Sustained Intratumoral Administration of Agonist CD40 Antibody Overcomes Immunosuppressive Tumor Microenvironment in Pancreatic Cancer

Abstract

Agonist CD40 monoclonal antibodies (mAb) is a promising immunotherapeutic agent for cold-to-hot tumor immune microenvironment (TIME) conversion. Pancreatic ductal adenocarcinoma (PDAC) is an aggressive and lethal cancer known as an immune desert, and therefore urgently needs more effective treatment. Conventional systemic treatment fails to effectively penetrate the characteristic dense tumor stroma. Here, it is shown that sustained low-dose intratumoral delivery of CD40 mAb via the nanofluidic drug-eluting seed (NDES) can modulate the TIME to reduce tumor burden in murine models. NDES achieves tumor reduction at a fourfold lower dosage than systemic treatment while avoiding treatment-related adverse events. Further, abscopal responses are shown where intratumoral treatment yields growth inhibition in distant untreated tumors. Overall, the NDES is presented as a viable approach to penetrate the PDAC immune barrier in a minimally invasive and effective manner, for the overarching goal of transforming treatment.

Keywords: drug delivery; immunotherapy; implantable device; pancreatic cancer; sustained release.

Figure 1

KPC tumor‐bearing mice were treated with CD40 mAb via systemic (IP) or local (NDES) administration. A) Rendering of NDES components. B) Scanning electron microscopy (SEM) image of nanochannels in the nanofluidic membrane. C) In vitro cumulative release of CD40 mAb from NDES (n = 4). Data are expressed as mean ± standard deviation. D) Representative ex vivo tumor IVIS images (n = 6). E) Bar‐graph depicted the CD40‐AF700 signal. Data are expressed as mean ± standard deviation. Significance was analyzed by 2‐way ANOVA, ***p < 0.0005; ****p < 0.0001. Sidak correction was applied for multiple comparison. F) KPC tumor subcutaneous inoculation on right flank and treatment schematic. G) In vivo measurement of KPC tumor growth curve (n = 8/group). H) Bright field pictures of ex vivo tumors on day 14. I) Ex vivo tumor weight of all groups on day 14. J) Percentage change in tumor volumes on day 14 relative to their respective baselines (day 0). Each bar represents one mouse tumor. Data are expressed as mean ± standard deviation. Significance was analyzed by 2‐way ANOVA, ****p < 0.0001. The detailed significance between groups provided in Table S3, Supporting Information.

Figure 2

Histological analysis of tumors and treatment‐associated adverse events. A) H&E (top panel) and Masson's trichrome (bottom panel) staining shows the tumor composition of cancer cells (yellow arrows), fibroblast (blue staining), and infiltrating immune cells (white arrows). The length of scale bar for the bottom left image is 50 µm. B) Serum cytokine levels on day 1 as an indication of cytokine storm. n = 4/group. Data are expressed as mean ± standard deviation. Significance was analyzed by unpaired t‐test, *p < 0.05, **p < 0.005. C) Liver histopathology scoring. D) Representative H&E staining of liver samples. Black arrows indicate the inflamed loci.

Figure 3

Tumor‐infiltrating immune cells assessment, including myeloid cells and lymphocytes on day 14. Myeloid cells population: A) DC, B) activated CD80+ DC, C) M1, and D) M2 cells. Lymphocytes population: E) CD8 effector T cells, F) CD4 effector T cells, G) IFNγ+ expressing CD4 T cells, and H) FoxP3+ Tregs via flow cytometry analysis. Data are expressed as mean ± standard deviation. Significance was analyzed by one‐way ANOVA. *p < 0.05; **p < 0.005; ***p < 0.0005; ****p < 0.0001. Tukey's correction was applied for multiple comparison.

Figure 4

IMC analysis of TIME. Representative clusters from PhenoGraph. A) CD163+PD‐L1+ (M2 macrophages); B) B220+ (B cells); C) Foxp3+ (immunosuppressive intracellular marker); D) CD8+PD‐1+ (CD8+ exhausted T cells); E) PanCK+ (cancer cells); F) αSMA+ (endothelial cells). Cell density was averaged across 3 random fields of view per tumor (n = 2/group). Significance was analyzed by one‐way ANOVA. *p < 0.05; **p < 0.005; ***p < 0.0005. Tukey's correction was applied for multiple comparison.

Figure 5

Neighborhood analysis of cellular interactions from IMC are presented as a heatmap. The color panel indicates significantly neighbored (red) or avoided (blue) cell‐cell interactions in A) UnTx, B) IP CD40, and C) NDES CD40 groups. Highlighted squares indicate directional interaction between cell phenotypes. Phenotypes #5 and #11 showed avoidance with phenotype #13 (green boxes) in NDES group compared to UnTx and IP CD40 groups. Phenotype #15 (CD8+PD1+) showed enrichment in cancer cells (#21, PanCK+) in UnTx TIME compared to NDES CD40 and IP CD40 groups (black circle). Phenotypes #11, 12, 13 (CD4+, CD4+PD1+, Foxp3+, respectively) showed significant enrichment in cancer cells (#21), in the UnTx group (dotted rectangle).

Figure 6

Combinational treatment efficacy of single dose of radiation and CD40 mAb treatment in KPC mice. A) Tumor growth curve (n = 7–8/group). Data are expressed as mean ± standard deviation. Significance was analyzed by two‐way ANOVA, ****p < 0.0001. The detailed significance between groups on each day was provided in Table S4, Supporting Information. Immune cells population in combination treated KPC mice: B) M2 macrophages in TdLN, and activated DCs in C) TdLN and D) tumor. Data are expressed as mean ± standard deviation. Significance was analyzed by one‐way ANOVA. *p < 0.05; **p < 0.005; ***p < 0.0005; ****p < 0.0001. Tukey's correction was applied for multiple comparisons.

Figure 7

Bilateral KPC tumor model for assessing abscopal response. A) KPC tumor inoculation and treatment schematic on C57BL/6 mice. B) Growth curve of primary (treated) tumors. C) Growth curve of distant (UnTx) tumors. Data are expressed as mean ± standard deviation. Significance was analyzed by 2‐way ANOVA. ****p < 0.0001. The detailed significance between groups on each day is provided in Table S5, Supporting Information. Tukey's correction was applied for multiple comparison. D) Kaplan–Meier survival rate of UnTx, NDES, and IT groups. Significance was analyzed by log‐rank test; n = 8/group; **p < 0.001; ***p < 0.0005).

Figure 8

Panc02 tumor model for assessing CD40 mAb treatment administered via IP or NDES. A) Panc02 tumor inoculation and treatment schematic of C57BL/6 mice. Radiated groups received single 5 Gy dose one day prior to CD40 mAb treatment. B) In vivo tumor growth curve of each group. Significance was analyzed by 2‐way ANOVA and compared to UnTx group. ****p < 0.0001. IP CD40 versus NDES CD40 and Rad only: ****p < 0.0001; NDES CD40 versus Rad + IP CD40: ****p < 0.0001; NDES CD40 versus Rad + NDES CD40: ***p < 0.0005; Rad only versus Rad + IP CD40: ****p < 0.0001; Rad only versus Rad + NDES CD40: **p < 0.001. The detailed significance between groups on each day is provided in Table S6, Supporting Information. Tukey's correction was applied for multiple comparisons. C–H) Individual tumor growths of each group (N = 8/group).