical trials 6 . Each antibody was used at 10 μg/ml and cytokine levels were normalized to isotype control conditions. Means, standard deviation, and individual data points are shown, n = 2-3. * p < 0.05, *** p < 0.001 over isotype control by one-way ANOVA with Dunnett’s multiple comparisons test. IFN-γ IL-2 TGI 40% TGI 50% TGI 61% Volumes of subcutaneous EMT6 tumors were monitored every other day during treatment with the indicated antibodies, administered biweekly alone or in combination depending on the treatment group (200 μg per antibody for each mouse). Tumor volumes are shown for each animal (one line per animal, n = 10) and tumor growth inhibition (TGI) was calculated. Dendritic cells NK cells COAD, colon adenocarcinoma, n = 95; KIRC, kidney clear cell carcinoma, n = 54; LUAD/LUSC, lung adenocarcinoma and squamous cell carcinoma, n= 54; ESCA, esophageal carcinoma, n= 21; CESC, cervical squamous cell and adenocarcinoma, n = 11; UCEC, uterine corpus endometrial carcinoma, n = 10; STAD, stomach adenocarcinoma, n = 9; BLCA, bladder urothelial carcinoma, n = 7; HEP, liver cancers, 7; HN, head and neck carcinomas, n = 7; PAAD, pancreatic adenocarcinoma, n = 6; OV, ovarian adenocarcinoma, n = 5; BRCA, breast invasive carcinoma, n = 2; SKME, skin cutaneous melanoma, n = 2; PRAD, prostate adenocarcinoma, n = 1. P365 Total immune infiltrate
Transcript of P365 Triple checkpoint blockade targeting PD -1, TIM-3 ... · PD-1, TIM-3, and LAG-3, eventually...
Triple checkpoint blockade targeting PD-1, TIM-3, and LAG-3 improves T cellreinvigoration and antitumor efficacy over single and double combinations
Johanna K. Kaufmann, Geeta Sharma, Srimoyee Ghosh, Sujatha Kumar, Kevin G. Coleman, Sridhar Ramaswamy, David JenkinsTESARO, Inc., Waltham, MA, USA
33rd Annual Meeting of the Society for Immunotherapy of Cancer, Washington, DC, November 7–11, 2018
• Triple blockade of PD-1, TIM-3, and LAG-3 resulted in highly effective reversal of T cell exhaustion in a murine model.
• Triple blockade improved tumor growth control over single or double combinations.
• Targeting immune checkpoints expressed on multiple cell types engaged additional mechanisms of tumor immune control.
• Triple combination treatment with checkpoint inhibitors was able to reinvigorate human TILs most effectively.
• Overall, these findings support the concept of double and triple combinations of antibodies that block PD-1, TIM-3, and LAG-3.
1. Koyama S et al. Nat Commun. 2016;7:10501.2. Shayan G et al. Oncoimmunology. 2017;6:e1261779.3. Taube JM et al. Clin Cancer Res. 2015;21:3969-3976.4. Thommen DS et al. Cancer Immunol Res. 2015;3:1344-1355.5. Ghosh S et al. Cancer Res. 2018;78 suppl 13:5690.6. NCT02715284, NCT02817633, NCT03250832, NCT03307785, NCT03574779,
NCT03602859, NCT03651206, NCT03680508.
Writing and editorial support, funded by TESARO, Inc. (Waltham, MA, USA) and coordinated by Hemant Vyas, PhD, of TESARO, Inc., was provided by Ashfield Healthcare Communications (Middletown, CT, USA).
In vitro exhaustion assaySplenocytes from MBP-Tracker mice with CD4+ T cell receptor transgenic T cells
72 h peptide stimulation with APL-MBP
Exhausted T cells
72 h restimulation of CD4+ T cells with APL-MBP peptide presented on irradiated APCs with immune checkpoint inhibitors
Measure T cell activity by IFN-γ releaseSplenocytes were isolated from mice engineered to express a CD4+ T cell receptor recognizing an epitope of myelin basic protein (MBP). Cells were stimulated with either MBP peptide to generate effector T cells or with a superagonistic altered peptide ligand (APL) of MBP to generate exhausted T cells. Immune cells were isolated by density gradient centrifugation and rested for 4 days before restimulation with APL-MBP-loaded and irradiated antigen-presenting cells (APCs). IFN-γ release was quantified in supernatants by ELISA (Figure 3).
Ex vivo TIL reinvigoration
Primary resections from treatment-naïve ovarian cancer patients were dissociated and CD45+ tumor infiltrating lymphocytes (TILs) were enriched by magnetic bead isolation. TILs were then restimulated with S. aureus enterotoxin B (SEB) and cytokine release into the supernatant was quantified by cytometric bead array (CBA, Figure 6).
Ovarian cancer tumor resection at clinical site
Dissociation into single cell suspension
48 h restimulation with 100 μg/ml SEB with immune checkpoint inhibitors
TIL enrichment
Measure T cell activity by IFN-γ and IL-2 release
Syngenic in vivo model for efficacy and pharmacodynamics
EMT6 murine mammary carcinoma cells were implanted subcutaneously (sc.) into flanks of Balb/c mice and animals were randomized across treatment groups on day 4 post-inoculation, when tumor volumes ranged between 50 and 80 mm3 (n = 10). Animals were dosed biweekly (BIW) with single or multiple antibodies at 200 μg per antibody for each mouse. Tumor growth and peripheral and local immune cell composition on day 20 were monitored as indicated (Figures 4 and 5).
-4 0 4 9 13 18 20
immunophenotyping ofspleens and TILs
antibody treatment, BIWEMT6 sc. in Balb/c
tumor volume measurement
METHODS CONCLUSIONS
REFERENCES
ACKNOWLEDGEMENTS
Figure 2. Co-expression of PD-1, TIM-3, and LAG-3 on human tumor-infiltrating lymphocytes (TILs) isolated from multiple tumor tissues
Fresh tumor tissue samples across 15 different tumor types were collected and dissociated to single cell suspensions before immune profiling by flow cytometry. For each tumor type, the frequency of PD-1, TIM-3, and LAG-3 co-expressing cells was determined for 4 TIL subsets, defined by classical lineage markers.
Here, we explore the functional effects of triple blockade of PD-1, TIM-3, and LAG-3 on T cell reinvigoration, tumor immune contexture, and antitumor activity in preclinical models.
RESULTS
BACKGROUND
OBJECTIVES
Figure 1. Therapeutic concept
T cell
Anti-PD(L)-1
PD-1
TIM-3 α-TIM-3
Tumor cellsPDL-1
α-PD-1
TIM-3
Boost anti-tumor activity
T cell invigoration
Amplify immune function
Overcome secondary resistance
LAG-3
α-LAG-3
LAG-3
α-PD-1α-LAG-3
α-TIM-3
T cells upregulate multiple checkpoint receptors in response to chronic activation, including PD-1, TIM-3, and LAG-3, eventually resulting in a dysfunctional or exhausted phenotype. Triple blockade targeting all three of these immune checkpoint receptors may therefore be more effective at T cell reinvigoration. In addition, increases in TIM-3 and LAG-3 expression have been described in the context of acquired resistance to PD-1/PD-L1 axis blockade, which may be overcome by combination checkpoint blockade.1-4 Furthermore, TIM-3 and LAG-3 have been described to modulate activity of other cell types, including NK and dendritic cells, suggesting potential additional mechanisms of action during triple immune checkpoint blockade.
ExhaustedT cell
ExhaustedT cell
CD8+ T cells CD4+ T cells
Figure 3. Triple combination treatment was able to fully reverse the exhausted phenotype in an antigen-specific CD4+ T cell model
(a) In vitro exhausted CD4+ T cells were restimulated for 72 h with a superagonisticpeptide in the presence of single, double, and triple antibody combinations at 0.32 µM total antibody concentration. For single and double combinations, the remaining thirds of antibody were topped off with isotype control. Means and standard deviations of normalized IFN-γ secretions are shown, n = 4. The dotted line indicates IFN-γ secretion by non-exhausted T cells. * p < 0.05, ** p < 0.01, **** p < 0.0001 by one-way ANOVA with Tukey’s multiple comparisons test. (b) A dose-titration of triple antibody combination and isotype control was performed in the same assay as in (a). Means and standard deviations of normalized IFN-γ secretions are shown, n = 4. The dotted line indicates IFN-γ secretion by non-exhausted T cells * p < 0.05, *** p < 0.001, **** p < 0.0001 over isotype control at the same concentration by two-way ANOVA with Sidak’smultiple comparisons test. One of two representative experiments is shown.
a b
Figure 4. Maximum TGI was observed in the triple combination group in an EMT-6 syngenic mouse breast cancer model
Triple combination treatment targeting human PD-1, TIM-3, and LAG-3 using TSR-042, TSR-022, and TSR-03, respectively, was also most effective in an MDA-MB-436 triple-negative breast cancer model in humanized NOG-EXL mice.5
Isotype control α-PD-1 α-TIM-3 α-LAG-3
α-PD-1 + α-TIM-3 α-PD-1 + α-LAG-3 α-PD-1 + α-LAG-3 + α-TIM-3
TGI 24% TGI 6% TGI 11%
Figure 5. Triple combination treatment modulated multiple aspects of the tumor microenvironment
EMT6 tumors were harvested on day 20 from all remaining animals. Tumors were dissociated into single-cell suspension and infiltrating immune cells were characterized by flow cytometry. Frequencies of selected cell populations of interest are shown as means and standard deviations, along with values for individual animals (n = 7-9). ** p < 0.01, **** p < 0.0001 by one-way ANOVA with Dunnett’s multiple comparisons test vs isotype control.
T cellsM
yeloid cells
Figure 6. Triple combination treatment is able to reinvigorate TILs from human ovarian cancer patients
TILs were isolated from primary ovarian cancer resections from treatment-naïve patients. IFN-γ and IL-2 secretion was quantified after restimulation with 100 μg/ml SEB for 48 h in the presence of immune checkpoint inhibitory antibodies targeting PD-1 (TSR-042), TIM-3 (TSR-022), and LAG-3 (TSR-033). All three proprietary antibodies (TESARO, Inc.) are being evaluated in clinical trials6. Each antibody was used at 10 μg/ml and cytokine levels were normalized to isotype control conditions. Means, standard deviation, and individual data points are shown, n = 2-3. * p < 0.05, *** p < 0.001 over isotype control by one-way ANOVA with Dunnett’s multiple comparisons test.
IFN-γ IL-2
TGI 40% TGI 50% TGI 61%
Volumes of subcutaneous EMT6 tumors were monitored every other day during treatment with the indicated antibodies, administered biweekly alone or in combination depending on the treatment group (200 μg per antibody for each mouse). Tumor volumes are shown for each animal (one line per animal, n = 10) and tumor growth inhibition (TGI) was calculated.
Dendritic cellsNK cells
COAD, colon adenocarcinoma, n = 95; KIRC, kidney clear cell carcinoma, n = 54; LUAD/LUSC, lung adenocarcinoma and squamous cell carcinoma, n= 54; ESCA, esophageal carcinoma, n= 21; CESC, cervical squamous cell and adenocarcinoma, n = 11; UCEC, uterine corpus endometrial carcinoma, n = 10; STAD, stomach adenocarcinoma, n = 9; BLCA, bladder urothelial carcinoma, n = 7; HEP, liver cancers, 7; HN, head and neck carcinomas, n = 7; PAAD, pancreatic adenocarcinoma, n = 6; OV, ovarian adenocarcinoma, n = 5; BRCA, breast invasive carcinoma, n = 2; SKME, skin cutaneous melanoma, n = 2; PRAD, prostate adenocarcinoma, n = 1.
P365
Total immune infiltrate
